Judge LIVINGSTON concurs in part and dissents in part in a separate opinion.
POOLER, Circuit Judge:Appeal from United States District Court for the Eastern District of New York (Joanna Seybert, /.) judgment in favor of plaintiffs-appellees John Restivo and Dennis Halstead following a jury trial. Defendant-appellant Carolann Hessemann, as executrix of the Estate of Joseph Volpe, (“Volpe”) challenges the district court’s grant of Restivo and Halstead’s motion for a new trial; several evidentiary rulings at second trial; the district court’s holding that Volpe was not deprived of a fair trial because his counsel was allegedly operating under a conflict of interest; the district court’s denial of Volpe’s motion for remitti-tur and a setoff of damages; and the district court’s grant of attorneys’ fees to counsel for Restivo and Halstead. For the reasons that follow, we affirm the judgment of the district court.
BACKGROUND
In 1984, sixteen-year-old Theresa Fusco was raped and murdered in Nassau County. As will be described in greater detail below, then-21-year-old John Kogut confessed to the Nassau County Police, stating that he, John Restivo, and Dennis Hal-stead had participated in the rape of Ms. Fusco, and then murdered her. All three men were charged with the rape and murder. Restivo and Halstead were tried jointly, and in 1986, a jury found both men guilty of rape and second degree murder; they were sentenced to 33-and-one-third years to life. Kogut was tried separately, and was convicted by jury; he was sentenced to an indeterminate term of imprisonment of 37-and-one-half years to life.
*553But in 2003, DNA testing was conducted on a slide containing cells taken from a swab of the victim’s vagina after her body was found. The testing revealed the presence of DNA from two individuals: Ms. Fusco and an unknown man. The testing excluded Kogut, Halstead, and Restivo from having contributed that DNA. The DNA also did not match that of 86 other individuals known to Ms. Fusco. Based on this evidence, Halstead, Restivo, and Ko-gut’s convictions were vacated. The State went to trial again against Kogut, primarily based on his confession, but, when he was acquitted, the State dropped the charges against Restivo and Halstead. Ultimately, Halstead and Restivo spent eighteen years in prison—two pretrial and sixteen following conviction—for these crimes. This civil suit, alleging malicious prosecution and denial of fair trial claims, among other claims, followed.
1. The Crime, Investigation, and Criminal Trial
In 1984, sixteen-year-old Theresa Fusco went missing. She was last seen on November 10, 1984 at 9:47 PM, when she clocked out of her job at Hot Skates, a roller rink, in Lynbrook, New York. A missing persons investigation began the next day, with then-Nassau County Homicide Detective Joseph Volpe acting as lead detective.2 Several weeks after Ms. Fusco went missing, on December 6, 1984, her body was found, naked, near the Lynbrook railroad tracks. As the medical examiner testified, the autopsy showed that she had been raped and that the cause of death was ligature strangulation.3 The medical examiner also determined that Ms. Fusco was most likely strangled with a rope, approximately one inch thick. On cross-examination, the medical examiner testified, based on the injuries to Ms. Fusco, that she believed that Ms. Fusco was strangled with pressure applied to the front of her neck with “rope or ligature, going from the back toward the neck, possibly crisscrossing with a lot of movement, and then both hands extending in the back of the neck.” Trial Tr. at 719.4 She further testified that the rope “should be less than six feet, depending on how it was applied” because “both ends of the rope never touched the skin in the back,” and that she would expect that there would be blood on the rope visible to the naked eye. Trial Tr. at 720-22. On redirect, she testified that, considering that thickness of the rope, it would have been easier to strangle Ms. Fusco if the rope were shorter, but clarified, “I don’t know the length of the rope. I have no idea.” Trial Tr. at 726.5
In late January 1985, Volpe heard that a man named Harry Smyle had been making suspicious statements about the murder. After an eleven hour interrogation and being told that he was a suspect, Smyle *554told police that Restivo had made a suspicious statement to him about the murder. Volpe then picked up Restivo. Volpe and Robert Dempsey, another Nassau County detective, interrogated Restivo for approximately eight hours over March 5, 1985 and March 6,1985 and allegedly physically assaulted him, Restivo then signed a statement, which stated that he heard Halstead make admissions about the Fusco murder but that did not implicate Restivo. Specifically, the statement said:
I would like to say that sometime back possibly November, December 1984, I stopped by my friend Dennis Halstead’s apartment. He lives above the store on Atlantic Avenue.
[[Image here]]
When I saw him, I realized that he was also high. We were talking about 10 to 15 minutes and at this point and kind of out of the blue Dennis started to talk strange.
He started talking about a broad. Dennis said he was with a broad, a girl, and that he was either by a cemet[e]ry, in a cemet[e]ry, across from the cemet[e]ry. He said he tried to fuck her. Then he had to fuck her up. But when he said that, he didn’t tell me how he fucked her up. He then told me that he strangled her and killed her.
Appellees’ Supp. App’x at 270. After Resti-vo was released, he contacted his attorney, Theodore Robinson, who called the homicide bureau and informed them he was representing Restivo and Halstead, and objected to the interrogation as coercive. On March 21, 1985, police officers came to the home of John Kogut, who worked with Restivo and Halstead, and asked him to come to the police station for questioning in relation to Ms. Fusco’s murder. Kogut complied; he denied knowledge of the crime, but agreed to come back on March 25. He returned on March 25, and was interrogated through the night. Kogut testified that Volpe and Dempsey screamed at him, threatened him, and told him that they had scientific evidence, witnesses, and statements demonstrating that he committed the crime. According to Ko-gut, Volpe stated to him, “I’m going to tell you how you did it because I already know how it happened.” Appellees’ Supp. App’x at 197. Volpe then told various stories, until Kogut finally gave in and agreed; Volpe wrote out a confession, which Kogut signed at approximately 9:00 AM on March 26. The written statement said:
My name is John Kogut[.] I am 21 years of age, being born on 11-29-63. I live with my girlfriend Lisa Price and her father at 161 Traymore BLVD, Island Park. I am currently employed by Frank Bertini Landscaping, 2988 Bay View Court, Oceanside, phone RO-4-7483.
I have been told by the detective that I have the right to remain silent and that any statements I make may be used against me in court. I have been told that I have the right to talk with a lawyer before answering any questions or to have the lawyer present at any time. Further, I have been advised that if I cannot afford to hire a lawyer, one will be furnished [to] me and I have the right to keep silent until I have had the chance to talk with a lawyer.
I understand my rights and make the following statement freely and voluntarily. I am willing to give this statement without talking with a lawyer or having one present.
I would like to say that sometime back in Nov 1984, about a week or two before my birthday I was with these two guys. One was John Restivo, who I worked for[,] and the other was Dennis Hal-stead, a guy from Lynbrook I know, who also worked for Restivo. On this night, it *555was between 8:00 pm and 10 pm, I was with John Restivo and Dennis Halstead in John’s van. It’s a blue [Fjord step van, with a side door that opens. This door is located back from the right front passenger door. I seem to remember that we were coming down from a moving job, possibly Hempstead, and we were drinking beer and smoking pot. John was driving, Dennis was in the right front passenger seat, and I was sitting on a cushion seat right behind the passenger seat. We were coming from East Rockaway where John keeps his trucks up Ocean Ave. [A]t the intersection of Merrick Road John turned left. We were heading west on Merrick Road and the cemet[e]ry was on our right side. At this point there’s a girl walking on Merrick Road by the cem-et[e]ry and heading towards McDonalds. John pulls up and stops along side her. John and Dennis were saying let[’]s see if she wants to party, meaning maybe smoke or have some beers. I felt that one of them knew her by the way they were talking, but I can’t remember her or their exact words. When I heard this conversation, I moved up in the seat to see the girl and hear what[’]s happening. Either John or Dennis invited her in, not to party, but for a ride home. I then opened the side door and I see this girl. She was about 15 or 16 years old, dark hair, medium long. She had on a blue denim dungaree jacket, I think a dark top, dark pants and white high top sneakers.
She knelt down in between Dennis and John. I was sitting behind her. Dennis, as John’s driving around, says to the girl “You want to party,” “forget about getting fired,” “do you want to do the right thing.” Do the right thing on the street means to get laid. With that she said, “stop the fucking van,” let me out. Dennis then jumped out of his seat and grabbed her. The girl started screaming “leave me alone, let me out.” She was fighting Dennis, but he was to[o] big for her and had a good grip on her. As I moved up to where she and Dennis were, she turned on me and smacked me in the face. With this I freaked out, I got crazy and I punched her with my left fist, hit her on the right side of her face. She falls out of Dennis’[s] grip to the floor of the van. I would like to say that I hit her with a left and a right before she fell from Dennis[’s] grip. I jumped on her upper body and she was trying to throw punches and kicks. At this point Dennis started taking off her pants and underpants, and I was taking her jacket shirt and bra. I’m telling her to shut up, calm down, and I realize that Dennis had put his penis inside of her. While Dennis fucked her I held her upper body down. At this point she wasn’t fighting to[o] much. By this time, I mean after Dennis fucked her, we were already in the cemet[e]ry. John had stopped the van and yelled back to me and Dennis, “let me get a piece.” I looked down at her and she was almost unconscious, I mean, she was fainting. Dennis pulled his pants up and was sitting in a seat by the passenger seat, John was now fucking her while she was unconscious, and I got out the side door and I took a blanket out that was in the van. I spread it out on the ground. This blanket was quilted with a different col- or border. Dennis and John then carried her out of the van and laid her face up on the blanket. I remember she had some jewelry on. I recall seeing a gold colored chain with what looked like a double heart on it with a piece broken off of it. I think there were other charms on the chain but I don’t remember what they looked like. I then ripped the heart and chain off and I put it in my pocket. *556John was in the van pulling his pants up and putting his shoes on, and Dennis was taking the rest of her jewelry and rings off. Now she starts to regain consciousness. She was a little dazed, but she was saying “I gotta tell,” “I gotta tell,” and she was crying. She was still laying on the blanket with no clothes on. Dennis, John and I decided she had to be killed. We were afraid she would tell on us. She started to come to and she was getting frantic. I got on top of her, put my knees on her shoulder, and covered her mouth. My back was to John and Dennis and one of them threw me a rope. One of them said “Do what you gotta do,” but first, before the rope was thrown to me Dennis, while standing over her told her, “You have to die.” They both went back to the van. I took the rope, which was a hard nylon type. I wrapped it double around her neck, then I twisted it like a cork screw. I twisted it for a few minutes until her body went limp, and I felt she was dead. I rolled her body up in the quilt and I threw her over my shoulder and to the van. I threw the rope into the van and then I dropped the body into the floor of the van. I got into the van, closed the door, and John took off out of the cemet[e]ry. Dennis was sitting up front. As we were leaving the cemet[e]ry we discussed that we had to get rid of the body. While we were driving and talking, I was putting the clothing into a plastic garbage bag. In addition to her clothing, I took her pocketbook and put it into the bag. It was maroon or black with a strap. I don’t remember what I did with the rope, and I left the plastic bag with her clothing in it behind the driver[’]s seat. I told John he had better get rid of the clothing and he said he’d take care of it. We then went down a street across from the cemet[e]ry that dead ends by the railroad tracks. I don’t remember who picked out the street but it was dark there. John [and] Dennis took her body out the side door and still wrapped in the quilt, they walked into the wooded area along the tracks and I was behind them. It was very dark and very heavy brush. John was carrying her by her head and walking backwards and Dennis had her feet. It seemed like 5 or 10 minutes to walk in before we got to an opening. At this opening we saw some wooden [pallets]. When they put her down, they both pulled the blanket and she rolled out. She ended up face down. We all started kicking leaves on her and I suggested we cover her with the [pallets], which we did. At this point I told Dennis and John I was leaving. As I started out from where her body was, I threw the jewelry I had by the tracks. I walked along the tracks until I came to a street. I turned right and I was on Sunrise [Highway].
I walked down Sunrise to Union Ave. I crossed over Sunrise on Union to Oakland. At this time I was living with my friend Brian Skellington at 66 Oakland Ave, Lynbrook. The front door was open, I went and went to sleep.
I am presently at the Homicide Squad, I have given this statement to Det. Volpe who has written it for me, I’ve read it and it is the truth.
App’x at 314-20; see also First Trial Tr. at 1383-1413. Kogut also gave a videotaped confession, confirming that he had been advised of and understood his constitutional rights and repeating the above story in material part. Kogut was then arrested and indicted on three counts: first-degree rape; second-degree murder in the course of rape; and intentional second-degree murder. Following a jury trial, Kogut was convicted on all counts and was sentenced to serve 37-and-one-half years to life in prison.
However, even with Kogut’s confession, police recognized that they would have a *557very weak case against Restivo and Hal-stead unless Kogut were ultimately to testify against them at trial (which he did not). See Appellees’ Supp. App’x at 225-26 (Detective Spillane, Volpe’s supervisor, testifying that, as of March 26, after Kogut confessed and was arrested, the case against Halstead and Restivo was “very weak if not nonexistent”). Restivo’s van was seized pursuant to a warrant soon thereafter. Certain hairs were allegedly found in Restivo’s van pursuant to this search. Several of these hairs, the “questioned hairs” or “Q hairs,”6 were determined to be consistent with Ms. Fusco’s hairs.7 These hairs provided the only forensic evidence linking Restivo and Hal-stead to the crime.
Restivo and Halstead were then also arrested and indicted with counts of first-degree rape, second-degree murder in the course of rape, and intentional second-degree murder. Restivo and Halstead were tried jointly. As will be discussed below, Kogut’s confession was inadmissible against Halstead and Restivo at their criminal trial, because Kogut did not testify. Nor was Restivo’s statement implicating Halstead introduced at trial. Instead, the State introduced the Q hair evidence and evidence of statements allegedly made by Restivo or Halstead to third parties. Restivo and Halstead were convicted of all counts, and each was sentenced to 33-and-one-third years to life.
II. The DNA Evidence
Beginning in 1993, samples of semen obtained from Ms. Fusco’s body were subjected to DNA testing. Initial tests excluded Halstead, Restivo, and Kogut as the source of the semen, but plaintiffs’ motions to vacate their convictions were denied. A new round of DNA testing was conducted in 2003, which again demonstrated that neither Halstead, nor Restivo, nor Kogut was the source of the semen found in Ms. Fusco’s body. Thereafter, on June 11, 2003, all three men’s convictions were vacated. The Nassau County District Attorney retried Kogut, based on his confession and, in December 2003, Kogut was acquitted. The District Attorney then moved to dismiss the indictment against Restivo and Kogut, on the ground that the DNA evidence disproved the state’s theory of the case. The indictments were dismissed on December 29,2005.
III. Civil Suit
Kogut filed suit on December 19, 2006, and Restivo and Halstead filed suit on December 21, 2006. The district court consolidated the suits, which alleged numerous constitutional claims pursuant to 42 U.S.C. Section 1983, as well as state law claims related to the Fusco investigation and their prosecution.8
A. First Trial
Following an extensive Daubert hearing, which will be discussed below, and motion *558practice, the case proceeded to trial on claims that defendants Joseph Volpe, Robert Dempsey, Frank Sirianni, Thomas Allen, Wayne Birdsall, and Charles Fraas maliciously prosecuted Restivo, Halstead, and Kogut and deprived them of their constitutional right to a fair trial, as well as a failure to supervise claim against defendant Shaun Spillane9 and a municipal liability claim against Nassau County. The jury found for defendants on all counts.
Thereafter, all three plaintiffs moved for a new trial on various grounds. The district court granted Restivo and Halstead’s new trial motion, but denied Kogut’s.10 In particular, the district court ruled that it had erred in its treatment of Kogut’s confession as to Restivo and Halstead. At trial, the confession was admitted without any limiting instruction to establish probable cause for the prosecution of Restivo and Halstead. The district court concluded that it was error to allow the jury to consider the confession to establish probable cause for Restivo and Halstead’s prosecution because the confession was inadmissible against them in their criminal trials, relying on Boyd v. City of New York, 336 F.3d 72 (2d Cir. 2003). The district court read Boyd as holding that defendants cannot use evidence that was inadmissible in a criminal case to establish probable cause to prosecute. The court concluded that its error affected Restivo and Halstead’s substantial rights, given that the confession “was arguably one of the central pieces of evidence in this case.” Special App’x at 62.
B. Second Trial
The case then proceeded to a second trial, at which Restivo and Halstead greatly narrowed their claims. At the second trial, plaintiffs claimed that Volpe and Fraas, the hair evidence technician who worked on the Fusco investigation, unconstitutionally deprived them of a fair trial under Section 1983 and that Volpe and Fraas maliciously prosecuted them under Section 1983 and New York state law. Both claims were based on the same two factual theories: that Volpe suppressed exculpatory evidence and/or that Volpe, either independently or with the assistance of Fraas, planted the hairs allegedly found in Restivo’s car.
Pursuant to its ruling that “in attempting to show probable cause, Defendants [are] limited to evidence that was both included in their interrogatory responses and admissible in Halstead and Restivo’s criminal proceedings,” Special App’x at 61, and finding the confession irrelevant to the fair trial claims, the court excluded Ko-gut’s confession from coming into evidence at the second trial. The court also excluded Restivo’s statement to police regarding in-culpatory statements allegedly made by Halstead on the ground that the statement was inadmissible in the criminal trial.
Ultimately, the jury found that Volpe, but not Fraas, maliciously prosecuted plaintiffs and deprived them of them right to a fair trial. A special verdict was not requested, and the jury therefore did not specify whether it found against Volpe on the theory that he suppressed exculpatory evidence, planted the hair evidence, or both. Following a separate damages trial, the jury awarded each plaintiff a damages award of $18 million.
1. Evidence at Trial
We turn now to the evidence presented at trial.
*559a. Plaintiffs’ Claims of Innocence
Plaintiffs were not required to prove their innocence to win on their claims at trial. However, the evidence of their innocence provides an important backdrop for their claims at trial. As noted, the DNA testing in 2003 demonstrated that neither Restivo, nor Halstead, nor Kogut contributed the semen found in Ms. Fusco after her death. Volpe asserts that this does not establish plaintiffs’ innocence, and speculates that plaintiffs raped the victim without ejaculating or while wearing condoms and that the sperm came from a consensual sexual partner. Volpe baselessly speculates that, although there was no evidence that Ms. Fusco was sexually active, she may have been having sex with her boyfriend, who was black, and she may not have told her friends because “it was socially unacceptable for white girls to date black guys on Long Island in the 1980s.” Appellant’s Reply Br. at 4. However, not only does this speculation lack any basis in record evidence, it is also rendered implausible by testimony at trial.
First, plaintiffs’ expert witness, Dr. Charlotte Word, testified about the DNA evidence in this case. She testified that vaginal swabs and smears taken from Ms. Fusco at her autopsy and later put on slides showed DNA from two sources. First, there was Ms. Fusco’s DNA, demonstrating that these slides were taken from her and were not mislabeled or misplaced. Second, there was semen from only one male donor. As noted, the testing on the semen definitively excluded Restivo, Ko-gut, and Halstead as a source. The semen was also tested against DNA profiles of 86 individuals known to Ms. Fusco, including various people she had dated, and there were no matches. Nor was there a match when the semen was tested against the Combined DNA Index System database kept by the Federal Bureau of Investigation. Further, Dr. Word testified that there was a large quantity of sperm present and that the DNA in the sperm head was well preserved. The quantity of sperm indicated that the “murder occurred very soon after the deposition of the sperm.” Trial Tr. at 435. Additionally, Dr. Word testified that the Nassau County medical examiner’s office records indicate that a very large number of sperm were collected from the vaginal cavity on the swabs and, “when looked at microscopically, two different individuals were able to see that the tail portion of the sperm was also still attached to the sperm head.” Trial Tr. at 436. Dr. Word testified that “the presence of tails on the sperm means that the sperm had not been in the vaginal cavity for a very long time[,] [b]ecause normally, in a normal healthy female, probably within six hours or less, maybe at the outside ten to twelve hours, those sperm tails are lost.” Trial Tr. at 436. In sum, she testified that the sperm was deposited “[p]robably no more than eight to twelve hours, but more likely less than six” hours of the time of her death. Trial Tr. at 443. Furthermore, the physical evidence rebutted defendants’ hypothesis that plaintiffs raped Ms. Fusco while wearing condoms after a consensual sexual partner deposited the semen. In particular, Dr. Word gave the following explanation:
[T]he physical act of two [or] more, or whatever more of vaginal penetration and sexual intercourse with her [even with condoms] would ... contribute to the loss of sperm from the vaginal cavity. So under that scenario, my expectation would be that we would have a much smaller number of sperm present on the samples collected. It wouldn’t be consistent with that hypothesis.
Trial Tr. at 444.
Nor is there any evidence that Ms. Fus-co was sexually active. The undisputed ev*560idence, based on notes from the investigating detectives, Ms. Fusco’s diary, Ms. Fusco’s mother, and Ms. Fusco’s best friend, was that Ms. Fusco was a virgin prior to the rape. Moreover, as noted, DNA testing done on 86 individuals, including various boyfriends and ex-boyfriends of Ms. Fusco, did not reveal a match.11
Further, defendants introduced only limited evidence at trial to establish Restivo and Halstead’s guilt. Indeed, the primary evidence they sought to introduce to establish plaintiffs’ guilt was Kogut’s confession and Restivo’s statement implicating Hal-stead; as discussed in detail below, both of these statements were hearsay that the jury could not consider for the truth of the matters asserted, that is, for the truth of whether Halstead and Restivo committed the crimes. The only other evidence introduced at trial tending to show their guilt was the testimony of Restivo’s friend Michael Cockerel, who testified that, before Ms. Fusco’s body was found, Restivo commented to him that the body would probably be found at Lynbrook Cemetery; that Restivo’s brother assaulted him, and was then charged with witness tampering; that, when he asked Restivo if he committed the crime, Restivo responded, “Didn’t matter who did it. She was a black person lover,” Trial Tr. at 2430; and that, the day after Ms. Fusco’s disappearance, Restivo told him that he “got his dick wet” the prior night, Trial Tr. at 2431. But on cross-examination, Cockerel admitted that he gave various statements to police that had contradictory information, that he lied to the grand jury, and that he was interviewed by police for more than ten hours, during which time poliee accused him of being involved in the crime.12
b. Suppression of Exculpatory Evidence
We now turn to the theories underpinning plaintiffs’ malicious prosecution and denial of fair trial claims. First, plaintiffs claim that Volpe suppressed exculpatory evidence regarding a lead that Volpe developed—the “French lead.”
Volpe testified at his deposition that, when she went missing, Ms. Fusco was wearing dark and light striped blue jeans with stripes two to three inches apart, and that the jeans were tapered at the ankles with cuffs. Accordingly, a missing person flier, which stated that the victim was wearing striped blue jeans when she went missing, was posted around town.
Shortly after Ms. Fusco’s body was found, a man named John French called to alert the police that his car had been stolen near Hot Skates around the same time of Ms. Fusco’s disappearance.13 Speciflcal*561ly, the report, which was made on December 6, 1984, stated that French’s car had been stolen between 9:30 PM and 11:05 PM on November 10,1984 less than a mile away from Hot Skates.14 French found his car about a week later, before Ms. Fusco’s body was discovered. He found the car near railroad tracks in Lakeview, near Woodfield Road, approximately one and one-half to two miles away from where it had been stolen. When he found the car, the license plates had been changed. Additionally, his sister, Lori French Gabberty, found a pair of women’s or girls’ jeans with stripes under the right passenger seat; at least one leg of the pants was inside out. These jeans were not in French’s car before it was stolen. The jeans were discarded inadvertently by the Lynbrook Police15 before Ms. Fusco’s body was found. French also stated that, when he found his car, a rope that had previously been there was missing. The car was taken for processing, and was combed for forensic evidence. Volpe’s handwritten notes indicate that he got hair samples from people who may have had access to French’s car, and that investigators found a hair that was “more than 50% similar to deceased & less than 100% similar” to Ms. Fusco’s in the car. Appellees’ Supp. App’x at 1569.
Two documents were created as a result of Volpe’s investigation, Plaintiffs’ Exhibits 161 and 163. Exhibit 161 is a report providing information about the location and time of the theft of French’s car, where the car was found, and the discovery of women’s blue jeans with stripes, inside out under the seat. Exhibit 163 is a statement from French, wherein he describes being shown a length of rope and a brown felt pouch, which he identified as belonging to him. See App’x at 375 (photograph of rope).
At his deposition, Volpe testified that he thought that the French lead was Brady material, but that he turned it over to the District Attorney. There is evidence, however, that Volpe in fact did not turn over the French lead material. This evidence included testimony from Restivo that he did not know about this evidence at the time of his trial, testimony from Restivo’s criminal trial counsel that he did not know about this evidence at the time of trial, evidence that Volpe did not disclose that striped jeans were found in French’s car or that a rope was missing when he informed his supervisor, Lieutenant Spillane, that he was closing the French lead, testimony from the Assistant District Attorney on the case that he did not remember whether Volpe had told him about the French lead, and testimony from Judge Edward W. McCarty, who was the Assistant District Attorney on the case, that he did not remember being told about the French lead at the time of the investigation and that he never saw Exhibits 161 or 163 or learned the information contained therein before the civil trial. The parties also stipulated that Nassau County District Attorney’s Office Chief of Appeals Bureau Peter Weinstein, who oversaw the State’s response to Halstead and Restivo’s requests for post-conviction relief would testify that “[t]o the best of [his] knowledge, th[e] information about the French car/striped jeans lead was never disclosed by police to the prosecutors.” Appellees’ Supp. App’x at 1736. And the county attorney previously representing defendants in the civil suit testified that when she received the DA’s files and reviewed them, she did not recall seeing Exhibits 161 or *562163 or other evidence about the French lead.
Further, although Volpe admitted at his deposition that he knew the French lead was exculpatory, his attorney now argues that it was not, relying on the following evidence. First, at the civil trial, Gabberty testified that the jeans she found in 1984 had stitching, and she could not remember if there were stripes. But in 1984 she told police that the jeans were striped, and she testified at the civil trial that her memory was much fresher at the time she first spoke to detectives than at the civil trial decades later. Volpe also argues that the rope could not have been used in the murder because it had no blood on it and the medical examiner testified that the rope was too long, relying apparently on the image of the rope in evidence. But at trial, as noted, the medical examiner testified that although using a shorter rope would have been more efficient at killing Ms. Fusco, she “[doesn’t] know the length of the rope; [she] [has] no idea.” Appellees’ Supp. App’x at 568. Further, the rope went missing at some point and was not tested for blood.
The materiality of the French lead was also reinforced by testimony from plaintiffs’ police practices expert, Russell Fischer. Fischer worked in law enforcement for 33 years, ultimately achieving the rank of chief of criminal investigations in the Miami-Dade Police Department, one of the seven largest police departments nationwide. Fischer now works as an expert witness and assists in training law enforcement officers. At trial, Fischer testified to minimally accepted police practices, “a baseline type of behavior or protocols that law enforcement officers follow,” which differ from best practices. Trial Tr. at 1776. He testified that these minimally accepted police practices derive from a number of sources including court decisions, articles written by police practitioners, model policies produced by the International Association of Chiefs of Police as well as other groups, as well as practices from major police departments.
Fischer testified that, based on the evidence he had, the French lead was clearly exculpatory, and should have been documented and sent to a prosecutor, based on minimally accepted police practices in the 1980s. In particular, Fischer concluded that based on the time and location of the theft, the fact that the car’s license plates were removed, the fact that a rope was in the back seat at the time of the theft and a rope was used as a ligature to strangle Ms. Fusco, and the fact that a pair of women’s or girls’ striped jeans were found in the back seat of the vehicle, with at least one leg turned inside out, this lead was clearly exculpatory. Fischer testified that under minimally accepted practices, exculpatory evidence does not need to definitively prove innocence in order for it to have to be documented and disclosed, and that the combination of all of these factors meant that the evidence should have been disclosed under minimally accepted police practices.
c. Planting of Hair Evidence
The second theory underlying plaintiffs’ malicious prosecution and fair trial claims was that Volpe, either alone or with the assistance of Fraas, planted the Q hairs belonging to Ms. Fusco that were allegedly found in Restivo’s van. According to plaintiffs’ theory of the case, Volpe removed several hairs from an envelope containing hairs removed at Ms. Fusco’s autopsy and planted them in an envelope containing hairs collected from Restivo’s van. By contrast, Volpe contends that Restivo and Halstead are actually guilty of the crime, and Ms. Fusco’s hairs fell out in Restivo’s van during the commission of the crime.
*563Several pieces of evidence support plaintiffs’ theory of the case. When Ms. Fusco’s body was brought to the Medical Examiner’s office, after photographs were taken, her hair was washed with water and the debris and leaves were removed. One of plaintiffs’ experts testified that there was no significant mechanical damage to the Q8 or Q4 hairs or any debris on the hairs. Ms. Fusco’s known hairs removed at the autopsy were in the same condition, basically without debris or mechanical damage. By contrast, the other hairs removed from Restivo’s van, which were not consistent with Ms. Fusco’s hair, “had some mechanical damage, which is typical[ly] what happens to hair when it is laying around in an area where people work or walk or travel through, etcetera,” and “also had some debris on it.” Trial Tr. at 921.
Additionally, extensive expgrt testimony was given regarding the presence of postmortem root banding (“PMRB”) in the Q hairs, which testimony was the subject of a five-day Daubert hearing. PMRB is a type of decomposition that occurs in hairs that are attached to dead bodies. Plaintiffs sought to introduce expert evidence that the Q hairs displayed PMRB and that PMRB takes days, if not weeks, to develop. This evidence would tend to show that the Q hairs were planted autopsy hairs; because the medical examiner testified that Ms. Fusco was placed where she was found within an hour of her death, even if Ms. Fusco was in Restivo’s van after she had been killed, according to plaintiffs’ experts, the Q hairs still must have been planted because PMRB could not develop in her hair within an hour of her death.
Plaintiffs proffered three experts in the field of PMRB: Dr. Max Houck, a forensic anthropologist and trace evidence analyst; Nicholas Petraco, a former trace evidence analyst with the New York Police Department (“NYPD”) and a consultant for the NYPD’s Forensic Investigation Division; and Dr. Peter DeForest, who taught crimi-nalistics at the John Jay College of Criminal Justice for nearly forty years. Following a five-day Daubert hearing, the district court ruled on the admissibility of these experts and their testimony. The district court found that Dr. Houck, Petraco, and Dr. DeForest qualified as experts, rulings not challenged on appeal by Volpe.16
The district court summarized the evidence presented by plaintiffs’ experts at the Daubert hearing as follows. Dr. Houck defined PMRB as “an opaque ellipsoidal band which appears to be composed of a collection of parallel elongated air spaces near the root of a hair, appearing as a dark or blackened band in the hair shaft.” Kogut v. County of Nassau, 894 F.Supp.2d 230, 235 (E.D.N.Y. 2012) (internal quotation marks omitted). He explained that PMRB is an artifact of decomposition: hairs in the active growing stage when in the time of death “go through changes in their root ends related to the decomposition of the surrounding skin and follicle.” Id.
According to Houck, the transformation of the putrid root only occurs in roots that remain in the scalp of a decomposing body; the changes do not occur if the hair is plucked (or shed) prior to death and allowed to deteriorate. He asserts that, according to the literature on the topic, for a hair to exhibit PMRB three conditions must be met: the hair must have been (1) in the active growing phase prior to an individual’s death; (2) *564in the skin while the body was decomposing; and (3) in the decomposing skin for a minimum of 7 days. Based on Houck’s understanding of the prosecution’s theory of the Fusco Homicide, according to which Fusco was in Restivo’s van for perhaps less than an hour, the Q hairs could not have come from Fusco on the night she disappeared. Houck concludes: Based on the known and documented scientific clinical studies on postmortem root banding relating to its timing, description, appearance, and conditions for existence, there is no known mechanism or reasonable explanation for PMRB to appear in Ms. Fusco’s hairs that were allegedly left in the blue van.
Id. (alterations, citations, and internal quotation marks omitted).
Next, Petraco also believed that the Q8 hairs could not have come from Ms. Fusco before she died or during the time between her death and when her body was left in the woods. He “opined that PMRB only develops in hairs while they are attached to a decomposing body and that the banding takes at least 8 hours after death to appear,” citing “two instances in which PMRB was observed in hairs 8-10 and 10-12 hours after death, respectively,” the shortest intervals before which PMRB has been observed. Id. at 236. Petraco also “observed that the Q8 hairs exhibited banding patterns that are consistent with the patterns on ‘known’ hairs collected during Fusco’s autopsy,” and “because hairs do not continue to develop PMRB once they are removed from the scalp, it is extremely unlikely, and probably impossible that the Q8 hairs—if they really came from Fusco either before or shortly after she died—would exhibit PMRB consistent with [the] degree of banding seen on the autopsy hairs taken weeks after Fusco was murdered.” Id. (citation and internal quotation marks omitted). Finally, Dr. DeForest stated that PMRB is a recognized phenomenon in the scientific community of hair examiners, although scientists do not fully understand the mechanisms that cause it. See id. at 237. He concluded that, based on the degree of PMRB of the Q8 hairs, and the fact that it was “similar to the greatest degree” of PMRB of Ms. Fus-co’s known hairs from the autopsy—the “K hairs”—that “to a reasonable degree of scientific certainty that the Q8 hairs exhibiting PMRB came from the sample of known hairs taken at the autopsy of the homicide victim, Theresa Fusco.” Id. (alteration omitted). He also concluded that the Q4 hair exhibited PMRB beyond a reasonable scientific certainty. See id.
Considering Federal Rule of Evidence 702, Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), the district court concluded that “[t]he idea that PMRB takes several days to develop (and thus that it could not have developed in the short time Ms. Fusco was alleged to be in Restivo’s van) has not yet been established by scientific standards of proof.” Id. at 240. The court reached this conclusion because although this theory was falsifiable, it had not been tested. Further, although the experts agreed that PMRB takes days, not hours, to develop, this hypothesis was not firmly grounded in the sparse academic literature or scientific studies on the topic. Additionally, there were two cases in the 1980s where PMRB was observed between 8-12 and 10-12 hours after death respectively, which would contradict the theory that it takes days for PMRB to develop, assuming those reports were accurate. The court concluded therefore that the experts could not testify that their conclusions regarding the timing of PMRB or the ultimate issue of whether Ms. Fusco’s hairs *565were planted were sound to a reasonable degree of scientific certainty.
However, although the evidence did not pass muster as “scientific” evidence under Daubert, the court held that the witnesses could testify based on their “technical or other specialized knowledge” under Federal Rule of Evidence 702. Id. at 242-43. That is, the court reasoned, “much of what Plaintiffs’ Experts have to say is grounded in sound science, and the last leap—the timing—is justified by their training and experience.” Id. at 243. The court found that, aside from the timing issue, the testimony on PMRB is supported by many of Daubert’s indicia of reliability. The district court found that there was evidence that PMRB can be distinguished from other types of environmentally caused banding within an acceptable rate of error—99.5% accuracy when one person worked alone, or 100% when two examiners double checked each other’s work. This was reinforced by graduate student Alison Domzal-ski’s thesis, which stated that although environmental factors can cause changes to scalp hair roots, environmental banding is visually different, occurring at a different place on the hair than PMRB. Additionally, the experts’ opinions were consistent with the academic literature on the topic, and PMRB, as plaintiffs’ experts described it, is a generally accepted phenomenon within the forensic science community. In sum, the court concluded that “although these facts do not add up to scientific proof, they supply a reasonable basis for forensic experts to conclude that PMRB is an artifact of decomposition and that, consistent with the speed at which other effects of decomposition appear on a corpse, it does not appear immediately after death.” Id. at 244.
Before the second trial, plaintiffs moved for reconsideration on the ground that a new peer reviewed study on PMRB, the Koch study, addressed many of the deficiencies in the literature that previously troubled the district court. In that study, FBI and Bureau of Alcohol, Tobacco, Firearms, and Explosives criminalists working at the University of Tennessee’s “body farm” tested approximately 24,000 hair roots from 23 human cadavers permitted to decompose in varying environmental conditions. In cadavers outside on the ground, PMRB generally did not develop until 6-10 days after death. The earliest development of PMRB in the study, which occurred in a cadaver placed in the trunk of a car in August, occurred after 4 days. Upon reconsideration, the district court found that this new evidence addressed some of its concerns about PMRB evidence, but ultimately held, “You can bring it in and testify to it, but they can’t use the magical words ‘to a degree of scientific certainty.’ ” Appellees’ Supp. App’x at 440. Volpe challenges this decision on the grounds that the Koch report stated that it was unable to draw any conclusions as to the onset of PMRB. Appellant’s Reply Br. at 44. But the Koch study merely stated that the rate of putrefaction versus time and temperature could not be calculated.
At trial, plaintiffs’ experts testified regarding PMRB. Defendants raise issue with Dr. Houck’s and Petraco’s testimony, claiming that they violated the court’s ruling that they were not to testify that their opinion was to a reasonable degree of scientific certainty. In particular, Dr. Houck, in response to a series of hypothetical questions, stated that he did not think it was possible for Ms. Fusco’s hairs to have been left in Restivo’s van during the rape and murder because of the presence of PMRB, and because the best evidence shows that PMRB takes on the order of days to develop. In response to defendants’ concerns about Dr. Houck’s testimony, the district court charged the jury, “Ladies and gentlemen, certain witnesses are per*566mitted to testify as experts. The Court has ruled that such expert testimony, however you deem it or however you consider it, cannot be a substitute for your own independent evaluation of the facts in this case. Moreover, the witness is not testifying to a reasonable degree of scientific certainty.” Appellees’ Supp. App’x at 632. Volpe also takes issue with Petraco’s testimony that the presence of PMRB meant that the Q hairs could not have come from Ms. Fusco on the night she died, as well as his statement that he would testify about PMRB to a reasonable degree of scientific certainty, if he were permitted to. However, he later clarified on cross-examination that he was not testifying about the timing of PMRB to a reasonable degree of scientific certainty.
Volpe also sought to introduce evidence from an expert witness—Dr. Joseph B. Kadane. Dr. Kadane, a statistician, sought to testify that (1) postmortem root banding could not be distinguished from various types of ante-mortem root banding and that (2) the academic research that had been done to that point was insufficient to establish how long it took after death for PMRB to appear. With respect to the first question, he referred to the study by Dom-zalski, referenced above. Domzalski’s study cautioned that various environmental conditions can cause a type of root banding, which “could be confused” with PMRB, but she noted that environmental banding appears at a different location than PMRB. Kogut, 894 F.Supp.2d at 237. Based on Domzalski’s study, Dr. Kadane sought to testify: “[I]t is not unreasonable to suppose that the Q-hairs were also exposed to dirt in the van in which they were found. Since the mechanism(s) that lead to root banding is unknown, we are not in a position to determine whether the Q-hairs are pre- or post-mortem. Neither of these can be excluded.” Id. at 238. On the second issue, Dr. Kadane sought to testify that plaintiffs’ experts’ data fails to account for crime scene and autopsy delays, and that there is no statistical evidence to support plaintiffs’ experts conclusions on the timing of the development of PMRB. Id. With respect to Dr. Kadane, the court found that he “does not have the relevant expertise to offer a helpful opinion at trial.” Id. at 244. Specifically, the court found that although he was an accomplished statistician, he does not have more than a passing familiarity with hair microscopy and forensic science, and he had not conducted his own statistical analysis of PMRB. Id. at 244-45.
Next, defendants contend that there was no evidence that Volpe had the ability to plant the evidence. But there was evidence that he did have such an ability. The known hairs, or K hams, taken from the autopsy were kept in a set of clearly marked envelopes labeled “K” and with Ms. Fusco’s name, and the Q8 hairs were kept in an envelope stating “van” and “right front seat floor.” Appellees’ Supp. App’x at 1659-60. There was nothing on the envelopes stating how many K hairs were in each envelope, and therefore there was no way of knowing whether any of Ms. Fusco’s K hairs from the autopsy were removed from the envelope.
The envelopes containing the K hairs and the Q hairs were unsealed, and were kept by Fraas in the same unlocked cabinet above the table in the microscope room in the Scientific Investigation Bureau (“SIB”). The cabinet had a sliding glass door to close it, and did not have a lock. The envelopes were neither logged in and out nor resealed every time they were taken out. Further, although there was a log where Fraas and other technicians signed in and out when they arrived and left the lab, there was no log for visitors who had to be buzzed in to the lab; thus, if someone was buzzed in there was no docu*567mentation of that fact. Fraas testified that he could not recall specifically if Volpe came to the lab on March 26, but he may have- come then, or on March 27 or 28. Further, there was testimony that the door to the hair room was not always locked, so those who had access to the lab also had access to the hair room; accordingly those with access to the lab could have access to the unlocked cabinets where the Q and K hairs were kept. If a homicide detective, such as Volpe, buzzed to enter the lab, and if the person he asked for was in the lab, he would be let in without having to sign in. Depending on who greeted the homicide detective at the door, the detective would sometimes be escorted to where he was going, and sometimes would not be. According to testimony from Bird-sail, who worked in the lab as a serologist, homicide detectives were treated as trusted colleagues and were therefore generally not escorted in the lab. Further, when a homicide detective was in the lab and left, in 1984 and 1985, he would leave on his own, unescorted. And Volpe, himself, testified that he could visit the SIB at his leisure.
Finally, evidence regarding an affidavit sworn by Volpe supports the inference that Volpe may have planted the hams. On March 29, three days after the search of Restivo’s van, and two days after Fraas began examining the 80 hairs recovered from Restivo’s van, Volpe swore an affidavit in support of a search warrant stating, “A search of [Restivo’s] van has produced hair consistent with THERESA FUSCO’S and possible human blood.” Appellees’ Supp. App’x at 1556. Volpe testified that when he said “possible blood” he was “just a messenger from what [he] had heard from the lab,” which was that “it looked like it was blood.” Trial Tr. at 2573. When questioned about his statement that there was hair consistent with Ms. Fusco’s, Volpe stated that the source of that information was the lab, specifically, an oral conversation with Fraas. But when asked whether the microscopic examination had been finished when they had this conversation, Volpe testified, “I don’t believe it was even started. I think it was just a visual, you know, look-see.” Trial Tr. at 2574. Indeed, Fraas later testified that it took “a significant amount of time, days to weeks, to generate a conclusion” regarding the hairs.17 Appellees’ Supp. App’x at 714. In addition, Birdsall testified that he never told Volpe that he found possible human blood and that by midnight on March 26 he had made a preliminary determination that there was no blood in the van.
Fischer testified it would not be appropriate pursuant to minimally accepted practices for a homicide detective to swear an affidavit claiming possible human blood based on a stain, and rather that minimal practices require presumptive testing for blood, and that it would not be appropriate based on minimally accepted practices to swear out an affidavit saying “possible human blood” without talking to the person who did the testing. Similarly, Fischer testified that it would not be appropriate pursuant to minimally accepted practices to report that hairs were consistent without a report from a hair examiner that compared the hairs and found them similar, and that it would not be appropriate to report hair as consistent from merely looking at the hairs.
2. Verdict and Damages
On April 11, 2014, following the presentation of evidence and being charged, the *568jury found that Volpe, but not Fraas, unconstitutionally deprived Restivo and Hal-stead of their right to a fair trial and maliciously prosecuted them. A separate damages trial was then held, where extensive testimony was given about the effects of the incarceration on the men. The jury was instructed to determine an award that fairly compensates each plaintiff for the damages they suffered. The jury was instructed to award only compensatory damages, and only for the injuries that plaintiffs proved were proximately caused by Volpe's wrongful conduct. Further, the court instructed: “Although defendant Volpe was found liable for two causes of action, each plaintiff is entitled to one recovery. The damages that you award must be fair compensation for all of the plaintiffs damages, no more and no less.” Damages Trial Tr. at 579. On April 17, 2014, the jury awarded $18 million each to Resti-vo and Halstead.
C. Post-Trial Issues
1. Motions to Reduce the Jury Award
Following the trial, Volpe moved to reduce the jury’s damages award on several grounds, each of which was rejected by the district court.
Volpe first claimed that he was entitled to relief based on a prior settlement between plaintiffs and New York State. Before bringing suit in federal court, plaintiffs brought suit against New York State in the New York State Court of Claims pursuant to the Unjust Conviction and Imprisonment Act, New York Court of Claims Act Section 8-b. Under this law, individuals who can demonstrate that they were wrongly convicted of state crimes are entitled to compensation, regardless of whether there was any wrongdoing by a government official. Accordingly, plaintiffs sought compensation for the sixteen years they spent in jail following their convictions. Plaintiffs settled with the State for $2.2 million each. As is relevant on appeal, Volpe argued that allowing each plaintiff to keep the $18 million awarded by the jury, while also keeping the $2.2 million awarded in the State settlement, violates the rule against double recovery, and that the $18 million awards should be reduced by the amount of the settlements. The district court rejected this argument, concluding that a setoff in this case would violate federal policies underlying Section 1983.
Volpe also argued that the damage award was excessive and moved for remit-titur. The district court considered both the federal and state standards for exces-siveness, and concluded that the award was not excessive under either standard.
2. Attorneys’ Fees
Following the second trial and motion practice on this issue, the district court awarded plaintiffs’ attorneys at the law firm Neufeld Scheck & Brustin, LLP (“NSB”) $4,997,914.55 in fees. In assessing the reasonableness of NSB’s fee application, the district court emphasized that this case has been litigated for eight years, with twenty-three separate discovery requests and fifty-four days of depositions. The first trial lasted thirty-one trial days over eleven weeks, and included the presentation of forty-two witnesses. The second trial, which both followed and was followed by extensive and complicated post-trial motions, lasted seventeen trial days, and included the presentation of forty witnesses. There was extensive expert testimony, as well as litigation over the admissibility of expert testimony, culminating in a five-day Daubert hearing. Based on these eight years of work, NSB requested compensation for 11,222.6 hours *569of work.18 NSB requested that attorneys be compensated at the rates prevailing in the Southern District of New York; specifically, NSB requested attorney rates ranging from $250 an hour to $700 an hour, and paralegal rates of $125 an hour, NSB requested that, if the district court applied Eastern District rates, the court apply rates ranging from $200 to $600 an hour for attorneys, and paralegal rates of $100 an hour. Ultimately, the district court granted NSB’s fee application, reimbursing NSB for 11,222.6 hours of attorney and paralegal work at the Southern District of New York rates it had proposed.
3. Indemnification
Before passing away, Volpe agreed to have the County Attorney represent him. Initially, the Nassau County Police Indemnification Board determined that Volpe and all other officer-defendants were acting within the scope of their employment during the Fusco investigation and therefore were fully indemnified, and attorneys from the law firm Freeman Nooter & Ginsberg were hired by the County to represent all defendants. At a March 2015 hearing, long after the second trial and the jury verdict against Volpe, the district court was alerted that Nassau County might take the position that it was not obligated to indemnify Volpe’s estate. At the hearing, the district court held that “[bjecause the estate and the County’s interests] were aligned at trial, it was appropriate for the County and the estate to put on a joint defense. However, it appears that the interests of the County and the estate have now diverged.” Appellees’ Supp. App’x at 1819. The district court therefore disqualified Freeman Nooter & Ginsberg from representing Volpe’s estate, based on the “unwaivable direct conflict” due to the law firm representing Volpe while at the same time contending that the County did not need to indemnify the estate. Appellees’ Supp. App’x at 1810. Thereafter, plaintiffs, the conflict counsel for the estate, and counsel for the County of Nassau entered into a binding agreement, so ordered by the district court on July 1, 2015, that the County would fully indemnify the estate.
DISCUSSION
I. Ruling on Rule 59(a) Motion for a New Trial
Volpe first challenges the district court’s grant of Restivo and Halstead’s motion for a new trial under Federal Rule of Civil Procedure 59(a) following the defense verdict at the first trial.
In general we review for abuse of discretion a district court’s denial of a motion for a new trial pursuant to Rule 59, but where the basis of the Rule 59 motion is an erroneous jury instruction, we review the jury instructions de novo. A jury instruction is erroneous if it misleads the jury as to the correct legal standard or does not adequately inform the jury on the law. An erroneous jury instruction requires a new trial, unless the error is harmless.
Velez v. City of New York, 730 F.3d 128, 134 (2d Cir. 2013) (internal quotation marks and citations omitted).
Because a malicious prosecution plaintiff must show the absence of probable cause for prosecution, see Torres v. Jones, 26 N.Y.3d 742, 47 N.E.3d 747, 760 (2016), whether the police defendants had *570probable cause to proceed with the prosecution of Restivo and Halstead was of central importance at both the first and second trials in this case. The district court’s decision to grant the Rule 59(a) motion turned precisely on this point. It first held that, because Kogut’s confession was inadmissible at the criminal trial against Res-tivo and Halstead, pursuant to Boyd v. City of New York, 336 F.3d 72 (2d Cir. 2003), Volpe could not use that confession to defend against Restivo and Halstead’s assertion that probable cause for the prosecution was lacking. The district court concluded that it had erred in framing the jury instructions so broadly as to imply that Kogut’s confession, even if admissible at the first trial for other reasons, could, in fact, be used for this purpose. It therefore granted Restivo and Halstead’s Rule 59(a) motion for a new trial.
We conclude that the district court’s reading of our limited holding in Boyd was incorrect. Nevertheless, the court was correct in concluding that, in the context of this case, the instruction on probable cause at the first trial was inadequate, and in a manner implicating the fairness of that proceeding. We therefore affirm the district court’s ruling on an alternative ground. Standard Inv. Chartered, Inc. v. Nat’l. Ass’n of Sec. Dealers, Inc., 560 F.3d 118, 126 (2d Cir. 2009).
As to Boyd, we held only that in the particular circumstances that case presented, a statement that constituted the sole evidence of probable cause and that would be inadmissible in the criminal case (from an ex ante perspective) did not warrant the grant of summary judgment to the civil defendants. This holding turned on an unusual set of facts in which “the only evidence the police had that could reasonably indicate that Boyd knew the [car] was stolen was his statement” to the police. Id, at 77. Moreover, not only was the evidence tying Boyd to the crime limited, Boyd had also produced evidence to establish that his statement was inadmissible under Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966)—evidence which, if credited by the jury, would have shown that police understood at the time that this statement could not be used to prove a criminal case. We concluded that in these unusual circumstances, the district court erred in granting summary judgment to the defendants. In other words, Boyd did not hold that inadmissible evidence cannot be used in evaluating probable cause for a prosecution, but only that where the sole evidence of a defendant’s guilt is a single statement that police would have understood at the time could not be used in a criminal case (a circumstance largely limited to the facts of Boyd itself) such evidence is not alone sufficient to defeat a malicious prosecution claim.19
*571Kogut’s confession, in contrast, was only one of several pieces of evidence that police officers, including Volpe, had before them in deciding whether the totality of the evidence was indicative of Restivo and Halstead’s guilt and that a prosecution against them would succeed. Cf. Walczyk v. Rio, 496 F.3d 139, 156 (2d Cir. 2007) (“[P]robable cause ... is demonstrated ... [by] the totality of circumstances”). Even if ultimately deemed inadmissible, a confession such as Kogut’s, looked at ex ante, appeared to support the determination that Restivo and Halstead were complicit in Fusco’s rape. “Probable cause does not require proof sufficient to warrant a conviction beyond a reasonable doubt but merely information sufficient to support a reasonable belief that an offense has been ... committed by the suspectf ].” Torres, 47 N.E.3d at 760 (internal quotation marks omitted).
Moreover, at the time members of the police, including Volpe, made the decision to refer Restivo and Halstead for prosecution, it was not yet clear whether Kogut’s confession would be admissible at their criminal tiñáis. Kogut ultimately refused to testify against Restivo and Hal-stead, and, as a result, his confession was inadmissible under Bruton v. United States, 391 U.S. 123, 88 S.Ct. 1620, 20 L.Ed.2d 476 (1968), because to admit it would have violated Restivo and Halstead’s Sixth Amendment Confrontation Clause rights. Id. at 135-36, 88 S.Ct. 1620. If, however, Kogut had cooperated against Restivo and Halstead and testified at their trial, there would have been no Confrontation Clause problem. Thus, there is a temporal component to the analysis as well: The fact that evidence ultimately proves inadmissible does not make it irrelevant to the evaluation of probable cause at the time police are presented with it. Put differently, the fact that Kogut’s confession was eventually deemed inadmissible does not undercut probable cause at the time the decision to prosecute Restivo and Hal-stead was made. See Lowth v. Town of Cheektowaga, 82 F.3d 563, 571 (2d Cir. 1996) (holding that where the probable cause determination was made before an intervening fact came to light the analysis of a malicious prosecution claim looks back to the facts as they were known at the time the determination was made).
The district court’s erroneous interpretation of Boyd notwithstanding; the determination that its instructions were inadequate at the first trial was correct. In the context of that trial, Kogut’s confession was not just any piece of evidence. As the district court observed, “the Kogut confession was arguably one of the central pieces of evidence in this case,”'Special App’x at 62, arid, as this Court recognized, due to its “devastating detail,” the confession “posed a clear[] danger of unfair prejudice” to Restivo and Halstead. Kogut, 789 F.3d at 44. The confession played a central role in evidentiary hearings leading up to both trials, and whether it should be admitted, with respect to whom, and for what purpose was sharply contested. Indeed, Volpe explicitly presses this point, asserting that the jury’s judgment was swayed by the district judge’s failure to admit Kogut’s confession at the second trial. See Appellant’s Br. at 60-61 (arguing that admitting the confession “would have resulted in a defense verdict, as it did at the first trial” and that the confession was “vitally important to rebut Restivo and Halstead’s testimonies that they were actually innocent”).
*572Given the specific and unusual context of this civil case, we can discern no abuse of discretion in the district court’s underlying determination that its failure properly to instruct the jury as to how it should consider the Kogut confession deprived the plaintiffs here of a fair trial in the first proceeding. The district court failed to make clear, by direct instruction, that, while the jury could permissibly apply the confession directly against Kogut, it could only take Kogut’s confession against Resti-vo and Halstead into account to the extent that the confession undermined specific elements of Restivo and Halstead’s malicious prosecution claim.20 In the absence of a clear instruction on these issues, the jury was faced with evidence of a confession indicting Restivo and Halstead whose proper use—despite Restivo and Hal-stead’s initial objection, see Rosenfeld v. Basquiat, 78 F.3d 84, 90 (2d Cir. 1996)— was never explained.21 As we have said in the past, a new trial is properly granted where “taken as a whole, the jury instructions gave a misleading impression or inadequate understanding of the law,” BAII Banking Corp. v. UPG, Inc., 985 F.2d 685, 696 (2d Cir. 1993), especially where the “lack of instruction undoubtedly left the jury confused” about a “potentially disposi-tive” issue in the case, Pitasi v. Stratton Corp., 968 F.2d 1558, 1563 (2d Cir. 1992). In this case, and given these unusual circumstances, we cannot conclude that the district court abused its discretion in determining that a new trial was warranted.
II. Evidentiary Rulings at the Second Trial
Next, Volpe challenges a number of evi-dentiary rulings at the second trial, and contends that a third trial is required. In particular, Volpe challenges (1) the district court’s exclusion of several inculpatory statements by Kogut, Restivo and Hal-stead; (2) the district court’s admission of plaintiffs’ expert testimony on PMRB analysis; (3) the district court’s exclusion of Volpe’s expert Kadane who sought to testify about PMRB analysis; and (4) the district court’s admission of testimony from plaintiffs’ police practices expert, Fischer.
In assessing these challenges, it is important to note that the jury found against Volpe and for Restivo and Halstead on the malicious prosecution claim and the deprivation of a fair trial claim, and that each of these claims is independently sufficient to sustain the damages award against Volpe because there was a single injury: the wrongful imprisonment.22
A. Exclusion of Inculpatory Statements by Kogut, Restivo, and Hal-stead
Volpe first argues that the district court denied him a fair trial by refusing to admit Kogut’s confession and certain admissions allegedly made by Restivo and Halstead. To be admissible, evidence must be relevant, under Federal Rules of Evidence 401 and 402, and even relevant *573evidence may be excluded under Rule 408 if “its probative value is substantially outweighed by a danger of one or more of the following: unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.” We apply “abuse-of-discretion review to a trial court’s eviden-tiary rulings.” Lore v. City of Syracuse, 670 F.3d 127, 155 (2d Cir. 2012). “The hallmark of abuse-of-discretion review is deference_” Id. In particular, when assessing claims of relevance and undue prejudice, our review “is highly deferential in recognition of the district court’s superi- or position to assess relevancy and to weigh the probative value of evidence against its potential for unfair prejudice.” United States v. Coppola, 671 F.3d 220, 244 (2d Cir. 2012) (internal quotation marks omitted). We will therefore only hold that the district court abused its discretion in making an evidentiary ruling where “the ruling was arbitrary and irrational.” Id. Further, as noted, “an erroneous evidentiary ruling warrants a new trial only when a substantial right of a party is affected, as when a jury’s judgment would be swayed in a material fashion by the error.” Lore, 670 F.3d at 155 (internal quotation marks omitted).
First, Volpe challenges the fact that eight alleged statements that tend to show that Restivo and Halstead were guilty of the crimes were not admitted at the second trial. However, as Volpe himself admits, he never sought to introduce these pieces of evidence. Volpe claims that he did not offer the alleged statements or call the relevant witnesses at the second trial because the district court precluded all but the evidence admitted at the criminal trial. But Volpe misstates the court’s ruling. The district court did not rule that only evidence actually admitted at the criminal trial could be admitted in this civil case; rather, the district court clearly ruled that only evidence that would have been admissible at the criminal trial, regardless whether it was actually admitted, could be admitted in the civil trial in order to prove probable cause. See Special App’x at 60-61 (agreeing with plaintiffs’ arguments that “Boyd ... assert[s] that probable cause to commence proceedings means the belief that the prosecution would succeed based on admissible evidence,” and holding, therefore, that “Defendants were limited to evidence that was both included in their interrogatory responses and admissible in Halstead and Restivo’s criminal proceedings”). If Volpe failed to offer evidence based on his misinterpretation of the district court’s clear ruling, he has only himself to blame.
Volpe next challenges two evidentiary rulings actually made by the district court: its exclusion of the Kogut confession and of the statement signed by Restivo, which implicated Halstead in the crime. Because the jury found for plaintiffs and against Volpe on the malicious prosecution claim and the fair trial claim, each of which is sufficient independently to sustain the jury award, we consider only whether the “jury’s judgment would be swayed in a material fashion by the error.” Arlio v. Lively, 474 F.3d 46, 51 (2d Cir. 2007).23
First, the district court did not err in excluding the Kogut confession from consideration with respect to the fair trial claim because the confession was not relevant to plaintiffs’ claims that Volpe planted evidence and suppressed Brady material. *574Volpe argues that the confession is relevant to the fair trial claim for two reasons. First, he contends, the confession explains how the victim’s hair came to be found in Restivo’s van, rebutting plaintiffs’ claim that the hair was planted. But this is only the case if the confession was admitted for its truth, and it was not shown to be admissible for this purpose. Second, Volpe claims that the confession established the reasonableness of Volpe’s belief that the French lead was not Brady material. But whether Volpe had an honestly held and even an objectively reasonable belief that plaintiffs were guilty is not relevant to the question of whether evidence was exculpatory. Moreover, Volpe himself admitted in a deposition that he believed that the French lead was Brady material, but he claimed that he turned it over to the district attorney.
Finally, we consider Volpe’s challenge to the exclusion of Restivo’s statements made on March 5 and 6, 1985, following an overnight interrogation. At some point during the eight-hour interrogation, Restivo signed the following statement:
I would like to say that sometime back possibly November, December 1984, I stopped by my friend Dennis Halstead’s apartment. He lives above the store on Atlantic Avenue.
[[Image here]]
When I saw him, I realized that he was also high. We were talking about 10 to 15 minutes and at this point and kind of out of the blue Dennis started to talk strange.
He started talking about a broad. Dennis said he was with a broad, a girl, and that he was either by a cemet[e]ry, in a cemet[e]ry, across from the cemet[e]ry.
He said he tried to fuck her. Then he had to fuck her up. But when he said that, he didn’t tell me how he fucked her up. He then told me that he strangled her and killed her.
Appellees’ Supp. App’x at 270. As with the Kogut confession, this statement might have some relevance to the hair-planting claim if the jury were permitted to consider the statement by Restivo for its truth. That is, if the jury were to consider this statement for the truth of whether Hal-stead committed the crime, the statement might be relevant in determining whether the hairs allegedly found in Restivo’s van were planted. However, the jury was not permitted to consider the statement for the truth of whether Halstead was guilty. The fact that Restivo made this statement was admissible against him, as a statement by a party opponent under Rule 801(d)(2)(A). But there is a second level of hearsay incorporated into this statement— the statements that Restivo attributed to Halstead.24 No exception to the hearsay rule was shown to apply that would permit admitting this evidence for its truth against Restivo, and the jury could therefore not consider the statements attributed to Halstead for this purpose.25 The same applies for Halstead himself.
B. Expert Witness Rulings
Volpe next challenges several rulings by the district court on testimony of expert witnesses, specifically (1) the district court’s admission of plaintiffs’ expert testimony on post-mortem root banding (“PMRB”) analysis, (2) its exclusion of the *575proposed PMRB defense expert, and (3) its admission of testimony from plaintiffs’ police practices expert, Fischer.
1. Admission of Plaintiffs’ Expert Testimony on PMRB Analysis
As noted, the only physical evidence linking plaintiffs to the crime were the Q hairs that were allegedly recovered during a search of Restivo’s blue van on March 26,1986, almost five months after the murder. At trial, plaintiffs’ experts testified that the presence of PMRB, as well as the extent of the PMRB in the hairs, indicated that the Q hairs were hairs taken from Ms. Fusco during her autopsy and then placed in an envelope containing hairs found in Restivo’s van. Volpe raises two challenges with respect to plaintiffs’ experts’ PMRB testimony. First, he argues, because the court found that the timing of PMRB had not been established to a degree of scientific certainty, the court was not permitted to compromise and allow the evidence to come in so long as the experts did not testify to a degree of scientific certainty. Second, Volpe claims that two of the experts—Dr. Houck and Petraco—did not comply with the limitation set by the court.
The same abuse-of-discretion standard of review applies to rulings on “the admissibility of expert testimony.” Lore, 670 F.3d at 165. Again, this is a highly deferential standard, and “a ruling on the admissibility of expert testimony ‘is to be sustained unless manifestly erroneous.’ ” Id. (quoting Salem v. United States Lines Co., 370 U.S. 31, 36, 82 S.Ct. 1119, 8 L.Ed.2d 313 (1962)). “Significantly, the abuse of discretion standard ‘applies as much to the trial court’s decisions about how to determine reliability as to its ultimate conclusion.’” Amorgianos v. Nat’l R.R. Passenger Corp., 303 F.3d 266, 265 (2d Cir. 2002) (quoting Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999)). “Thus, in analyzing the admissibility of expert evidence, the district court has broad discretion in determining what method is appropriate for evaluating reliability under the circumstances of each case.” Id. “Further, an erroneous evidentiary ruling warrants a new trial only when a substantial right of a party is affected, as when a jury’s judgment would be swayed in a material fashion by the error.” Lore, 670 F.3d at 155 (internal quotation marks omitted).
Under the Federal Rules of Evidence, expert witnesses may testify under the following conditions:
A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if:
(a) the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product of reliable principles and methods; and
(d) the expert has reliably applied the principles and methods to the facts of the case.
Fed. R. Evid. 702. In Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786,125 L.Ed.2d 469 (1993), the Supreme Court ruled that Rule 702 assigns to district courts the gatekeeper function—“ensuring] that any and all scientific testimony or evidence admitted is not only relevant, but reliable.” Id. at 589, 113 S.Ct. 2786. Under Daubert, factors relevant to determining reliability include “the theory’s testability, the extent to which it ‘has been subjected to peer review and publication,’ the extent to which a technique is subject to ‘standards control*576ling the technique’s operation,’ the ‘known or potential rate of error,’ and the ‘degree of acceptance’ within the ‘relevant scientific community.’ ” United States v. Romano, 794 F.3d 317, 330 (2d Cir. 2015) (quoting Daubert, 509 U.S. at 593-94, 113 S.Ct. 2786). But the inquiry is a “flexible one,” Daubert, 509 U.S. at 594, 113 S.Ct. 2786, and the “factors [Daubert] mentions do not constitute a ‘definitive checklist or test.’ ” Kumho Tire Co., 526 U.S. at 150,119 S.Ct. 1167 (quoting Daubert, 509 U.S. at 593, 113 S.Ct. 2786).
In Kumho Tire Co., the Supreme Court ruled “that Daubert’s, general holding— setting forth the trial judge’s general ‘ga-tekeeping’ obligation—applies not only to testimony based on ‘scientific’ knowledge, but also to testimony based on ‘technical’ and ‘other specialized’ knowledge.” 526 U.S. at 141, 119 S.Ct. 1167 (quoting Fed. R. Evid. 702(a)). When determining the admissibility of non-scientific expert evidence, “a trial court may consider one or more of the more specific factors that Daubert mentioned when doing so will help determine that testimony’s reliability.” Id. “But, as the Court stated in Dau-bert, the test of reliability is ‘flexible,’ and Daubert’s list of specific factors neither necessarily nor exclusively applies to all experts or in every case.” Id. “Rather, the law grants a district court the same broad latitude when it decides how to determine reliability as it enjoys in respect to its ultimate reliability determination.” Id. at 142,119 S.Ct. 1167.
Volpe claims that, because the district court concluded that certain aspects of PMRB had not been established to a degree of scientific certainty, the district court should not have permitted plaintiffs’ experts to testify about PMRB at all. We disagree. Rule 702 “makes no relevant distinction between ‘scientific’ knowledge and ‘technical’ or ‘other specialized’ knowledge,” and “makes clear that any such knowledge might become the subject of expert testimony.” Kumho Tire Co., 526 U.S. at 147, 119 S.Ct. 1167. There is no basis in the rule or in our case law to suggest that a scientist whose testimony could not pass muster under Daubert as “scientific knowledge” could not testify to their technical or other specialized knowledge, so long as that testimony was reliable under Rule 702 and Kumho. “Experts of all kinds tie observations to conclusions through the use of what Judge Learned Hand called ‘general truths derived from ... specialized experience.’ ” Id. at 148, 119 S.Ct. 1167 (quoting Learned Hand, Historical and Practical Considerations Regarding Expert Testimony, 15 Harv. L. Rev. 40, 54 (1901)). And, just as non-scientist experts can testify about their opinions, so too can scientists, when their opinions are based on reliable technical or specialized knowledge, though not scientific fact. As explained in In re Ephedra Products Liability Litigation, 393 F.Supp.2d 181 (S.D.N.Y. 2005), “[t]he Court in Kumho Tire rejected the view that Rule 702 sets a lower standard for witnesses with ‘technical or other specialized knowledge’ than for scientists. But its reasoning cuts both ways: if art appraisers, handwriting experts or economists may express professional opinions that fall short of definitive proof, so may scientists.” Id. at 188; see also id. at 190 (reasoning that Daubert was “never intended to keep from the jury the kind of evidence scientists regularly rely on in forming opinions of causality simply because such evidence is not definitive,” and holding that “Rule 702, a rule of threshold admissibility, should not be transformed into a rule for imposing a more exacting standard of causality than more-probable-than-not simply because scientific issues are involved”).
*577Thus, if a scientist wants to testify to an opinion or conclusion that has not been established to a degree of scientific certainty under Daubert, the court may admit the testimony as non-scientific expert testimony under Rule 702 and Kumho Tire. That is, the court must still assess whether the expert employs “the same level of intellectual rigor that characterizes the practice of an expert in the relevant field,” Kumho Tire Co., 526 U.S. at 152, 119 S.Ct. 1167, and may consider the Dau-bert factors in making this determination or other relevant factors, id. at 141-42,119 S.Ct. 1167. In particular, we have held that a “trial judge should exclude expert testimony if it is speculative or conjectural or based on assumptions that are so unrealistic and contradictory as to suggest bad faith or to be in essence an apples and oranges comparison.” Zerega Ave. Realty Carp. v. Hornbeck Offshore Transp., LLC, 571 F.3d 206, 214 (2d Cir. 2009) (internal quotation marks omitted). By contrast, “other contentions that the assumptions are unfounded go to the weight, not the admissibility, of the testimony.” Boucher v. U.S. Suzuki Motor Carp., 73 F.3d 18, 21 (2d Cir. 1996) (internal quotation marks omitted). Additionally, the district court may consider the gap between the data and the conclusion drawn by the expert from that data, and exclude opinion evidence where the court “concludefs] that there is simply too great an analytical gap between the data and the opinion proffered.” Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997). Frequently, though, “gaps or inconsistencies in the reasoning leading to [the expert’s] opinion ... go to the weight of the evidence, not to its admissibility.” Campbell ex rel. Campbell v. Metro. Prop. & Cas. Ins. Co., 239 F.3d 179, 186 (2d Cir. 2001).
With that background, we turn to the question whether the district court abused its discretion in admitting plaintiffs’ experts’ testimony, while prohibiting those experts from testifying to a reasonable degree of scientific certainty. The district court, having found that the timing of PMRB was not scientifically established, applied the Daubert factors and found that they overall supported the admissibility of plaintiffs’ experts’ testimony. The district court found that there was evidence that PMRB can be distinguished from other types of environmentally caused banding within an acceptable rate of error—99.5% accuracy when one person worked alone, or 100% when two examiners double checked each other’s work. See Kogut, 894 F.Supp.2d at 235. This was reinforced by Domzalski’s thesis, which stated that although environmental factors can cause changes to scalp hair roots, environmental banding is visually different, occurring at a different place on the hair than PMRB. Id. Additionally, the experts’ opinions were consistent with the academic literature on the topic, and PMRB, as plaintiffs’ experts described it, is a generally accepted phenomenon within the forensic science community. Id. In sum, the court concluded that “although these facts do not add up to scientific proof, they supply a reasonable basis for forensic experts to conclude that PMRB is an artifact of decomposition and that, consistent with the speed at which other effects of decomposition appear on a corpse, it does not appear immediately after death.” Id. at 244. The district court, before the second trial, also considered the Koch study, which, as noted, involved a test of roughly 24,000 hair roots from 23 human cadavers permitted to decompose in varying environmental conditions. In cadavers outside on the ground, PMRB generally did not develop until 6-10 days after death. The earliest development of PMRB in the study, which occurred in a cadaver placed in the trunk of a car in August, *578occurred after 4 days. Based on this evidence, we find no abuse of the district court’s discretion in holding that this evidence was sufficiently reliable to be admitted, and the analytical gaps in the reasoning of Plaintiffs’ experts were not too great.
Finally, we find no reversible error with respect to Volpe’s claims that Dr. Houck and Petraco violated the court’s instruction not to state that their testimony was to a degree of scientific certainty. To the extent that Dr. Houck or Petraco violated this instruction, any error was harmless. With respect to Dr. Houck, the district court gave an instruction that he was not testifying to a reasonable degree of scientific certainty, and Petraco testified on cross-examination that he was not testifying to a reasonable degree of scientific certainty. Finally, to the extent that Volpe’s argument on appeal turns on the fact that these experts testified to their opinion on an ultimate issue, that is, whether the hairs were planted, Federal Rule of Evidence 704(a) instructs that “[a]n opinion is not objectionable just because it embraces an ultimate issue.”
2. Exclusion of Defense Expert on PMRB
Volpe next claims that the district court incorrectly precluded him from introducing testimony from his proposed expert, Dr. Joseph B. Kadane. Dr. Kadane, a statistician, sought to testify that (1) post-mortem root banding could not be distinguished from various types of ante-mortem root banding and that (2) the academic research that had been done to that point was insufficient to establish how long it took after death for PMRB to appear.
With respect to the first point, Dr. Kadane referred to the Domzalski study, which found that various environmental conditions could cause root banding that “could be confused” with PMRB, although environmental banding appeared nearer to the root of the hair than PMRB. Kogut, 894 F.Supp.2d at 287. Based on this study, Kadane sought to testify that it cannot be determined whether the Q hairs were removed post-mortem or ante-mortem. Id. at 237-38. The district court correctly excluded this line of testimony. Reasoning that “[a]lthough there is no dispute that Ka-dane is an accomplished statistician, it is equally beyond debate that he lacks more than a passing familiarity with hair microscopy and forensic science,” the court found him not equipped to testify about this matter. Id. at 244. The question whether experts can distinguish between ante- and post-mortem root banding is a topic clearly outside Dr. Kadane’s area of expertise. See Kmnho Tire Co., 526 U.S. at 149, 119 S.Ct. 1167 (requiring the district court to determine “whether the testimony has a reliable basis in the knowledge and experience of the relevant discipline” (alteration and internal quotation marks omitted)). Further, this opinion was not based on “sufficient facts or data,” Fed. R, Evid. 702(b), as it was based only on an article that stated that the existence of post-mortem root banding is well-documented in the literature and found that ante- and postmortem root banding were visibly different, indicating that trained experts could distinguish the different types of banding. Motion to Strike, Kogut v. Cty. Of Nassau, No. 06-CV-6695 (E.D.N.Y. Apr. 30, 2012), EOF No. 205, Ex. G„ at 16 (discussing PMRB); id. at 49 (“[S]ome of the patterns that arose from environmental exposure appeared similar to those encountered in postmortem samples, which has very important forensic implications for interpretation of hair evidence. While these patterns, arising from environmental exposure, could be confused for postmortem root bands, it is important to point out *579that the root bands that appeared in this study were very proximal to the anagen root end. This is not an accepted criterion for postmortem root banding.”); id. at 54-55 (“In this study, some hairs produced patterns that could have been construed as postmortem root changes. One should bear in mind that banding patterns in postmortem hair roots appear more distal to the anagen root end as opposed to the banding patterns in the environmentally exposed hairs, which appeared proximal to the ana-gen root end.”).26
With respect to the second proposed area of testimony, Dr. Kadane sought to testify about experimental flaws in the research relied on by plaintiffs’ experts. In particular, Dr. Kadane listed factors that were not accounted for in the studies relied on by plaintiffs’ experts. The district court again held that Kadane’s “expertise is simply not useful in attempting to refute Plaintiffs’ Experts’ opinions about PMRB,” reasoning that Dr. Kadane “was free to conduct his own statistical analysis of PMRB but did not do so.” Special App’x at 40. Even assuming, ar-guendo, that the district court abused its discretion in excluding Dr. Kadane from testifying about experimental flaws and the value of the data on which plaintiffs’ experts relied, any error was harmless. Dr. Kadane sought only to critique the testimony of plaintiffs’ expert witnesses, and Volpe was able to effectively cross-examine plaintiffs’ experts based on the statistical issues raised by Dr. Kadane. See Beck Chevrolet Co. v. Gen. Motors LLC, 787 F.3d 663, 680 (2d Cir.) (holding that any error was harmless where an expert report was excluded but counsel cross-examined the opposing expert about the topics contained in the report), certified question on an unrelated issue accepted, 25 N.Y.3d 1057, 33 N.E.3d 496 (2015), and certified question answered, 27 N.Y.3d 379, 53 N.E.3d 706 (2016).
3. Police Practices Expert Testimony
Finally, Volpe claims that the district court abused its discretion in admitting certain testimony from plaintiffs’ police practices expert, Russell Fischer. First, Volpe claims that Fischer misinformed the jury about the relevant legal Brady standard that was in place during the time of the investigation.27 Second, Volpe argues that Fischer was improperly permitted to define materiality and opine on the materiality of the French lead. Third, Volpe argues that Fischer should not have been permitted to testify about the words “possible human blood” and “hair consistent with” because those terms have no special law enforcement meaning. Finally, Volpe argues that Fischer should not have been permitted to testify about the effects of lengthy police interrogation on reliability because this was a veiled attack on the credibility of Cockerel, which was the province of the jury.
First, the district court did not abuse its discretion in admitting Fischer’s testimony with respect to Brady and the materiality of the French lead. Contrary to Volpe’s assertions, Fischer did not testify to the legal Brady standard or define materiality. Rather, he testified to minimally accepted police practices, “a baseline type *580of behavior or protocols that law enforcement officers follow.” Trial Tr. at 1776. He testified only about what minimally accepted police practices required and about his opinion that Volpe’s conduct departed from accepted police practices. “When an expert offers an opinion relevant to applying a legal standard such as probable cause [or Brady], the expert’s role is limited to describing sound professional standards and identifying departures from them,” Jimenez v. City of Chicago, 732 F.3d 710, 721 (7th Cir. 2013) (internal quotation marks omitted), which is precisely what Fischer did in this case. This expert testimony on applicable professional standards is relevant because it “can give a jury a baseline to help evaluate whether a defendant’s deviations from those standards were merely negligent or were so severe or persistent as to support an inference of intentional or reckless conduct that violated a plaintiffs constitutional rights.” Id. at 721-22. It is true that Fischer’s “opinions had direct implications for applying legal standards,” but that is exactly “why his testimony was relevant.” Id. at 721. The court correctly limited Fischer’s testimony to professional practices and departures in this case, and Fischer did not stray from these bounds.
Volpe’s claim that Fischer should not have been permitted to define “possible human blood” and “hair consistent with” similarly lacks merit. Experts are permitted to testify about “custom and usage” of terms in particular industries. See SR Int'l Bus. Ins. Co. v. World Trade Ctr. Props., LLC, 467 F.3d 107, 134 (2d Cir. 2006). If, as Volpe claims, those terms have no special law enforcement meaning, then Volpe should have introduced testimony to that effect, expert or otherwise. That does not render Fischer’s testimony inadmissible.
Finally, Fischer testified about potential risks inherent in long interrogations, and testified that if a witness’s story changed over time, under minimally accepted police practices, an officer should record the changes and corroborate the statements. Volpe challenges this testimony as going to witness credibility. But, testimony about appropriate responses to inconsistencies in a witness’s story is permissible where that testimony is introduced to inform the jury “what a reasonable police investigator ... [does] when presented with ... conflicting and/or inculpatory statements during [a] murder investigation.” Jimenez, 732 F.3d at 723. Accordingly, we find no abuse of discretion in the district court’s decision to admit Fischer’s testimony.
III. Conflict of Interest
Next, Volpe argues that a conflict of interest deprived him of a fair trial. In particular he claims that because counsel represented both Volpe and Fraas, counsel did not argue that, if anyone planted the hair evidence, it was Fraas. Volpe also claims that there was a second conflict because counsel also represented the county, which had agreed to indemnify both Volpe and Fraas, and thus preferred the argument that no evidence was planted, rather than the argument that Fraas and not Volpe planted evidence. Volpe argues that this prejudiced him at trial and that this prejudice is not remedied by Nassau County indemnifying him because the county could again deny indemnification and because Volpe’s family has an interest in restoring his reputation.
Volpe relies primarily on Dunton v. Suffolk County, 729 F.2d 903 (2d Cir. 1984), amended on other grounds, 748 F.2d 69 (2d Cir. 1984), for this argument. In Dun-ton, a Suffolk County police officer defendant, Officer Pfeiffer, allegedly assaulted the plaintiff, Dunton, after the plaintiff allegedly made improper advances toward Officer Pfeiffer’s wife. See id. at 905. Dun-*581ton brought suit against Officer Pfeiffer, his wife, and Suffolk County. Under local law, Suffolk County provides for representation of employees sued under Section 1983. The county attorney’s answer to Dunton’s complaint raised the affirmative defense that Officer Pfeiffer was “acting in good faith pursuant to his official duties and responsibilities,” but at trial, the county attorney’s theory of the case was that Officer Pfeiffer was acting as an “irate husband rather than a police officer.” Id. at 906 (internal quotation marks omitted). The county attorney’s position at trial was therefore against Officer Pfeiffer’s interest. “Specifically, Officer Pfeiffer claim[ed] that it was in his interest to assert his immunity from section 1983 liability based on good faith actions within the scope of his employment.” Id. at 907. By contrast, the attorney undermined the good faith immunity defense by repeatedly stating that Pfeiffer acted not as a police officer but as an irate husband, a defense that would help the county escape liability. Thus, at trial a serious conflict actually surfaced “when the County Attorney stated that Pfeiffer was not acting under color of state law but rather as an ‘irate husband.’” Id. at 907. Our Court reasoned,
As soon as the County Attorney began to undermine Officer Pfeiffer’s good faith immunity defense by stating that Pfeiffer acted as an “irate husband” and not as a police officer, he was not only failing to act as a conscientious advocate for Pfeiffer, but was acting against Pfeif-fer’s interest. The seriousness of this conflict made disqualification appropriate.
Id. at 908. Further, this conflict was prejudicial because the jury did not consider Officer Pfeiffer’s good faith defense and the county indemnified Officer Pfeiffer only for compensatory damages, not punitive damages, and “[i]f the jury found that Pfeiffer was acting in good faith as a police officer, it might not have awarded punitive damages.” Id. at 909.
We find no actual and serious conflict requiring a new trial here. First, the “particular conflict cited in Dunton as inherent in Section 1983 actions against municipalities, namely that the municipality can escape liability by arguing that its employees were not acting within the scope of official employment while the employee can escape liability by arguing the opposite, is simply not present here.” Patterson v. Balsamico, 440 F.3d 104, 115 (2d Cir. 2006). Unlike in Dunton, in this case the County was not a party at trial facing a Monell claim, and thus would not benefit from the argument that Volpe was acting outside the scope of his employment. Further, prior to trial, Nassau County had taken the position that it was fully indemnifying Volpe, and there was no indication that it might not until after the trial ended. At trial, therefore, the County and Volpe’s interests aligned. Unlike in Dunton where counsel argued a theory that benefited one defendant and harmed another, counsel’s defense theory—that there was no suppression of exculpatory evidence or planting of evidence—was a unified theory of defense benefitting Volpe, Fraas, and the indemnitor, the County. Finally, we find unpersuasive Volpe’s argument that an un-conflicted attorney would have pointed the finger at Fraas and that this necessitates a new trial. Indeed, defense counsel argued that, if someone had planted the hair, it could not have been Volpe, as he lacked the necessary access to do so.
IV. Setoff
We now turn to issues related to the jury’s damages award. Restivo and Hal-stead each received $2.2 million in compensation in a settlement with New York State pursuant to New York Court of Claims Act Section 8-b, which provides for *582compensation by the State upon a showing of wrongful conviction without proof of wrongdoing by any government official. Volpe argues that the jury’s damages award should be set off by the amount received in the State settlement.
As to Restivo and Halstead’s recovery on Section 1983 claims, we first must decide what law to apply in deciding whether a setoff is available. Under Section 1988 of Title 42, jurisdiction over Section 1983 claims
shall be exereised and enforced in conformity with the laws of the United States, so far as such laws are suitable to carry the same into effect; but in all cases where they are not adapted to the object, or are deficient in the provisions necessary to furnish suitable remedies and punish offenses against law, the common law, as modified and changed by the constitution and statutes of the State wherein the court having jurisdiction of such civil or criminal cause is held, so far as the same is not inconsistent with the Constitution and laws of the United States, shall be extended to and govern the said courts in the trial and disposition of the cause, and, if it is of a criminal nature, in the infliction of punishment on the party found guilty.
42 U.S.C. § 1988(a); see also Davis v. Rodriguez, 364 F.3d 424, 433 n.7 (2d Cir. 2004); Valley Disposal Inc. v. Cent. Vt. Solid Waste Mgmt. Dist., 113 F.3d 357, 362 (2d Cir. 1997) (“Section 1988 ... provides that where there are gaps in federal law with respect to the availability of suitable remedies for civil rights violations, the courts should look to state law insofar as it is not inconsistent with federal law.”).
This statutory scheme establishes a three part test for finding substantive law. First, if federal law is neither deficient nor inapplicable, it will apply. Second, if federal law does not apply, state law does apply, unless, third, state law would be inconsistent with the Constitution and laws of the United States.
Dobson v. Camden, 705 F.2d 759, 762 (5th Cir. 1983), modified on other grounds on reh’g, 725 F.2d 1003 (5th Cir. 1984); see also Robertson v. Wegmann, 436 U.S. 584, 588, 98 S.Ct. 1991, 56 L.Ed.2d 554 (1978).
Turning to the first question, we find that federal law is deficient because federal statutory law is silent on the question of the effect of a settlement on the damage award against a non-settling defendant in a Section 1983 case, and because federal common law is not clear on this point. Compare Miller v. Apartments & Homes of N.J., Inc., 646 F.2d 101, 110 (3d Cir. 1981) (holding in a Fair Housing Act and Section 1982 case, that, under the federal common law, “where one or more defendants have settled with a plaintiff, the damages recoverable by that plaintiff shall be reduced by the amount of the settlement received”), with Dobson, 705 F.2d at 768 (holding, in a Section 1983 case, that a settlement by one tortfeasor has no effect on the liability of another, with “each non-settling tortfeasor” being responsible for a “proportional share of the damages”); see also Dobson, 705 F.2d at 762 (finding federal law deficient on this issue); Johnson v. Rogers, 621 F.2d 300, 304 n.6 (8th Cir. 1980) (“In the instant case, as federal law does not answer the question of the effect of a plaintiffs settlement with one defendant on an award of damages against a non-settling defendant, state law should serve as the federal rule of decision unless inconsistent with the policies underlying a [Section] 1983 cause of action.”); Mason v. City of New York, 949 F.Supp. 1068, 1077-78 (S.D.N.Y. 1996) (finding federal law deficient due to inconsistent pronouncements); Hoffman v. McNamara, 688 F.Supp. 830, 833 (D. Conn. 1988) (same).
*583The dissent, in contrast, concludes that federal law is not deficient because some federal caselaw suggests that setoff is appropriate in the context of certain Section 1983 actions. We disagree with the dissent, and continue to find federal law is deficient and inapplicable, for three basic reasons: (1) federal caselaw makes clear that the State of New York is not a proper defendant in a Section 1983 action and, therefore, whether and for what amount the State of New York has settled claims with the Section 1983 plaintiffs is irrelevant with respect to other defendants’ liability for plaintiffs’ Section 1983 claims; (2) no federal caselaw suggests that setoff is appropriate where a settling party’s liability is never considered at trial by a jury; and (3) federal caselaw prohibiting double recovery is inapplicable where no total amount of loss can be ascertained.
First, it is axiomatic that a state is not a proper defendant in an action brought pursuant to Section 1983. Will v. Mich. Dep’t of State Police, 491 U.S. 58, 71, 109 S.Ct. 2304, 105 L.Ed.2d 45 (1989) (holding that a “State” is not a “person under [Section 1983]”). At trial, the jury awarded Restivo and Halstead $18 million each from Volpe pursuant to Section 1983. Supp. App’x at 1528. None of that amount could even theoretically have been attributable to the State of New York, whether or not the State of New York had previously settled, because the State of New York is not a proper party under Section 1983. The dissent’s reliance on cases that address setoff in situations where the settling party could have been held liable under the same theory or theories of liability as those considered at trial are therefore wholly inapposite to the situation here.
Second, none of the cases to which the dissent points address the issue of settlement before trial of a party in a Section 1983 action where the liability of the settling party is then never considered by the jury at trial. This distinction is crucial— the settling defendant, here the State of New York, was never included in the charge to the jury as a liable defendant. See Supp. App’x at 1332-33. This is in no small part because the State of New York is not a proper defendant in a Section 1983 action. Assuming, arguendo, that the State was a proper party vis á vis the claims on which the jury rendered its damages award, however, federal law would still be deficient here. This is because New York Court of Claims Act Section 8-b, under which the State of New York settled, has no fault requirement. Plaintiffs’ Section 8-b action did not turn on a wrongful or malicious conviction (as would be required for a finding of liability under Section 1983), but simply on an incorrect one. The State, therefore, never admitted liability and so could not have been found liable by the jury unless it had been included as a defendant, which it was not.
The absence of the State of New York from the instant action at any phase of the case, as well as the inability of the State to be held liable under plaintiffs’ asserted causes of action, also distinguishes the instant case from McDermott, Inc. v. Am-Clyde, 511 U.S. 202, 205-06, 114 S.Ct. 1461, 128 L.Ed.2d 148 (1994) on which the dissent relies. In McDermott, the jury considered the fault of and apportioned damages between all parties to the action, including the petitioner and a group of defendants with which the petitioner had already settled, who had all been sued under the same causes of action. Id. Here, in contrast, the jury never considered what fault, if any, it might have attributed to the State. This is in part because the State could not have been held liable under the causes of action presented to the jury, and in part because the liability-phase jury *584verdict sheet in the instant trial shows that only the liability of Volpe and Fraas was considered by the jury, not the liability of either the State of New York or of Nassau County (which would have been a proper defendant). See Supp. App’x at 1332-33. Further, once the jury had found at the liability phase that only Volpe was hable to Restivo and Halstead, then, at the damages phase, all that the jury considered was the amount of damages that Volpe, as an individual, owed to Restivo and Halstead as a result of their eighteen-year incarceration. See Supp. App’x at 1528. Thus, the dissent’s background rule of no double recovery for Section 1983 and state-law claims arising out of the same operative facts, or even the rule that defendants’ share of damages may be reduced by those of settling defendants, has no application to the instant facts where the settling party was not a proper defendant under Section 1983, never admitted liability, and was wholly absent from the trial. On this issue, federal law is deficient.
Third and lastly, the dissent also posits that federal common law barring double recovery in order to ensure that the victim does not receive more than he or she lost is applicable here. See United States v. Nucci, 364 F.3d 419, 423 (2d Cir. 2004). We disagree. Nucci, emblematic of this line of cases, based its holding in part on the observation that the “effect of joint liability in a tort context is to excuse one defendant from paying any portion of the judgment if the plaintiff collects the full amount from the other.” Id. (internal quotation marks omitted). This holding, however, is based on the premise that there is a known “full amount” of loss to be reimbursed through restitution. There is no known full amount of loss here. The jury’s determination that Volpe owed Restivo and Halstead each $18 million is not the same as finding that the total amount of damages that Restivo and Halstead were owed for the eighteen years each spent in prison was $18 million. Instead, the jury could have determined that that was the amount which Volpe alone owed each man under Section 1983. Since there is no way to know the amount of Restivo and Hal-stead’s “loss,” it is inappropriate to apply rules ensuring an individual does not recover in tort or after a criminal case more than he “lost” where one defendant is neither included in the trial nor could be held liable under the same cause of action as the one on which the jury based its damages award.
Accordingly, we hold that federal law is deficient with respect to a scenario in which an absent settling party settles a claim distinct from the claim which is sent to the jury, the jury was never asked to consider or determine the liability of the absent settling party, and the full amount of loss is unknowable.
Finding federal law deficient on this point, we turn to New York state law and the question whether New York law is inconsistent with federal policy underlying Section 1983. See Robertson, 436 U.S. at 590, 98 S.Ct. 1991 (“In resolving questions of inconsistency between state and federal law raised under [Section] 1988, courts must look not only at particular federal statutes and constitutional provisions, but also at ‘the policies expressed in [them].’ ” (alteration in original) (quoting Sullivan v. Little Hunting Park, Inc., 396 U.S. 229, 240, 90 S.Ct. 400, 24 L.Ed.2d 386 (1969))). New York General Obligations Law Section 15-108 provides:
When a release or a covenant not to sue or not to enforce a judgment is given to one of two or more persons liable or claimed to be liable in tort for the same injury, or the same wrongful death, it does not discharge any of the other tort-feasors from liability for the injury or *585wrongful death unless its terms expressly so provide, but it reduces the claim of the releasor against the other tortfea-sors to the extent of any amount stipulated by the release or the covenant, or in the amount of the consideration paid for it, or in the amount of the released tortfeasor’s equitable share of the damages under article fourteen of the civil practice law and rules, whichever is the greatest.
N.Y. Gen. Oblig. Law § 15-108(a). This law, which provides for either a pro tanto (dollar for dollar) setoff or a setoff of the settling tortfeasor’s equitable share of damages, whichever is greater, is inconsistent with federal policy underlying Section 1983.
“The legislative history of [Section] 1983 ... demonstrates that it was intended to create a species of tort liability in favor of persons who are deprived of rights, privileges, or immunities secured to them by the Constitution.” Carey v. Piphus, 435 U.S. 247, 253, 98 S.Ct. 1042, 55 L.Ed.2d 252 (1978) (alteration and internal quotation marks omitted); see also Hardin v. Straub, 490 U.S. 536, 539 n.5, 109 S.Ct. 1998, 104 L.Ed.2d 582 (1989). “Thus, [Section] 1983 addresses constitutional principles that reach profoundly to the core of our notion of justice and underpin our legal system’s groundings on the rule of law.” Banks ex rel. Banks v. Yokemick, 177 F.Supp.2d 239, 260 (S.D.N.Y. 2001). Accordingly, the policies underlying Section 1983 include not only compensation of injured parties, but also “deterrence of future abuses of power by persons acting under color of state law.” City of Newport v. Fact Concerts, Inc., 453 U.S. 247, 268, 101 S.Ct. 2748, 69 L.Ed.2d 616 (1981); see also Owen v. City of Independence, 445 U.S. 622, 651, 100 S.Ct. 1398, 63 L.Ed.2d 673 (1980) (“[Section] 1983 was intended not only to provide compensation to the victims of past abuses, but to serve as a deterrent against future constitutional deprivations, as well.”); Robertson, 436 U.S. at 590-91, 98 S.Ct. 1991 (“The policies underlying [Section] 1983 include compensation of persons injured by deprivation of federal rights and prevention of abuses of power by those acting under color of state law.”); Ciraolo v. City of New York, 216 F.3d 236, 243 (2d Cir. 2000) (Calabresi, J., concurring) (noting that Section 1983 was “designed to deter constitutional torts by making the tortfeasors liable in damages to their victims”); Weaver v. Brenner, 40 F.3d 527, 532 (2d Cir. 1994) (“The goal of [Section] 1983 is to deter public officials from violating citizens’ federal rights and to compensate victims of such official wrongdoing.”); Dobson, 705 F.2d at 765 (“Section 1983 ... embodies two policies: compensation of plaintiffs and deterrence of wrongdoers.”). The deterrence goal of Section 1983 is best served by “internalizing to the violator the costs of violating federal rights.” Dobson, 705 F.2d at 765; see also Ciraolo, 216 F.3d at 243 (Calabre-si, J., concurring) (discussing the importance of a wrongdoer internalizing the costs of his conduct to effectively deter future conduct). Moreover, “[a] rule that removes the burden of damages from the wrongdoer certainly conflicts with the policy of deterrence.” Dobson, 704 F.2d at 766.
By contrast, New York General Obligation Law Section 15-108(a) has “two purposes: first, to encourage settlements, and, second, to ensure that nonsettling tortfeasors are not required to bear more than their equitable share of liability.” Apple v. Jewish Hosp. & Med. Ctr., 829 F.2d 326, 331 (2d Cir. 1987) (citation omitted). “At its most basic level, the driving force embodied in G.O.L. [Section] 15-108 is that of ensuring the compensation of victims while preventing their unjust enrichment.” Banks, 177 F.Supp.2d at 260. “[T]he statute, in the interest of minimiz*586ing the potential for over-compensation, fosters equitable, financial and judicial economy policies that appear to convey more solicitude towards fairness to the nonsettling tortfeasor than to the injured party.” Id.
The New York rule is inconsistent with the deterrent goal of Section 1983, as it allows nonsettling tortfeasors to bear less than the full cost of the harm they inflicted if settling tortfeasors settle for more than their proportional share of liability. “A rule that removes the burden of damages from the wrongdoer certainly conflicts with the policy of deterrence,” and “[w]e must always be conscious that one of the functions of section 1983, if not the primary function, is therapeutic, seeking to eradicate the disease of violations of federal rights under color of state law.” Dobson, 705 F,2d at 766. The district court was therefore correct in not applying the New York rule in determining whether Volpe was entitled to a setoff. See id. (holding that a pro tanto setoff provision under state law “would contravene the deterrence policy of section 1983, and thus is impermissible under section 1988”); Banks, 177 F.Supp.2d at 261 (holding that G.O.L. Section 15-108 is inconsistent with federal policy because it “allow[s] the nonsettling wrongdoer full credit for the amount others paid, thereby entirely relieving him from any obligation to compensate the extent of the injuries he caused” and “effectuate[s] a windfall to the defendant” “in the name of avoiding unjust enrichment for the plaintiff”).
Having concluded that state law is inconsistent with federal policy, we now assess whether the district court was correct in applying a policy of proportional reduction based on the nonsettling party’s proportionate share of liability. “Elemental notions of fairness dictate that one who causes a loss should bear the loss.” Owen v. City of Independence, 445 U.S. 622, 654, 100 S.Ct. 1398, 63 L.Ed.2d 673 (1980); see also Dobson, 705 F.2d at 769 (“The fairness of a tortfeasor paying an injured party rests not solely on the fact that the injured party deserves compensation, but also on the fact that the tortfeasor deserves to pay.”). It follows therefore, that a nonsettling party is entitled to a setoff only of a settling tortfeasor’s proportionate fault. See McDermott, 511 U.S. at 204,114 S.Ct. 1461 (holding, under federal common law in admiralty suit, that the liability of the nonsettling defendants should be calculated with reference to the jury’s allocation of proportionate responsibility, not a dollar for dollar setoff); Dobson, 705 F.2d at 767; Banks, 177 F.Supp.2d at 263-64.
This then leads to the question whether Volpe is entitled to a proportionate reduction. “Under the proportionate share approach, the allocation will take place at trial.” McDermott, 511 U.S. at 217, 114 S.Ct. 1461. That is, the jury is to decide the relative faults of the tortfeasors. The State has never been found to have been at fault, as it was sued under and settled pursuant to a state law that provided for compensation without any showing of fault by the State. See N.Y. Ct. of Claims Act § 8-b (providing for compensation by the State upon a showing that the plaintiff was incarcerated for a state crime, was innocent, and did not, through his own conduct, cause the conviction). Further, Volpe did not seek to introduce evidence tending to show that the State bore a share of the responsibility for plaintiffs’ incarceration, nor did he request an instruction that the jury should determine the State’s proportionate share of liability or consider the effect of a prior settlement when determining damages.28 For these *587reasons, there is no basis for applying a setoff to the damages awarded on Restivo and Halstead’s Section 1983 claims. Further, even if a setoff were available as to the state law claims, Restivo and Halstead are entitled to the full jury award, because “where only a single award of damages, not segregated into separate components, is made, the ... rule ... is that the successful plaintiff be paid under the theory of liability that provides the most complete recovery.” Magee v. U.S. Lines, Inc., 976 F.2d 821, 822 (2d Cir. 1992). The district court therefore correctly denied Volpe’s request for a setoff.
V. Remittitur Motion
Volpe next argues that the jury award was excessive. Following the trial, the district court denied Volpe’s motion for remittitur, reasoning that the award was reasonable under either the federal standard or the state standard. The federal standard provides that “we may set aside a jury’s award only if it is so high as to shock the judicial conscience and constitute a denial of justice.” Zeno v. Pine Plains Cent. Sch. Dist., 702 F.3d 655, 671 (2d Cir. 2012) (internal quotation marks omitted). By contrast, New York state law provides that an appellate court “shall determine that an award is excessive or inadequate if it deviates materially from what would be reasonable compensation.” N.Y. C.P.L.R. § 5501(c). Although a federal district court sitting in diversity considering a jury award for a claim under New York law must apply the state law standard of review, see Gasperini v. Ctr. for Humanities, Inc., 518 U.S. 415, 437-38, 116 S.Ct. 2211, 135 L.Ed.2d 659 (1996), district courts within this circuit have differed in their approach for reviewing jury awards based on both state law and federal law claims. Compare, e.g., Abel v. Tomi Sports Int'l, LLC, No. 09 Civ. 10388, 2012 WL 6720919, at *14 (S.D.N.Y. Dec. 18, 2012) (“As the jury in this case made its award under both federal and state law, and as the successful plaintiff should be paid under the theory of liability that provides the most complete recovery, the Court reviews the award under the federal standard.” (citations, alteration, and internal quotation marks omitted)), with Mason v. City of New York, 949 F.Supp. 1068, 1075 (S.D.N.Y. 1996) (“Because the jury awarded lump sum damages encompassing both the federal and state claims, the Court must determine excessiveness under both federal and state law.”). Because the result is the same whether we apply the state standard or the federal standard, we need not decide this issue here.
“We have long held that, when damages are awarded, calculation of damages is the province of the jury.” Zeno, 702 F.3d at 671 (internal quotation marks omitted). Therefore, “in reviewing damages awards, we accord considerable deference to the factual findings of both judge and jury. Although a review of comparable cases is appropriate, we need not average the high and low awards; we focus instead on whether the verdict lies within the reasonable range.” Id. (citation, alterations, and internal quotation marks omitted). “This Court’s function upon review of the *588district court’s decision is limited to determining whether the district court abused its discretion.” Patterson, 440 F.3d at 119-20.
The district court here plainly did not abuse its discretion in holding that the amount awarded, $1 million per year of wrongful incarceration per plaintiff, neither shocks the conscience nor materially deviates from reasonable compensation. Indeed, the jury award is in line with other approved awards in wrongful conviction cases. See, e.g., Limome v. United States, 579 F.3d 79, 103-07 (1st Cir. 2009) (approving district court award of $1,000,000 per plaintiff per year for wrongful incarcerations of between eighteen and thirty-three years, but noting that this was at the high end of what would be permissible); Smith v. City of Oakland, 538 F.Supp.2d 1217, 1242 (N.D. Cal. 2008) (“[A] number of verdicts of approximately $1 million per year have been awarded in cases involving periods of wrongful incarceration which span a significant amount of time.” (collecting cases)); Sital v. City of New York, 60 A.D.3d 465, 875 N.Y.S.2d-22, 23 (1st Dep’t 2009) (acknowledging reasonableness of a $150,000 award for a false arrest claim with an associated twenty hours imprisonment and approving a $1.6 million award for a malicious prosecution claim with an associated eleven months imprisonment); Haynes v. City of New York, 29 A.D.3d 521, 815 N.Y.S.2d 143, 145 (2d Dep’t 2006) (acknowledging reasonableness of a $1 million award on a malicious prosecution claim where the plaintiff was incarcerated for four months and of a $250,000 award for a false arrest claim). Plaintiffs and the district court cited scores of cases, federal and state, with awards in a similar range. The fact that some state cases, cited by Volpe, yielded lower damage awards does not mean that the district court’s conclusion was an abuse of discretion, as we “need not average the high and low awards,” and rather focus on “whether the verdict lies within the reasonable range.” Zeno, 702 F.3d at 671; see id. at 673 (rejecting contention that an award shocks the conscience when it was “located within the range of permissible decisions”). The evidence shows that, as can be expected, plaintiffs suffered grave harm from their 18 years of wrongful incarceration, including adverse psychiatric effects, loss of relationships with family members, and stigma due to the nature of the crimes they were convicted of—and resulting violence against them while incarcerated.29 The district court therefore did not abuse its discretion in holding that the jury award did not shock the conscience or materially deviate from what would be reasonable compensation.
VI. Attorneys’ Fees
Finally, Volpe argues that the district court’s award of $4,997,914.55 in attorneys’ fees to plaintiffs’ attorneys from the law firm Neufeld Scheck & Brustin, LLP (“NSB”) should be reduced to no more than $3,000,000. As discussed above, the district court compensated NSB attorneys at the Southern District of New York rates they had proposed, ranging from $250 an hour to $700 an hour for attorneys, and *589$125 an hour for paralegals, for 11,222.6 hours of attorney and paralegal work.
Section 1988(b) of Title 42 provides that, as is relevant here, in actions brought to enforce Section 1988, “the court, in its discretion, may allow the prevailing party, other than the United States, a reasonable attorney’s fee as part of the costs.” “[A] ‘reasonable’ fee is a fee that is sufficient to induce a capable attorney to undertake the representation of a meritorious civil rights case.” Perdue v. Kenny A. ex rel. Winn, 559 U.S. 542, 552, 130 S.Ct. 1662, 176 L.Ed.2d 494 (2010). The district court here derived the fee award through application of the lodestar method, multiplying the number of hours the court found that NSB attorneys reasonably spent litigating the ease by hourly rates it found reasonable. The Supreme Court has held that “the lodestar method yields a fee that is presumptively sufficient to achieve [the] objective” of Section 1988(b). Id.; see also Stanczyk v. City of New York, 752 F.3d 273, 284 (2d Cir. 2014). Volpe does not contest the method by which the court calculated attorneys’ fees, but contends that both the hourly rates and the number of hours reimbursed were excessive.
The Supreme Court has emphasized that “the determination of fees ‘should not result in a second major litigation.’ ” Fox v. Vice, 563 U.S. 826, 838, 131 S.Ct. 2205, 180 L.Ed.2d 45 (2011) (quoting Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S.Ct. 1933, 76 L.Ed.2d 40 (1983)). Instead, it has instructed:
The fee applicant ... must, of course, submit appropriate documentation to meet the burden of establishing entitlement to an award. But trial courts need not, and indeed should not, become green-eyeshade accountants. The essential goal in shifting fees ... is to do rough justice, not to achieve auditing perfection. So trial courts may take into account their overall sense of a suit, and may use estimates in calculating and allocating an attorney’s time.
Id. (citations and internal quotation marks omitted).
Further, our review of the district court’s fee award is limited and deferential, as we review only for abuse of discretion. See id. (“[AJppellate courts must give substantial deference to these determinations [regarding fee awards], in light of the district court’s superior understanding of the litigation. We can hardly think of a sphere of judicial decisionmaking in which appellate micromanagement has less to recommend it.” (internal quotation marks and citations omitted)); Merck Eprova AG v. Gnosis S.p.A., 760 F.3d 247, 265-66 (2d Cir. 2014) (“Given the district court’s inherent institutional advantages in this area, our review of a district court’s fee award is highly deferential.” (internal quotation marks omitted)). “Indeed ‘abuse of discretion’—already one of the most deferential standards of review—takes on special significance when reviewing fee decisions.” Goldberger v. Integrated Res., Inc., 209 F.3d 43, 47 (2d Cir. 2000). The district court has presided from the filing of the complaint in 2008 until the appeal was filed at the end of 2014, overseeing two trials, a lengthy Daubert hearing, and numerous pre- and post-trial motions, and clearly has an intimate familiarity with the record and lawyering of this case; as such, we should be especially deferential to the district court’s findings with respect to the reasonableness of attorneys’ fees. Bearing in mind this background, we turn to Volpe’s arguments regarding the district court’s fee award.
A. Hourly Rates
First, Volpe contends that the district court erred in awarding attorneys’ fees *590based on the higher rates typical of the Southern District of New York, where NSB is located, rather than the slightly lower rates awarded in the Eastern District of New York, where the case was tried.
When determining attorneys’ fees under the lodestar approach, courts apply the forum rule. “According to the forum rule, courts should generally use the hourly rates employed in the district in which the reviewing court sits in calculating the presumptively reasonable fee.” Simmons v. N.Y.C. Transit Auth., 575 F.3d 170, 174 (2d Cir. 2009) (internal quotation marks omitted); see also Gierlinger v. Gleason, 160 F.3d 858, 882 (2d Cir. 1998) (providing that, in calculating the lodestar amount, the rates used generally “are the market rates prevailing in the community for similar services by lawyers of reasonably comparable skill, experience, and reputation” (internal quotation marks omitted)). “[I]n order to receive an attorney’s fee award based on higher out-of-district rates, a litigant must overcome a presumption in favor of the forum rule, by persuasively establishing that a reasonable client would have selected out-of-district counsel because doing so would likely (not just possibly) produce a substantially better net result.” Simmons, 575 F.3d at 172; see also id. at 174 (“The court may apply an out-of-district rate (or some other rate, based on the aforementioned case-specific variables) if, in calculating the presumptively reasonable fee, it is clear that a reasonable, paying client would have paid those higher rates.” (alterations and internal quotation marks omitted)). Indeed, the “touchstone” of the forum rule doctrine, as well as for awarding attorneys’ fees in civil rights cases, is that “district courts should award fees just high enough to attract competent counsel.” Id. at 176 (some internal quotation marks omitted) (quoting Arbor Hill Concerned Citizens Neighborhood Ass’n v. County of Albany, 493 F.3d 110, 121 (2d Cir. 2007), amended on other grounds, 522 F.3d 182 (2d Cir. 2008), and 575 F.3d 170 (2d Cir. 2009). Therefore, to rebut the presumption that the forum rule applies, the party seeking higher fees “must make a particularized showing, not only that the selection of out-of-district counsel was predicated on experience-based, objective factors, but also of the likelihood that use of in-district counsel would produce a substantially inferior result.” Id. Relevant factors include whether counsel has “special expertise in litigating the particular type of case, if the case is of such nature as to benefit from special expertise,” whether any “in-district counsel possessed such expertise,” and whether “local counsel possessing requisite experience were unwilling or unable to take the case.” Id. at 175-76.
The district court, in response to NSB’s request for Southern District of New York rates, rather than Eastern District of New York rates, concluded that NSB had made the necessary particularized showing sufficient to overcome the forum rule. The court found that NSB has particular experience litigating wrongful conviction suits and has successfully litigated dozens of these cases nationwide, and that this experience was necessary given that “this case was exceedingly arduous and complex, involving extensive expert analysis and presentation of DNA evidence, spanning eight years and two trials.” Supp. Special App’x at 7. The district court further noted that defendants “do not dispute that no lawyers with primary offices in the Eastern District of New York have obtained a successful jury verdict in a [Section] 1983 wrongful conviction suit, as NSB did here.” Supp. Special App’x at 7. Finally, the court found that NSB achieved an excellent result in a particularly complicated case.
*591Having reviewed this record, we conclude that the district court did not abuse its discretion in applying higher out-of-district rates in the lodestar calculation. Volpe argues that the district court, in deciding to apply the higher rates, placed undue weight on the jury award obtained and its conclusion that NSB achieved an excellent result. We disagree. The district court properly considered the relevant factors, which can include litigation results where “they are a direct result of the quality of the attorney’s performance,” Millea v. Metro-North R.R. Co., 658 F.3d 154, 168 (2d Cir. 2011), and where litigation results can inform whether obtaining out-of-district counsel was likely to produce a substantially better net result. Volpe also claims that the case and underlying facts were not complex, and that expert counsel was not required. But the district court was in a better position than we are to assess the complexity of the case and the need for expert counsel, and the district court did not abuse its discretion in holding that this factor, as well as several others, pointed toward application of the forum rule being inappropriate.
Volpe also contends that a higher award was not necessary to attract out-of-district counsel because NSB took the case knowing that the forum rule meant that there was a presumption in favor of the lower Eastern District rates. Volpe further suggests that NSB has its own Eastern District rates, based on its fee submission to the district court. Contrary to Volpe’s intimations, NSB does not have a separate set of rates it charges to clients when litigating in the Eastern District of New York. Rather, the law firm provided rates it thought were reasonable if the district court chose to apply the forum rule, in addition to the rates it thought were reasonable if Southern District rates were applied. Further, if the standard were as Volpe claims, no out-of-district attorney would be able to rebut the forum rule’s presumption because of the mere fact that he took the case. The district court properly considered the relevant factors and evidence, and did not abuse its discretion in awarding the higher Southern District of New York rates.
B. Number of Hours
Volpe also challenges the number of hours reimbursed by the district court. Volpe contends that the hours for which NSB was reimbursed should have been reduced by at least 25% due to block billing and failure to keep contemporaneous time records, billing for claims that were dismissed or withdrawn, and excessive hours spent on the fee award application. As noted, NSB requested reimbursement for a total of 11,222.6 hours of attorney time over the eight years this case was litigated in the district court.
First, attorneys are required to keep and submit contemporaneous records with their fee applications, absent unusual circumstances outside the attorney’s control. See Scott v. City of New York, 643 F.3d 56, 58 (2d Cir. 2011). But, despite Volpe’s claims, the evidence demonstrates that NSB did keep contemporaneous records. Indeed, NSB did not request reimbursement for attorney time spent where the attorney or paralegal did not keep contemporaneous records or where the contemporaneous records were lost.
Volpe next asserts that block billing rendered NSB’s records inadequate. Although block billing is not preferred, it is permissible so long as the records allow the court to conduct a meaningful review of the hours requested. See Merck Eprova, 760 F.3d at 266 (affirming where the district court rejected defendant’s challenges to the use of block billing, vague descriptions of work, and overabundant staffing *592because “trial is a time consuming process” and the district court did not abuse its discretion in finding the hours expended were reasonable). The district court did not err in rejecting this challenge. The district court was able to adequately review the hours requested, struck hours of travel time billed as attorney work time, and did not abuse its discretion in finding the hours expended to be reasonable.
Next, Volpe challenges the district court’s decision to award fees based on claims other than plaintiffs’ prevailing claims against Volpe. “[Pjlaintiffs may receive fees under [Section] 1988 even if they are not victorious on every claim,” as a “civil rights plaintiff who obtains meaningful relief has corrected a violation of federal law and, in so doing, has vindicated Congress’s statutory purposes.” Fox, 563 U.S. at 834,131 S.Ct. 2205.
A court should compensate the plaintiff for the time his attorney reasonably spent in achieving the favorable outcome, even if “the plaintiff failed to prevail on every contention.” The fee award, of course, should not reimburse the plaintiff for work performed on claims that bore no relation to the grant of relief: Such work “cannot be deemed to have been expended in pursuit of the ultimate result achieved.”
Id. (quoting Hensley, 461 U.S. at 435, 103 S.Ct. 1933). Although attorneys should not be reimbursed for then- work on claims that “bore no relation to the grant of relief,” id. “[wjhere the district court determines that the successful and unsuccessful claims are inextricably intertwined and involve a common core of facts or are based on related legal theories, it is not an abuse of discretion for the court to award the entire fee,” Reed v. A.W. Lawrence & Co., 95 F.3d 1170, 1183 (2d Cir. 1996) (alteration and internal quotation marks omitted). The district court determined here that successful claims were inextricably intertwined with the unsuccessful claims, and that the successful and unsuccessful claims involved both a common core of facts and were based on related legal theories. Having reviewed the record, keeping in mind the superior position of the district court, we conclude that this finding was not clearly erroneous, and the district court therefore did not abuse its discretion in awarding the entire fee.
Finally, Volpe challenges the award of $97,132.50 for the time spent preparing the fee award application. In particular, Volpe contends that if NSB kept better time records, it would not have taken as long to prepare the fee award, and that NSB should not be reimbursed for time spent arguing that the forum rule should not apply. Reasonable attorneys’ fees for preparing the fee application are compensable. See Reed, 95 F.3d at 1183. Here, the district court rejected Volpe’s challenge that NSB spent an excessive number of hours preparing the fee application on the ground that the “brief does not cite to any relevant precedent and makes only conclusory arguments.” Supp. Special App’x at 14 n.6. This determination was not an abuse of discretion, as “when, as here, a fee target has failed to offer either countervailing evidence or persuasive argumentation in support of its position, [it is not] the district court’s job either to do the target’s homework or to take heroic measures aimed at salvaging the target from the predictable consequences of self-indulgent lassitude.” United States ex rel. Keshner v. Nursing Pers. Home Care, 794 F.3d 232, 236 (2d Cir. 2015) (alterations omitted) (quoting Foley v. City of Lowell, 948 F.2d 10, 21 (1st Cir. 1991)).
For the foregoing reasons, we affirm the district court’s award of $4,997,914.55 in attorneys’ fees to NSB.
*593CONCLUSION
We have considered all of Volpe’s arguments on this appeal, and have found no basis for reversal. Accordingly, the judgment of the district court is AFFIRMED.
. Volpe passed away on January 3, 2011. Thereafter, Carolann Hessemann, the executor of Volpe's estate, was substituted as a party. Before he passed away, Volpe was deposed over the course of several days by plaintiffs, but he passed away before he could be examined by defense counsel.
. At trial, Dr. Tamara Bloom, the supervising medical examiner of Nassau County, testified, as the doctor who had conducted Ms. Fusco's autopsy had since passed away. She based her testimony on the autopsy reports and photographs in this case.
. Unless otherwise noted, all transcript cites are to the transcripts of the second trial.
. On recross, she was asked, "Dr. Bloom, it is not that you have no idea. Your opinion of the ligature instrument, if it was a rope, was a shorter rope; is that correct?” Trial Tr. at 726. Dr. Bloom responded, “I would just picture it to be shorter in order that it was so efficient to have killed tills girl. I cannot exclude that it is a longer rope.” Trial Tr. at 726.
.The hairs found in Restivo’s van were called questioned hairs, or Q hairs. Two hairs belonging to Ms. Fusco were initially found in Restivo’s van. These hairs were found in the "Q8 quadrant” of the van, under the front passenger’s seat and were therefore called Q8 hairs, specifically Q8-A and Q8-B. At Kogut’s retrial, the prosecution also offered DNA evidence from a third hair purportedly found in Restivo’s van, the Q4 hair. We will refer to these three hairs matching Ms. Fusco collectively as the "Q hairs.”
. That is, these hairs could have come from Ms. Fusco. Following defendants’ convictions, DNA analysis was conducted on the Q hairs. Although the results of these DNA tests do not appear in the record, there is no dispute between the parties that the hairs actually belong to Ms. Fusco.
. Later, after the district court granted Resti-vo and Halstead, but not Kogut, a new trial, the cases were unconsolidated.
. Spillane was sued as Sean Spillane, but his name is Shaun Spillane.
. Our Court affirmed the district court’s decision as to Kogut in Kogut v. County of Nassau, 789 F.3d 36 (2d Cir. 2015).
. This includes the man with whom Volpe speculates that Ms. Fusco might have been having sex without telling her family or friends because of his race.
. Russell Fischer, plaintiffs’ police practices expert, who is discussed in detail below, also testified about the police’s interrogation of Cockerel. Fischer testified about risks from long interrogations, and testified that if a witness’s story changed over time, under minimally accepted police practices, an officer should record the changes and corroborate the statements. Fischer further testified: "I can't tell who’s telling the truth and not the truth. I’m here to talk about police procedures. All I am saying is that an individual who repeatedly changes their story, adds things, deletes things, that testimony if it isn’t corroborated, isn’t verified by some independent party, has questionable reliability.” Ap-pellees’ Supp. App'x at 914.
.French reported that his car was stolen to the Lynbrook Police immediately after he realized it was missing. Apparently, later, either French or Gabberty saw or heard of news coverage related to Ms. Fusco’s murder and surmised that the theft might have been connected to the murder.
. As noted, Ms. Fusco was last seen leaving Hot Skates at 9:47 PM that night.
. The Nassau County homicide detectives operated independently from the Lynbrook Police.
. On the topic of these witnesses’ expertise, Volpe argues that Dr. DeForest lacks the qualification and expertise to opine on the decomposition process, citing an admission by Dr. DeForest that the causes of PMRB are not his area of expertise. But Dr, DeForest was not called on to testify about the causes of PMRB.
. Fraas had testified at the criminal trial that he could not possibly have finished the microscopic hair comparison in four days.
. Initially, NSB had also requested reimbursement at the attorney rate for some hours of travel time, but when the issue was raised by Volpe, NSB "aslc[ed] the Court to omit these entries completely from its time request,” which the court did. Supp. Special App’x at 13-14.
. We reject several competing readings of Boyd offered by the parties and the district court. The district court's reading of Boyd as excluding all evidence that was inadmissible in the criminal trial places too high of a burden on defendants in malicious prosecution cases, as it prohibits them from relying on evidence that only retrospectively was deemed inadmissible, even if, at the time the prosecution was brought, the defendants reasonably believed the evidence would ultimately be ruled admissible.
Plaintiffs assert that our Court affirmed the district court’s reading of Boyd in Kogut v. County of Nassau, 789 F.3d 36 (2d Cir. 2015). We disagree. Although our Court assumed in Kogut that the district court’s reading of Boyd was correct, it did not have the opportunity to decide this issue. In that appeal, Kogut challenged the admission of three statements made by Restivo and Halstead, arguing that those statements should not have been admitted at the civil trial as evidence of probable cause. Our Court rejected this challenge, reasoning that Kogut did not establish that the admitted statements caused any actual prejudice, Id. at 44-45. The Court therefore did not assess the district court’s reading of Boyd. *571Finally, Volpe's reading of Boyd as excluding evidence that is only inadmissible at the civil trial both cannot be squared with the text of Boyd and renders the holding of Boyd a tautology: evidence that is inadmissible at the civil trial cannot be admitted at that trial.
. The Kogut confession was not shown to fit within any hearsay exception as to Restivo and Halstead.
. This error was compounded by repeated arguments by defendants, from opening statements through summations, that the confession demonstrated that the three plaintiffs were factually guilty of the crime. See, e.g., Appellees’ Supp. App’x at 292-97 (defense counsel arguing in summation at the first trial that the confession contained numerous facts known only to Kogut and not the police and therefore was accurate).
.Because the jury was instructed to award only compensatory damages, not punitive damages, and each plaintiff suffered only one injury from both claims, vacating the jury verdict on the malicious prosecution claim would not affect the jury's damage award.
. As discussed above, we have some doubts regarding the district court’s ruling on the admissibility of the Kogut confession to the malicious prosecution claim. But because the jury was not asked to return a special verdict, the jury award would not be affected so long as there were no reversible errors on the fair trial claim.
. Accordingly, the district court permitted defendants to elicit that Restivo made a statement to police and that he did not want others to know about it.
. To the extent that this statement could have been admitted as potentially relevant to showing Restivo's knowledge of the crime, the district court properly excluded it on Rule 403 grounds. Fed. R. Evid. 403.
. Further, we note that Domzalski’s study was introduced into evidence at the second trial,
. Volpe did not argue in his opening brief that the district court’s charge incorrectly defined the relevant Brady standard or the defí-nition of materiality. Accordingly, we decline to consider that argument. See McBride v. BIC Consumer Prods. Mfg. Co., 583 F.3d 92, 96 (2d Cir. 2009) ("[W]e ordinarily will not consider issues raised for the first time in a reply brief.”).
. Defendants requested an instruction on double recovery that plaintiffs were to be *587compensated only once for the injuries actually suffered, even though plaintiffs claimed their rights were violated in two ways. Accordingly, the district court instructed the jury, “You may award compensatory damages only for the injuries that plaintiffs prove were proximately caused by defendant Volpe's wrongful conduct. Although defendant Volpe was found liable for two causes of action, each plaintiff is entitled to one recovery. The damages that you award must be fair compensation for all of the plaintiff's damages, no more and no less," Appellees’ Supp. App’x at 1525, and the jury rendered one award of $18 million to Restivo, and one award of $18 million to Halstead.
. Volpe asserts that because Restivo and Halstead were adults at the time of their arrests, because their convictions allegedly resulted from the inculpatory statements they allegedly made, and because Volpe estimates Restivo and Halstead's lifetime earnings would likely have been between $50,000 and $75,000 a year, the $18 million awarded to each man must have included non-pecuniary damages so high as to shock the conscience. We easily reject this challenge, and defer to the jury and district court’s determination of the appropriate measure of damages.