Overview
Today we deal with the fallout from a deadly drug conspiracy in Puerto Rico involving a small army of criminals affiliated with “the Combo of Dr. Pila” (from now on, “the Combo”), a vicious gang named after a local housing project where members ran one of their many drug points. The five defendants whose joint trial led to these consolidated appeals are Acosta, Fournier, Castillo, Rodríguez, and Guzmán (their full names and aliases appear in our case caption).
A federal grand jury indicted each of them for conspiring to possess and distribute illegal drugs within 1,000 feet of a public-housing facility. See 21 U.S.C. §§ 841, 846, and 860. The grand jury also indicted Fournier, Rodríguez, and Guz-mán — but not Acosta and Castillo — for aiding and abetting the use or carrying of a firearm “during and in relation to” a drug *186crime or the possession of a firearm “in furtherance” of that crime. See 18 U.S.C. §§ 2 and 924(c)(1)(A). As shorthand, we shall refer to these counts as the drug-conspiracy count and the gun count.
Covering the period from January 2003 to July 2007, the indictment tagged Rodriguez and Guzmán as “leaders” in the Combo conspiracy, Acosta and Castillo as “sellers,” and Fournier as a “facilitator.” These five were not the only ones indicted. Far from it. The grand jury also indicted 90 others on similar charges. But some of them copped pleas and agreed to testify for the government at our defendants’ trial.
After hearing what these and other witnesses had to say, the jury filled out defendant-specific verdict forms, finding, essentially, each defendant guilty as charged and picking drug-weight ranges for the drugs each defendant conspired to possess and distribute — all while using a beyond-a-reasonable-doubt standard. The only slight wrinkle on the conviction front is that the jury found Acosta — and Acosta only — not guilty of participating in drug-related conspiracy activities within 1,000 feet of a public-housing project. Later, the district judge imposed the following prison sentences: Acosta, 151 months on the drug-conspiracy count; Fournier, 78 months on the drug-conspiracy count plus 60 consecutive months on the gun count; Castillo, 120 months on the drug-conspiracy count; Rodriguez, 240 months on the drug-conspiracy count and 60 consecutive months on the gun count; and Guzmán, life on the drug-conspiracy count plus 60 consecutive months on the gun count.
Their appeals raise a staggering number of issues for review, though not all require our extended attention. To make our opinion manageable, we sort the issues out person by person, highlighting only those facts needed to put things in perspective. And for anyone wishing to know our ending up front, we note that when all is said and done we affirm across the board.
Acosta
(1)
Public THal
Acosta starts things off by accusing the district judge of closing the courtroom to “the public” during the jury-selection process. The judge’s action, he insists, denied him his Sixth Amendment right to a public trial. See U.S. Const, amend. VI. The backstory, at least so far as the record discloses, may be swiftly summarized.
Just before picking the jury, the district judge called counsel to sidebar. “I’ve been informed by my [court-security officer],” the judge said, “that the marshals informed him that three buses” that looked like “school buses” had “arrived here with persons who have T-shirts saying, ‘Danny, we support you and we back you.’ ” “Danny” is defendant Guzmán. Anyway, “I’m not going to allow that,” the judge added, “and none of those persons are going to walk into the courtroom. They are going to be sent back[,] and they are going to be — ”. Guzmán’s counsel interrupted, saying, “I don’t think that’s appropriate. I had no idea. Send them back.” At the risk of stating the obvious, context makes clear that counsel was calling the bus-riders’ actions inappropriate, not the judge’s ruling. And while the judge did not blame the lawyers, he did stress that he would not “tolerate any activity like that from any of the defendants.” Tell “you[r] clients” to “behave,” he continued, or else “I’m going to exclude them from the courtroom,” install “a camera next” to “the holding cells,” and let them “watch the trial from there.” “Very well,” Rodriguez’s lawyer said, followed immediately by the judge’s saying, “I’m going to order *187the marshals to remove them from the court.”
Every criminal defendant has a Sixth-Amendment right to a public trial— a right designed to ensure a fundamentally fair process, since the public’s very presence there helps keep judges, prosecutors, and witnesses on their toes. See, e.g., Waller v. Georgia, 467 U.S. 39, 46-47, 104 S.Ct. 2210, 81 L.Ed.2d 31 (1984). But this public-trial right is not absolute and must be balanced against other important con siderations in the administration of justice. See, e.g., id. at 45, 104 S.Ct. 2210. For example, a judge may close the courtroom to all members of the public if he detects a compelling interest that needs protecting, considers sensible closure alternatives, ensures that closure is no broader than required, and makes findings sufficient to support his ruling. See, e.g., Presley v. Georgia, 558 U.S. 209, 213-14, 130 S.Ct. 721, 175 L.Ed.2d 675 (2010) (adding that when a defendant objects to a closure but does not offer alternatives, the judge must think of some on his own); Owens v. United States, 483 F.3d 48, 61-62 (1st Cir.2007). On the other hand, a judge may order a partial courtroom closure — partial, because only some members of the public are kept out — if he pinpoints a substantial interest that needs protecting and then does the other things that we just listed. See, e.g., Bucci v. United States, 662 F.3d 18, 23 (1st Cir.2011) (citing United States v. DeLuca, 137 F.3d 24, 34 (1st Cir.1998)).
Acosta is having a devil of a time explaining what type of closure happened here. First he says that the judge barred “the public in general” during this phase of the trial, not just Guzmán’s t-shirt-wearing supporters, apparently. Backing off a bit, then he says (emphasis ours) that at the very “least” the judge barred “defendants’ families.” Later still, he says that the judge “may” have barred his family, friends, and supporters (hardly a take-it-to-the-bank kind of statement). And he says all this without citing to the record.
Even pushing that failure aside, Acosta cannot overcome this problem: The judge put the exclusion matter squarely on the table for all the defendants’ lawyers at sidebar, explaining what he intended to do with the t-shirt wearers. Each attorney had the chance to speak up. And attorneys for two of Acosta’s codefendants did precisely that. But not Acosta’s lawyer— he said nothing, despite the judge’s placing the issue front and center and the other lawyers’ voicing their opinion on this weighty subject. Fournier’s attorney peeped no words of protest either, which is a problem for him, as we shall shortly see. Ultimately, then, given the particular facts of our case, we conclude that Acosta’s lawyer had to know that he had to chime in on the exclusion issue — as others had — or else waive any claim. So his silence constitutes classic waiver, rather than forfeiture, which means that he cannot challenge the judge’s ruling even as plain error. See, e.g., United States v. Christi, 682 F.3d 138, 142 (1st Cir.2012) (Souter, J.) (citing, among other cases, Levine v. United States, 362 U.S. 610, 80 S.Ct. 1038, 4 L.Ed.2d 989 (1960), which deems a courtroom-exclusion issue waived if counsel knew about the exclusion and “saw no disregard of a right” but now peddles the argument “as an afterthought on appeal”).
(2)
Alibi Witness
Acosta also challenges the judge’s decision banning him from calling his wife as an alibi witness at trial. Here is what happened.
Shortly after the indictment issued, the government, relying on Fed.R.Crim.P. 12.1, demanded that each defendant give *188notice of any alibi defense that they intended to use. Rule 12.1(a) provides that a request like that “must state the time, date, and place of the alleged offense.” And the version of the Rule in effect at the relevant time said that if the defendant wishes to raise an alibi defense, he must— within 10 (now 14) days — notify the government in writing of his intent to do so. And the government’s request here stated:
The defendant committed the charged offense throughout the years 2003, 2004, 2005, 2006, and until July 2007, during the day and night hours, in or within the Dr. Pila Public Housing Project; and/or within the Ponce Housing Public Housing Project; the José M. Gándara Public Housing Project; the Portugués Public Housing Project; the Los Rosales Public Housing Project; and within the Municipalities of Ponce and Juana Diaz, Puerto Rico.
One of Acosta’s then-codefendants, José Ríos Santiago (“Ríos”), who later pled guilty to the drug-conspiracy count, objected to the request, claiming that it was too broad because it covered four-plus years, every day and night. “If the government wants an alibi response,” Rios wrote, it must list the “dates” and “time” showing when he supposedly “was personally involved in the conspiracy charged,” and not use the entire period covered in the indictment — then and only then, he added, will notice be “adequate” under the Rule. Acosta neither filed his own objection nor joined Rios’s. Eventually, the judge granted Rios’s objection, using language indicating that the ruling applied only to him: “ORDER as to Jose L. Rios-Santiago GRANT[ED].” The government never amended its request. And Acosta never mentioned during pretrial that he intended to call his wife as an alibi witness.
We fast-forward to August 25, 2009, the thirteenth day of trial. A confidential informant named Ulises Martinez Camacho (“Martinez”) testified about his encounter with Acosta at a Combo-owned drug point — the “Coto Laurel” drug point — on July 11, 2007. Armed with police-supplied recording equipment and following the police’s marching orders, Martinez had journeyed there to buy drugs and ended up handing Acosta $80 for 13 bags of crack. Acosta grabbed the cash but never came back with the drugs. Martinez immediately complained about what had happened to Acosta’s brother, who in turn told defendant Rodriguez (who was also there at that time). Ticked off that “this guy” — meaning Acosta — had “done this again,” Rodriguez handled the problem by making sure that Martinez got what he had paid for. Importantly, Martinez had secretly caught nearly everything on audio and video tape — everything except a shot of Acosta, which Acosta’s counsel brought out on cross-examination.
After the judge recessed for the day, Acosta’s lawyer said that he wanted to call his client’s wife as an alibi witness. She would testify that Acosta had been with her on July 11, counsel explained, celebrating her birthday far from the Coto Laurel drug point. But prosecutors told the judge that they had given the defense copies of the July 11 recording way back in August 2007 — nearly 2 years before trial. And one can hear people saying Acosta’s name — “José David” or “David” — on the recordings, prosecutors stressed. They also said that they had given the defense a transcript of the audio recording in October 2007. And they noted that they had given the defense “Jencks Act” material regarding the July 11 transaction on August 1, 2009 — 6 days before trial started and 24 days before either Martinez’s direct testimony or Acosta’s in-court attempt to spring the alibi witness on the prosecution. For those uninitiated in the intricacies of federal criminal procedure, the Jencks Act *189entitles a criminal defendant to the “statement” of a government witness after the witness has testified on direct examination. See 18 U.S.C. § 8500(a).1
Acosta’s counsel did not deny any of this. Instead, he tells us, he tried to explain to the judge that, yes, prosecutors “may have” dropped off videos and transcripts of the July 11 doings, but he did not learn that Martinez would tie Acosta to the tape until he got the Jencks material about a week before trial. The judge was not impressed. “You had all that information” for such a long “time,” the judge found, yet you still did not give prosecutors the requisite notice so that they could “investigate” the alibi theory. “I have to consider the rights of your client,” the judge told Acosta’s attorney. And “I did,” the judge said, reminding counsel too that he had ordered prosecutors to produce all Jencks-Act statements well “before the time that the government ha[d] to turn it over.” Having concluded that the defense had failed to comply with Rule 12.1, the judge precluded Acosta’s wife from testifying. See Fed.R.Crim.P. 12.1(e) (providing that “a court may exclude the testimony of any undisclosed” alibi witness if a party does not comply with the Rule’s requirements).
Looking to undo the judge’s ruling, Acosta protests that he did not. have to give advance notice of his alibi witness because the prosecution had given him an inadequate Rule 12.1 request. But even if wrong on this point, Acosta contends that the judge should have used the power under Rule 12.1(d) to let the alibi witness testify anyway, particularly since the exclusion robbed him of his constitutional rights to present a meaningful defense, or so he says.2 Neither argument works.
Take the first one. At no time before dropping the alibi-witness bombshell mid-trial did Acosta object to the government’s Rule 12.1 request — despite then-eodefen-dant Rios’s objection and the judge’s ruling in Rios’s favor. Under these circumstances, Acosta has waived the argument. See, e.g., United States v. Valerio, 676 F.3d 237, 246 n. 2 (1st Cir.2012) (noting that arguments raised for the first time on appeal are deemed waived); United States v. Meade, 175 F.3d 215, 223-24 (1st Cir.1999) (same). Hoping to avoid this logic, Acosta faintly intimates a-slight whisper of a suggestion that Rios’s objection preserved the issue for all defendants. But he does not develop this piggyback theory, so we need say no more about that. See, e.g., United States v. Zannino, 895 F.2d 1, 17 (1st Cir.1990) (holding “that issues adverted to in a perfunctory manner, unaccompanied by some effort at developed argumentation, are deemed waived”).
Explaining the problem that sabotages Acosta’s second argument — that the judge should have plied his Rule 12.1(d) power to excuse the surprise-witness disclosure — requires a little more work on our part. We start with some basics. A defendant obviously has a right to offer witnesses in his defense, thanks to the Supreme Court’s reading of the Sixth Amendment.3 See Taylor, 484 U.S. at 409, *190108 S.Ct. 646; see also United States v. Portela, 167 F.3d 687, 704 (1st Cir.1999). But just as obviously, that right (like most rights) is not unlimited and may bow to other “[c]ompeting interests.” United States v. Brown, 500 F.3d 48, 57 (1st Cir.2007) (citing Taylor, 484 U.S. at 414-15, 108 S.Ct. 646). Among these are “the integrity of the adversary process, the danger of unfairly skewing the truth-determining function that lies at the epicenter of that process, and the efficient administration of justice.” Id. (citing Taylor again). Also relevant are “the willfulness [or not] of the violation, the relative simplicity of compliance, and whether or not some unfair tactical advantage has been sought.” United States v. Nelson-Rodriguez, 319 F.3d 12, 36 (1st Cir.2003) (quoting Chappee v. Vose, 843 F.2d 25, 29 (1st Cir.1988)). Not every factor applies in every situation, naturally. See Chappee, 843 F.2d at 29 (emphasizing that the Supreme Court has “declined to cast a mechanical standard to govern all possible cases”). And in the end, we give fresh review to the judge’s application of this standard, see Nelson-Rodriguez, 319 F.3d at 36, but with “considerable deference” to his factual determinations, see United States v. Levy-Cordero, 67 F.3d 1002, 1013 (1st Cir.1995).
Acosta’s effort to poke holes in the judge’s exclusion ruling first focuses on the willfulness factor: because, he says, there is zero indication that counsel had acted willfully when he disclosed the alibi witness midtrial, the judge’s edict must fall. What dooms his theory is that we have never held that the exclusion sanction is available only when a party willfully violates Rule 12.1. See Nelson-Rodriguez, 319 F.3d at 37. Next he argues that we must reverse given how badly he needed his wife’s testimony to rebut Martinez’s and so vindicate his constitutional right to mount an effective defense. But Acosta had more than enough time to reveal his alibi witness before day 13 of trial, as the judge supportably found after hearing prosecutors say how they had handed the defense incriminating July 11 evidence (recordings, transcripts, and reports), starting almost two years before the trial kicked off. Add to that the lack of a credible excuse for not complying with Rule 12.1 and the fact that allowing the surprise witness to testify would have delayed this multidefendant trial (to give prosecutors a chance to investigate the alibi), and it becomes clear that Acosta’s alibi-witness argument is a no-go. See generally Williams v. Florida, 399 U.S. 78, 81-82, 90 S.Ct. 1893, 26 L.Ed.2d 446 (1970) (noting that because an alibi defense is easily “fabricated,” the government’s “interest in protecting itself against an eleventh-hour defense is both obvious and legitimate,” adding that defendants must know that a criminal trial “is not ... a poker game in which players enjoy an absolute right always to conceal their cards until played”).
(3)
Sufficiency of the Evidence
Acosta contends that the evidence was not sufficient to support his drug-conspiracy conviction. To convict someone of that crime, the government must prove beyond a reasonable doubt that he knew about and voluntarily participated in the conspiracy, “intending to commit the underlying substantive offense” — and proof may come from direct evidence or circumstantial evidence, like inferences drawn “from members’ ‘words and actions’ ” and from “ ‘the interdependence of activities and persons involved.’” United States v. Ortiz de Jesús, 230 F.3d 1, 5 (1st Cir.2000) (quoting United States v. Boylan, 898 F.2d 230, 241-42 (1st Cir.1990)). Of course, *191winning a sufficiency challenge is hard to do: Acosta must show that after viewing the evidence and reasonable inferences in the light most flattering to the prosecution, no rational jury could have found him guilty beyond a reasonable doubt. See, e.g., United States v. Polanco, 634 F.3d 39, 45 (1st Cir.2011). And though we exercise de novo review, we can neither re-weigh the evidence nor second-guess the jury’s credibility calls. Id.
Acosta pins his principal hope on convincing us that two government witnesses who helped seal his fate were unworthy of belief. This argument is hopeless.
The first witness is Jayson Serrano, a confidential informant who had grown up with defendant Rodriguez. Serrano testified that Acosta was a drug “runner” for the Combo-owned Coto Laurel drug point: Rodriguez (the person in charge at that drug point) would get “bundles” of drugs from his Combo colleagues and hand them over to Acosta, who would then give them to sellers at the Coto Laurel locale.4 Early in his direct examination Serrano named two sellers at the Coto Laurel drug point — nicknamed “Wanda” and “Pucho” — neither of whom were Acosta. Later, though, Serrano did say that Acosta was a seller there as well. He knew this because he had bought drugs from him “several” times. Serrano also testified that he and Acosta had met with Combo leaders at Rodriguez’s house. Acosta tries to pour cold water on all this, noting that Serrano is a convicted thief, former drug addict, and paid government snitch who could not keep his story straight about whether Acosta was the seller. What Acosta says may be proper argument for a jury, but it is not proper argument here, given that we must resolve all evidentiary and credibility conflicts in the government’s favor. See, e.g., Polanco, 634 F.3d at 45; United States v. Manor, 633 F.3d 11, 14 (1st Cir.2011).
The second witness that Acosta targets is Martinez, the same Martinez involved in the July 11, 2007 incident discussed above — where Acosta agreed to sell Martinez $80 worth of crack but took off without handing over the drugs, and then defendant Rodriguez fixed things by making sure that Martinez got his crack. Well, Acosta says that that testimony shows only that he had preyed on hapless drug buyers, not that he was a conspiracy member. But his theory cannot fly, given our prosecution-friendly standard of review — which, again, requires us to choose from among competing inferences the one most compatible with the jury’s guilty verdict. See, e.g., Polanco, 634 F.3d at 45; Manor, 633 F.3d at 13-14. And having done so, we believe a sensible jury could conclude that Combo-bigwig Rodriguez helped Martinez out because Acosta was a drug seller in the Combo conspiracy and Rodriguez did not want any customer-relations problems on his watch. Like he did with Serrano, Acosta also harps on Martinez’s status as a professional stoolie and former drug addict — a credibility attack that fails, for the reason his attack on Serrano failed.
The upshot is that Acosta’s insufficient-evidence arguments misfire. So we trudge on.
(4)
Drug Quantity
For his final salvo, Acosta attacks the judge’s drug-quantity finding. As anyone familiar with this area of the law knows, *192sentence length in drug cases turns largely on the amount and type of drugs involved. Section 841(b)(1)(A) of Title 21, for example, lists amounts of different drugs — 1 kilogram or more of heroin, 5 kilograms or more of cocaine, for instance — that lead to sentences of 10 years to life (“unless death or serious bodily injury results from the use of such substance,” which leads to sentences of 20 years to life). Acosta says, basically, that the judge relied on conspiracy-wide amounts rather than on amounts attributable to him personally or reasonably foreseeable by him — a gaffe, he adds, that resulted in his getting a higher sentence. Commendably, he concedes that he did not object on this basis below and so must prove plain error — a famously difficult standard to meet, requiring him to show “error, plainness, prejudice to the defendant!,] and the threat of a miscarriage of justice.” United States v. Torres-Rosario, 658 F.3d 110, 116 (1st Cir.2011); accord United States v. Eisom, 585 F.3d 552, 556 (1st Cir.2009). This is a standard he ultimately cannot meet, it turns out.
Even a quick look at the record reveals that the jury made individualized drug findings for each defendant beyond a reasonable doubt: following the judge’s charge, the jury checked off lines indicating, for example, that Acosta had conspired to possess and distribute 1 kilogram or more of heroin, 5 kilograms or more of cocaine, 50 grams or more of crack, and less than 100 kilograms of marijuana. In sentencing him on the drug-conspiracy count, the judge used the jury’s 1-kilogram-or-more-of-heroin finding — a finding that triggered a statutory minimum-to-maximum prison range of 10 years to life, see 21 U.S.C. § 841(b)(1)(A), all without offending Alleyne v. United States, — U.S. -, 133 S.Ct. 2151, 186 L.Ed.2d 314 (2013), despite what Acosta says. Al-leyne held that facts triggering application of a “mandatory minimum sentence” generally “must be submitted to the jury” and established “beyond a reasonable doubt.” See 133 S.Ct. at 2163.5 That is precisely what happened here.6 Working with the 2009 version of the federal sentencing guidelines,7 the judge then pegged Acosta’s base-offense level at 32 and made no adjustments, either up or down. With a criminal-history category of III, Acosta’s guidelines-recommended sentencing range was 151 to 188 months of imprisonment, the judge concluded. And after considering relevant sentencing factors, see 18 U.S.C. § 3553(a), the judge selected 151 months — the very bottom of the applicable guidelines range and obviously well within the statutory range of 10 years to life.
Given this set of circumstances, the bottom line is very simple: the jury’s individualized drug-quantity findings still Acosta’s cry that no individualized findings drove this part of the judge’s sentencing decision. Ever persistent, Acosta suggests that the evidence before the jury concerning what amounts he had handled or were reasonably foreseeable by him was “iffy” at best, an argument that certainly sounds like a sufficiency challenge. But the al*193ready high bar for plain error becomes even higher when dealing with an unpre-served sufficiency-of-the-evidence claim (which this is), requiring a criminal defendant to show a “clear and gross injustice” for reversal. United States v. Pratt, 568 F.3d 11, 18 (1st Cir.2009). And significantly, Acosta makes zero effort to explain how — after taking the evidence and permissible inferences in a prosecution-friendly way — he satisfies this souped-up standard. And that means that he has not lived up to his obligation to “develop[] a sustained argument out of ... legal precedents,” which leads to waiver of this issue. See Town of Norwood v. Fed. Energy Regulatory Comm’n, 202 F.3d 392, 405 (1st Cir.2000).
Fournier
(1)
Public Trial
Like Acosta, Fournier first argues that the judge’s courtroom-closure order violated the Constitution’s public-trial guarantee. Fournier’s claim is really just a variation on Acosta’s and shares the same devastating waiver problem. So we soldier on.
(2)
In-Court Identification
Fournier next claims that a witness’s in-court identification of him occurred under circumstances so suggestive and unreliable as to deny him due process of the law. The background events are easily summarized. On the second day of trial, Miguel Lespier Velázquez (“Lespier”) took the stand as a government witness. Possessing a bachelor’s degree in criminal justice, Lespier was an early Combo leader. Eventually arrested and later indicted (together with our five defendants) for his role in this sordid affair, Lespier pled guilty and agreed to help the government prosecute other conspiracy members. After he had given a downright chilling account of a Combo killing spree, the prosecutor asked him who had supplied the guns. “Georgie Mesón,” Lespier said. Fournier sometimes goes by that name, everyone agrees, because he owned a sandwich shop called “El Mesón Bypass.” “Pistols, revolvers, and AK-47 rifles” are what Fournier provided, Lespier noted. And Lespier said that he knew him. The prosecutor then asked him to identify Fournier for the jury. But he could not, even after standing up and looking in the defendants’ direction. “I don’t see [Four-nier] right now,” he said. “I know him,” though, he stressed again.
Later in a sidebar conference, the prosecutor told the judge that he believed that Lespier could not make an identification because Fournier had kept his head down and a podium had blocked Lespier’s view. The solution, the prosecutor said, was to have the defendants move so that the witness could see them better. Fournier’s lawyer objected, and the judge denied the request.
The following day, the prosecutor continued his direct examination of Lespier. Fournier attended Combo meetings, Les-pier said. Other key attendees included representatives from different housing projects — like the Portugués Housing Project — that were part of the Combo network. Shifting gears, the prosecutor then asked him about defendant Rodriguez. Rodriguez “belonged to the Combo” since 2005 and ran the Coto Laurel drug point, Lespier said. And he identified Rodriguez for the jury — -“the person is wearing a polo shirt, white,” Lespier added — after leaving the stand (with the judge’s permission, of course) to get a closer look. “[Do] you recognize any other person?” the prosecutor asked. Lespier identified four of the *194defendants by their nicknames. “What about the person in the coat,” the prosecutor then asked, “do you recognize that person?” “Objection,” Fournier’s counsel said. “Suggestive.” After the judge overruled the objection, Lespier stated, “[tjhat’s Georgie Mesón,” i.e., Fournier, and then returned to the witness stand, where he continued testifying about Four-nier’s exploits- — including how Fournier supplied Combo cohorts with guns and cars to use in shootouts with Combo foes. Fournier even drove the getaway car following one gunfight, Lespier said.
A defendant challenging an in-court identification must show that the procedure used was unduly suggestive. See, e.g., United States v. Espinal-Almeida, 699 F.3d 588, 602 (1st Cir.2012), cert. denied, — U.S. -, 133 S.Ct. 1837, 185 L.Ed.2d 845 (2013). But even if he does, the identification is still admissible if the totality of the circumstances indicates that it was nonetheless reliable. Id. An identification is unreliable only when it poses a “very substantial” risk of “irreparable mis-identification.” Id. (internal quotation marks omitted). For preserved challenges like Fournier’s, we review the judge’s ruling de novo and his fact findings for clear error. See id. The judge in our case, however, made no fact findings on this issue.
Here, Fournier cannot get to first base, because we do not believe that Lespier identified him under suggestive conditions. Consider the context. The prosecutor and Lespier had been discussing Rodriguez, not Fournier, when the prosecutor asked him whether he could identify “that person” — meaning Rodriguez — and tell the jury “what that person” had on. And the prosecutor’s quite neutral “What about the person in the coat, do you recognize that person?” comment did not coach Lespier into fingering Fournier (whom Lespier knew anyway) as the person wearing the coat.
Undeterred, Fournier plays up how Les-pier could not identify him in court the day before — so, he argues, the prosecutor’s comment must have been suggestive. But even if we assume for argument’s sake that Lespier made the identification under suggestive circumstances, Fournier still cannot get the result he wants. Unlike us, the jurors here had front-row seats for Lespier’s in-court identification. Hearing his voice, seeing his eyes and facial expressions, and observing his body movements, they were perfectly positioned to spot any slight sign of uncertainty on his part when he did identify Fournier. Fournier’s attorney also had the chance to challenge the identification’s worth on cross-examination. And because jurors have a superior vantage point for sizing up the whole picture, the reliability of testimony like this is normally a matter for them. See, e.g., United States v. Jones, 689 F.3d 12, 18 (1st Cir.2012); United States v. Maguire, 918 F.2d 254, 264 (1st Cir.1990). Working with the totality-of-the-circumstances test, we have held time and again that “only in extraordinary cases” should identification evidence be kept from the jury. See Jones, 689 F.3d at 18 (quoting United States v. de Jesus-Rios, 990 F.2d 672, 677 (1st Cir.1993)); see also Espinal-Almeida, 699 F.3d at 602 (citing United States v. Rivera-Rivera, 555 F.3d 277, 282 (1st Cir.2009)). Nothing about this case screams “extraordinary,” however, so the judge’s identification ruling stands.
(3)
Brady Claims
As everyone knows, prosecutors must turn over to the defense exculpatory evidence that is material either to guilt or punishment. See, e.g., Brady v. *195Maryland, 373 U.S. 83, 87, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963). The duty to disclose covers impeachment evidence too, even if the evidence is not inherently exculpatory. See, e.g., Giglio v. United States, 405 U.S. 150, 154-55, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972). That is what the Fifth and Fourteenth Amendments’ Due Process Clauses require. See, e.g., Haley v. City of Boston, 657 F.3d 39, 47 (1st Cir.2011).
Broadly speaking, there are two types of Brady violations. See, e.g., United States v. González-González, 258 F.3d 16, 21-22 (1st Cir.2001). The first occurs when the undisclosed evidence shows that prosecutors knowingly used perjured testimony or allowed false testimony to go uncorrected. See, e.g., United States v. Agurs, 427 U.S. 97, 103-04, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976); Giglio, 405 U.S. at 153, 92 S.Ct. 763; see also González-González, 258 F.3d at 21 (citing Giglio too, among others cases). For this violation, undisclosed evidence is material “if there is any reasonable likelihood that the false testimony could have affected the judgment of the jury.” Agurs, 427 U.S. at 103, 96 S.Ct. 2392. The other type of Brady violation occurs when prosecutors suppress evidence favorable to the defense, even if the evidence does not involve false testimony. Id. at 104, 96 S.Ct. 2392; González-González, 258 F.3d at 22. For this violation, evidence is material “if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.” United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985). The “reasonable likelihood” and “reasonable probability” standards are synonymous. See GonzálezGonzález, 258 F.3d at 22.
Fournier makes an odd Brady/Giglio argument. Bear with us now as we run through it.
Testifying for the government on the eleventh day of trial, Ernesto David Vidró Díaz (“Vidró”) told the jury about how he had worked at Fournier’s eatery, El Me-són Bypass, from 2003 until 2004. During his time there, Vidró saw some Combo leaders — guys named “Potro, Bo, Ramon-cito, Burrito” — meet with Fournier at that locale two or three times a week. And on a couple of occasions he saw Burrito and Bo go into Fournier’s office and take out clear plastic baggies containing what looked liked powdered cocaine. The duo would then go to an out-of-the-way part of the restaurant and “deck” the cocaine— drug slang that means to prepare the drug for distribution.8 See United States v. García-Torres, 341 F.3d 61, 64 (1st Cir.2003). Vidró also saw Combo personnel score guns and cars there, thanks to Four-nier’s connections.
Vidró was no angel either, by the way. He had pled guilty to conspiring from 2007 through 2008 to sell drugs at the Pámpa-nos Housing Project (“Pámpanos”). Prosecutors had handed the defense FBI reports on interviews with him concerning the Pámpanos conspiracy, even though the reports involved a different drug organization and a different period than the ones involved here.
Answering questions posed by Fournier’s counsel on cross-examination, Vidró repeated that he had seen Burrito “decking” at El Mesón. Vidró also said that he had never seen Fournier “dealing with drugs,” though. “Were you aware if his *196son was dealing with drugs?” Fournier’s counsel then asked. Vidró replied that he knew nothing of the son’s drug-dealing ways until after he (Vidró) had left El Mesón Bypass’s employ in 2004. Sometime between then and when he “went to jail” for the Pámpanos drug conspiracy in 2008, Vidró explained, he learned that Fournier’s son “was dealing with us” as part of that conspiracy. Vidró also said that Fournier had owned “the drug point at Portugués” for a week.
On the next trial day, Fournier’s counsel blasted prosecutors at sidebar for not disclosing evidence favorable to the defense— namely, notes of local police officers who had interviewed Vidró as part of the Pám-panos-conspiracy investigation. The notes, counsel said, indicated that Vidró had mentioned Fournier’s son’s drug dealings with Potro, Bo, and Burrito — but not Fournier’s. A snippet from one of the undisclosed notes says: “George Fournier’s son used to give the drug” to these men, “who would cut the drug at the little room of the ‘El Mesón’ business.” So, counsel argued, Vidró did not testify truthfully when he described Fournier’s drug dealings with Potro, Bo, and Burrito. The judge rebuffed the defense’s attacks, explaining that the notes involved a different conspiracy and that Vidró testified that he had never seen Fournier deal with drugs.
Fournier presses the same Brady argument on appeal, but with a twist. He now says that prosecutors knew that Vidró had testified falsely. The government hints at the possibility that Fournier did not preserve the challenge for appeal — which if true would normally force us to review for plain error rather than abuse of discretion. See United States v. Prochilo, 629 F.3d 264, 268 (1st Cir.2011). But because his challenge is easily rejected on the merits, we skip over any waiver question. See United States v. Dávila-González, 595 F.3d 42, 49 n. 2 (1st Cir.2010) (taking that very tack).
The much-fought-over investigative notes — which again reflect that Vidró had accused Fournier’s son but not Four-nier of drug dealings with Combo leaders during the Pámpanos conspiracy sometime after 2004 but before 2008 — are neither exculpatory nor impeaching in this case. To recognize the obvious, Vidro’s statements to Pámpanos investigators touch on a different conspiracy at a different time, and his telling them that Fournier’s son gave drugs to persons who “decked” them at El Mesón Bypass in 2007-08 hardly lets Fournier off the hook for allowing them to “deck” drugs there in 2003-04. Also and importantly, the notes do not come within a country mile of contradicting Vidro’s testimony, which (we remind the reader) was that he had not seen Fournier “dealing with drugs.” So we do not see how the notes had a reasonable chance of changing the verdict here. And because Fournier has not shown that Vidró had perjured himself, that aspect of his Brady claim goes nowhere too.9
*197Fournier argues for the first time on appeal that we should analyze his disclosure claims under the Sixth Amendment’s Confrontation Clause. That argument is a nonstarter given Prochilo, where we noted that our judicial superiors have “thus far only evaluated [Brady] disclosure claims ... under the Due Process Clause of the Fifth and Fourteenth Amendments,” not the Sixth Amendment’s Confrontation Clause. See 629 F.3d at 271 (citing Bagley, 473 U.S. at 674-78, 105 S.Ct. 3375, and Pennsylvania v. Ritchie, 480 U.S. 39, 51-54, 107 S.Ct. 989, 94 L.Ed.2d 40 (1987)).
(4)
Sufficiency of the Evidence
Fournier insists that the evidence was too skimpy to support his conviction for conspiring to possess illegal drugs with intent to distribute. And that means, he argues (at least implicitly), that the evidence was too skimpy to support his conviction of aiding and abetting the possession of guns in furtherance of a drug crime — if he was not part of the drug conspiracy, the argument seems to go, then he did not commit a drug crime and so could not have aided and abetted the possession of weapons in furtherance of a drug crime. This sufficiency challenge— which we of course review de novo and in the light most favorable to the government, see Polanco, 634 F.3d at 45 — misses the mark.
As we said earlier, to prove a drug-conspiracy charge like the one here, the government must show beyond a reasonable doubt that (a) a conspiracy existed, (b) the defendant knew of it, and (c) he voluntarily participated in it, intending to commit the specified underlying crime. See Ortiz de Jesús, 230 F.3d at 5. Fournier thinks that prosecutors fell short of meeting requirements (b) and (c), because, he says, there is no evidence either that he had known about the “decking” at El Me-són Bypass or that he had bought and sold drugs for the conspiracy.
But recall Lespier’s testimony (sketched in a way most flattering to the prosecution): Fournier had attended Combo meetings, had given Combo personnel guns and cars to take on Combo rivals, and had taken part in a Combo shooting binge. And recall too Vidro’s testimony (also presented in a way most agreeable to the government): Fournier had met with Combo big shots at his sandwich shop, had given Combo cabalists guns and cars, and had given them a safe place (a secluded part of his restaurant) to “deck” drugs too. On top of that, Combo VIPs were not shy about flashing baggies of cocaine in his office, Vidró added. In the light cast by this evidence, a rational jury could conclude beyond a reasonable doubt that Fournier knew of and voluntarily joined the Combo conspiracy with an intent to further its goals. And as for his specific complaints: The jury was free to draw the commonsense inference that he knew about the “decking” happening right under his roof. Also, being a drug seller or buyer is not the only way to participate in a drug conspiracy. See, e.g., United States v. Aviles-Colón, 536 F.3d 1, 17-18 (1st Cir.2008); United States v. Portalla, 496 F.3d 23, 27 (1st Cir.2007). Doing an “ancillary” task — “accounting, communications, strong-arm, enforcement,” for example — can suffice, if done to further the conspiracy’s objective. Aviles-Colón, 536 F.3d at 15 (emphasis added) (quoting United States v. García-Torres, 280 F.3d *1981, 4 (1st Cir.2002)); see also Portalla, 496 F.3d at 27 (holding that the fact that the defendant himself did not peddle drugs, did not hold a “leadership position” within the conspiracy, “and as a provider of ‘peripheral’ services {viz., the provision of cell phones designed to elude law enforcement detection) was unaware of many details of the [conspirators’] drug business, would not foreclose a reasonable jury from convicting him” — given that one of the conspiracy’s goals “was the avoidance of police detection,” which he obviously helped out with). This describes Fournier’s situation to a “T.” And with his sufficiency-of-the-evidence attack on the drug-conspiracy count foiled, his passing attempt to undo his conviction for aiding and abetting the possession of a firearm in furtherance of a drug crime is a washout too.10
(5)
Courtroom Seating
Fournier says (through his pro se brief) that court security officers during the trial sat him and his codefendants 20 feet from counsel table and “cutoff all communication with their lawyers” — actions, he adds, that infracted the Sixth Amendment. See, e.g., United States v. Rodríguez-Durán, 507 F.3d 749, 776 (1st Cir.2007) (discussing a defendant’s Sixth-Amendment right to consult with counsel at trial). His lawyer never complained about this below, and the record discloses no whisper as to what actually happened. Consequently, we must dismiss these claims without prejudice to his litigating them (if he desires) in a 28 U.S.C. § 2255 proceeding. See, e.g., United States v. Bucci, 525 F.3d 116, 129 (1st Cir.2008). Obviously, nothing we have said or left unsaid should be considered as even a remote suggestion of how a petition like that might fare.
Fournier also says (again, through his pro se brief) that having the defendants sit together during the trial deprived him of the presumption of innocence. Plain-error review applies, because his lawyer made no mention of this below. See, e.g., Torres-Rosario, 658 F.3d at 116. That is probably because the argument is a clear loser under United States v. Turkette, 656 F.2d 5 (1st Cir.1981). There, we rejected the suggestion that a judge had offended the Constitution by sitting the defendants alone in the first row of the courtroom’s spectators’ section — a borderline “frivolous” claim, we said, because
the defendants had to be seated somewhere, and from the start of the trial the jury knew that all of them ... were charged with conspiracy. If the jury received a suggestion of guilt by association, and their verdict belies this, it was the result of the conspiracy charge, not the seating arrangement.
Id. at 10. Ditto here. Consequently, we see no error, plain or otherwise.
(6)
Closing Argument
Finally, Fournier alleges (also through his pro se brief) several problems with the prosecution’s closing argument. Because he did not object below, he must show plain error. See, e.g., United States v. Kasenge, 660 F.3d 537, 541 (1st Cir.2011); United States v. Kinsella, 622 F.3d 75, 84 (1st Cir.2010). But even then, reversal is justified only if the prosecution’s *199comments “so poisoned the well that the trial’s outcome was likely affected.” Kasenge, 660 F.3d at 542 (quoting United States v. Henderson, 320 F.3d 92, 107 (1st Cir.2003)).
Undaunted, Fournier condemns the prosecutor for saying that he had armed Combo members with “rifles,” had owned the “Portugués” drug point for a “week,” and had provided “narcotics” too. None of this crossed any line, however, because the record supports what the prosecutor said: Lespier had explained how Fournier had given Combo personnel “AK-47 rifles” and had owned the “Po-rtugués” drug point for a week, and prosecutors could argue the Fournier-sold-nar-cotics inference based on his having owned that drug point. See, e.g., Manor, 633 F.3d at 18 (concluding that a defendant’s prosecutorial-misconduct claim flopped because the “prosecutor had enough eviden-tiary support” to make the complained-of statement); United States v. Martínez-Medina, 279 F.3d 105, 119 (1st Cir.2002) (finding nothing objectionable with statements that “appear reasonably supported by the record or are within the prerogative of the prosecution to characterize the evidence presented at trial and argue certain inferences to the jury”).
Next Fournier excoriates the prosecutor for saying that an early Combo member called “Yiyito” (whom we have not yet mentioned) had worked at El Mesón Bypass in December 2005, when in fact he had not. Another Combo member, Burrito (whom we have introduced already), had worked here during that period, the evidence shows. Perhaps the prosecutor confused the two. But given how both Yiyito and Burrito were Combo members, how the slight misstatement touched on an incidental matter (i.e., whether a Combo member besides Fournier had worked at El Mesón Bypass), how the judge had cautioned jurors that counsel’s comments were not evidence and that only their recollection of the facts counted, and how there was enough evidence of Fournier’s guilt, we easily conclude that the complained-of remark did not prejudice Four-nier’s right to a fair trial. See, e.g., United States v. Page, 521 F.3d 101, 107 (1st Cir.2008); United States v. Carrasquillo-Plaza, 873 F.2d 10, 14 (1st Cir.1989).
Castillo
(1)
Confession
Castillo’s lead argument is that the judge botched the case by not suppressing his (allegedly) involuntary confession. But the evidence presented at the suppression hearing — which we are about to discuss— fairly supports the judge’s conclusion that his confession was voluntary.
Convening a hearing mid-trial following Castillo’s oral motion to suppress, the judge heard from one witness: Enrique Rodríguez, a Puerto Rico Police officer, who explained what went down on the heels of Castillo’s arrest. Agents at a local police office gave Castillo an oral Miranda warning sometime before noon while they were processing him.11 Castillo said that *200he understood his rights. Later, officers gave him a form advising him of his Miranda rights, and he signed it around 2 p.m. Later still, while the police were driving him to federal jail, Officer Rodriguez asked a nervous-looking Castillo if “something was wrong.” Responding, Castillo confessed that he had worked at a drug point in the Dr. Pila Public Housing Project.
After the hearing, the judge found that Castillo was twice Mirandized before saying anything. Also, the judge found that the police had not used any coercive tactics that would spoil the voluntariness of his confession. And so the judge denied Castillo’s motion to suppress.
For obvious reasons, Castillo spends considerable energy trying to convince us to reverse that ruling. We review de novo the judge’s legal conclusion about the voluntariness of the confession. See, e.g., United States v. Boskic, 545 F.3d 69, 77 (1st Cir.2008). But we must accept the judge’s factfinding on the circumstances surrounding the confession unless clearly erroneous, see id., because determining credibility, weighing the evidence, and drawing inferences from the evidence all fall within his province, see, e.g., United States v. Valle, 72 F.3d 210, 213-14 (1st Cir.1995).
Castillo argues that officers coerced him to confess in the car by not giving him food at any point before he spoke. There is simply no record support for that claim, however. True, Officer Rodriguez could “not remember” if anyone had “offered” Castillo food “or not.” But he stressed that “it is our regular” practice “to offer” persons in Castillo’s shoes something to eat. And he quickly added that Castillo never asked him for food, telling defense counsel that “the only thing that we talked about was what he admitted [to] me and nothing else.” “He never complained” about “anything,” Officer Rodriguez said, and he looked “normal” too. The judge was entitled to believe the officer’s testimony, obviously. Given all this, we cannot brand the judge’s no-coercion factfinding clearly erroneous.
Surprisingly, Castillo contends that the government did not prove when the agents Mirandized him. And, he adds, the evidence “contradicts” the government’s theory that he had been Mirandized before the confession. Surprisingly is the right word, because the record&emdash;Officer Rodriguez’s testimony (which the judge could credit without committing clear error) and Castillo’s signed Miranda-rights-waiver form&emdash; shows that officers had twice Mirandized him before he opened up to Officer Rodriguez. Again, we see no clear error.
Castillo, lastly, grumbles that he was arrested in the morning, without anyone taking him to a magistrate judge&emdash;a one-sentence-throwaway line appearing in his brief, supported by no analysis explaining why any of that should matter for purposes of assessing the voluntariness of his confession. Whatever point he is trying to make is waived. See, e.g., Zannino, 895 F.2d at 17 (deeming waived claims that lack coherence or developed argument).
The bottom line is that the judge did not slip in admitting Castillo’s confession.
(2)
Sufficiency of the Evidence
Castillo makes a sufficiency-of the-evidence claim, arguing that the government did not prove that he both had knowledge of the drug conspiracy and had acted with an intent to further its objectives. But his admission that he had sold *201drugs at what the evidence showed was a Combo-controlled drug point cuts the legs out from under this argument. Enough said.
(3)
Drug Quantity
At sentencing the judge calculated Castillo’s sentence on the drug conspiracy count this way. Using the sentencing guidelines’ 2009 edition, the judge noted that Castillo was responsible for (among other things) 1 kilogram or more of heroin. That amount — which activated a statutory range of 10 years to life, see 21 U.S.C. § 841(b)(1)(A) — resulted in a base-offense level of 32. After adding 2 levels for conduct within 1,000 feet of a public-housing facility, the guidelines level totaled 34, which, when cross-referenced against Castillo’s criminal-history category of I, produced a guidelines range of 151 to 188 months. But after accounting for the § 3553(a) factors, the judge imposed a 120-month prison sentence — a sentence 31 months below the low end of the guidelines.
An unhappy Castillo now accuses the judge of presiding over a process that flunked Alleyne, hitting him with the full weight of the conspiracy’s drug doings instead of relying on individualized drug amounts determined beyond a reasonable doubt. Because he débuts this claim here, our review is for plain error.12 But no plain error rears its ugly head.
As we observed in discussing Acosta’s sentencing challenge, the jury made individualized beyond-a-reasonable-doubt drug findings for each defendant, concluding when they got to Castillo that he had conspired to possess and distribute 1 kilogram or more of heroin, less than 500 grams of cocaine, less than 5 grams of crack, and less than 100 kilograms of marijuana. And the jury’s particularized beyond-a-reasonable-doubt findings sap the strength from Castillo’s Alleyne argument.
Hold on, says Castillo, the jury did not have enough evidence to make that finding. This is an unpreserved sufficiency argument reminiscent of Acosta’s. And like Acosta’s it is a deadbang loser, for the same reason. He says not a word about how — after considering the evidence and inferences in the light most sympathetic to the government — he can clear the very high elear-and-gross-injustice hurdle required for reversal, see Pratt, 568 F.3d at 18, which does not cut it, see Town of Norwood, 202 F.3d at 405.
Still hoping against hope for a reversal, Castillo argues that the judge bungled the instructions by not focusing the jury’s attention on whether he and his codefen-dants had “knowingly” conspired to possess and distribute the drug types and quantities alleged in the indictment. Hardly. Getting down to the nitty-gritty, we see that the judge told the jury that it could convict a defendant on the drug-conspiracy count only if it found “beyond a reasonable doubt” that the “defendant knowingly and willfully” conspired to possess and distribute the drug types and quantities “as charged” in “the indictment.” Keeping in mind too the special verdict forms that required the jury to make personalized drug-type-and-quantity findings for each of the five defendants *202beyond a reasonable doubt, it is readily evident that Castillo’s theory is a bad one.
Rodriguez
(1)
Variance
Rodriguez claims that a fatal variance existed between the drug conspiracy alleged in the indictment and the drug conspiracy proved at trial. To get his conviction reversed on this ground, Rodriguez must show two things: first, that a variance occurred&emdash;ie., that the facts proved at trial differed materially from those alleged in the indictment; and second, that the variance was prejudicial&emdash;ie., that it affected his substantial rights. See, e.g., United States v. Seng Tan, 674 F.3d 103, 110 (1st Cir.2012). Critically, there is no prejudicial variance if the indictment gave him enough details to prepare a defense and plead double jeopardy to prevent another prosecution for the same crime. See id.
Our review of this claim is similar to a sufficiency-of-the-evidenee challenge and is assessed de novo, if properly preserved. See, e.g., United States v. Dunbar, 553 F.3d 48, 61 (1st Cir.2009). The government floats the idea that maybe Rodriguez did not say enough in the district court to preserve the argument for appeal. Maybe, but because his challenge is a dud, we skirt the waiver question.
The gist of his theory is that the indictment “exaggerated [his] participation” in the drug conspiracy, linking him to three drug points while the trial evidence only linked him to one, the Coto Laurel drug point mentioned earlier in this opinion. Going by the “strict definition” of the word, one could argue “that proving fewer than all of the facts in an indictment&emdash;but adding nothing new&emdash;is not a variance at all.” United States v. Mueffelman, 470 F.3d 33, 38 n. 6 (1st Cir.2006). Yet “omissions could so seriously distort the picture presented ... as to raise questions of unfair prejudice,” making a variance analysis proper. Id. To move the analysis along, we will assume without deciding that Rodríguez can show a variance. Still, he must show prejudice sufficient to justify reversal. This he has not done. Things might be different if the indictment had not mentioned the Coto Laurel drug point or had “masked” its importance by putting “undue emphasis” on the other drug points. See id. at 39 (discussing the masking phenomenon). But the indictment did mention the Coto Laurel drug point specifically, explicitly, and repeatedly. And Rodriguez’s counsel did try to discredit the charges at trial where she could&emdash;challenging Serrano, for example, regarding his knowledge of the history of the Coto Laurel drug point (its ownership, how long it had been up and running, etc.).
The short of it is that the indictment gave Rodriguez fair warning that the government viewed his Coto-Laurel-drug-point ties as an important part of the overall conspiratorial scheme. And because he cannot credibly claim surprise, the alleged variance does not merit reversal. See id.; see also United States v. Rodriguez, 525 F.3d 85, 103 (1st Cir.2008) (concluding that even if the evidence varied from the charges, the government’s proving that the defendant “ran one of several drug points” instead of “the entire drug organization” caused no prejudice, because the indictment gave him sufficient knowledge of the charges against him so that he could craft a defense).13
*203(2)
Sufficiency of the Evidence
Rodríguez separately argues that insufficient evidence supports his conviction on the gun count. To prove guilt here, the government had to prove beyond a reasonable doubt, first, that Rodriguez knew “to a practical certainty” that someone would use or carry a firearm during and in relation to a drug-trafficking offense or possess a firearm in furtherance of one and, second, that he willingly took some action to facilitate the firearm’s use, carrying, or possession. See, e.g., United States v. AlverioMeléndez, 640 F.3d 412, 420 (1st Cir.2011); United States v. Negrón-Narváez, 403 F.3d 33, 37-38 (1st Cir.2005); see also United States v. Medina-Román, 376 F.3d 1, 6 (1st Cir.2004) (clarifying that knowledge meeting the “practical certainty” test will in many cases arise “from such an intimate involvement in the enterprise that the requirement for an affirmative action to facilitate the crime inevitably will be met”).14 At bottom, Rodriguez really trains his sights on one issue: Did prosecutors prove his facilitation? Perusing the evidence in a way most amiable to the government, see Polanco, 634 F.3d at 45, we answer yes.
Earlier we discussed how Lespier testified that Rodriguez ran the Combo-controlled Coto Laurel drug point. What we did not mention then is that Lespier also testified that Combo leaders like Rodriguez used armed “enforcers” all the time. Extraordinarily violent men (no doubt), enforcers protected Combo higher-ups and beat down rival drug gangs&emdash;kill-or-be-killed work, with guns among the tools of their trade. See, e.g., United States v. Correy, 570 F.3d 373, 400 (1st Cir.2009). Being in charge at the Coto Laurel drug point meant that Rodriguez had a hand in personnel decisions, bossing the help around with Jife-and-death consequences&emdash; or so a jury could find. For example, one enforcer, a person called “Gordo,” lost his life defending Rodriguez, Lespier said. Lespier also talked about seeing Rodriguez’s main enforcer, Miguel Pacheco, at the Coto Laurel drug point. And like other enforcers, Pacheco was packing heat. Considered in the proper light, the evidence and reasonable inferences supply enough proof of Rodriguez’s facilitation. See Badamo v. United States, 201 F.3d 426, 1999 WL 1338076, at *1, 2 (1st Cir.1999) (per curiam) (unpublished table decision) (finding adequate evidence of facili*204tation where (a) the defendant was responsible for having another help out with a drug-trafficking crime by posing as a police officer during a proposed robbery of a “stash house,” and (b) a jury could reasonably infer that a poser would have a gun on him, just like a real officer would).15
(3)
Drug Quantity and Role in the Offense
This brings us to Rodriguez’s challenges to his 240-month sentence for his conviction on the drug-conspiracy count. To understand his claims, some background is helpful.
The jury found beyond a reasonable doubt specific drug amounts attributable to Rodriguez — 1 kilogram or more of heroin, 5 kilograms or more of cocaine, 50 grams or more of crack, and less than 100 kilograms of marijuana. The first finding alone meant that he faced a statutory minimum sentence of 10 years and a maximum of life. See 21 U.S.C. § 841(b)(1)(A). With the statutory range set, the judge had to make defendant-specific drug findings by a preponderance of the evidence, see United States v. Rodríguez, 731 F.3d 20, 31 (1st Cir.2013), which is a more-likely-than-not standard, see United States v. Vixamar, 679 F.3d 22, 29 (1st Cir.2012). And using the 2009 iteration of the guidelines, the judge found that Rodriguez had been a leader of the Coto Laurel drug point and a leader in the Combo. From this finding the judge concluded that Rodriguez could have reasonably anticipated the amounts of drugs being distributed by the conspiracy each year — i.e., at least 182.5 kilograms of heroin, 182 kilograms of cocaine, 136.6 kilograms of crack, and 136 kilograms of marijuana, which is the equivalent of over 30,000 kilograms of marijuana, the judge noted.16 That led to an offense level of 38, the judge said, which he enhanced 2 levels for conduct within 1,000 feet of a public-housing facility and 3 levels for Rodriguez’s role as a manager or supervisor in the conspiracy. The result was an offense level of 43, which combined with his criminal-history category of I to give him a guidelines range of life in prison. The judge then considered the § 3553(a) factors before settling on the 240-month term.
Rodriguez essentially thinks that the judge stumbled twice — first by attributing conspiracy-wide drug quantities to him and then by imposing a 3-level sentencing enhancement on him for his supposed role as a manager or supervisor in the conspiracy. Spoiler alert — we see no error.
Absent individualized findings concerning the “amounts attributable to, or foreseeable by, that defendant,” the quantities ascribed “to the conspiracy as a whole cannot automatically be shifted to the defendant.” United States v. Colón-Solis, 354 F.3d 101, 103 (1st Cir.2004). Rodriguez concedes that the judge made particularized findings. He just believes that those findings are insupportable. Our review is for clear error, see, e.g., United States v. Correa-Alicea, 585 F.3d 484, 489 (1st Cir.2009), knowing that a finding is not clearly wrong simply because it strikes us as probably wrong — rather, “it must prompt ‘a strong, unyielding belief, based on the whole of the record,’ that the judge made a mistake,” Toye v. O’Donnell (In re O’Donnell), 728 F.3d 41, 45 (1st Cir.2013) (quoting Islamic Inv. Co. of the Gulf *205(Bah.) Ltd. v. Harper (In re Grand Jury Investigation), 545 F.3d 21, 24 (1st Cir.2008)); accord United States v. Cintrón-Echautegui, 604 F.3d 1, 6 (1st Cir.2010).
Plenty of evidence supports the judge’s finding that Combo-leader Rodriguez could have reasonably foreseen the amounts of drugs embraced by the conspiracy: Onetime-Combo leader Lespier testified that Rodriguez belonged to the Combo and ran the Coto Laurel drug point. Confidential-informer Serrano testified that Rodriguez got the drugs for that drug point from his Combo counterparts at the Dr. Pila Housing Project. The two were “united,” to use Serrano’s word. Another cooperating codefendant, Elvin Cart-agena Colón (“Cartagena”), whom we have not yet mentioned, saw Rodriguez hand a Combo leader a “tall bundle” of cash for drugs. Still another cooperating witness, Ediberto García Román (“García”), whose name we have not yet brought up, said Rodriguez met with Combo confederates to coordinate the drug supply for the Coto Laurel drug point. So Rodriguez did not just know Combo leaders. He was, as the judge found, a Combo leader too who did conspiracy-linked drug business with Combo members. And all of this was enough to satisfy the preponderance standard, despite what Rodriguez says.
As for his challenge to the manager-or-supervisor-in-the-eonspiracy enhancement, see U.S.S.G. § 3Bl.l(b), Rodriguez again must show clear error, see United States v. Garcia-Hernandez, 659 F.3d 108, 114 (1st Cir.2011). And again he cannot.
An enhancement under this guideline proceeds in a sequence of two steps. Id. First the sentencing judge must find that the underlying offense involved five or more persons (including the defendant) or was otherwise extensive. Id.; United States v. Conley, 156 F.3d 78, 85 (1st Cir.1998). Then the judge “must find that the defendant managed or supervised one or more of the other participants in that activity.” Garcia-Hernandez, 659 F.3d at 114. The preponderance-of-the-evidence standard of proof applies, not surprisingly. See, e.g., United States v. Alicea, 205 F.3d 480, 485 (1st Cir.2000).
Rodriguez does not contest the judge’s finding that the conspiracy involved at least five persons&emdash;indeed, his counsel specifically waived that argument at oral argument, reserving his objection for the judge’s conclusion that he played a managerial or supervisory role in the scheme. But gobs of testimony from multiple witnesses support the judge’s finding. Just from Lespier’s and Serrano’s testimony one can reasonably infer that Rodriguez&emdash;at a minimum&emdash;controlled the criminal activity of (a) two enforcers&emdash;one of whom, Gordo, died protecting him; (b) three drug sellers&emdash;Wanda, Pucho, and Acosta; and (c) one drug runner&emdash;Acosta, who took the bundle of drugs Rodriguez gave him and passed them on to the sellers. That will do. See, e.g., United States v. Cruz-Rodríguez, 541 F.3d 19, 33 & n. 12 (1st Cir.2008) (discussing what it means to manage or supervise a person). Desperately, Rodriguez essentially argues that neither Lespier nor Serrano were credible. But the judge had a better vantage to assess the credibility of all involved. And we see no reason to reverse. See generally United States v. Platte, 577 F.3d 387, 392-93 (1st Cir.2009) (noting that “credibility determinations are part of the sentencing court’s basic armamentarium”). The enhancement was justified.
Guzmán
(1)
Car Search
Guzman berates the judge for not suppressing the fruits from the search of a car *206(most notably an incredible amount of heroin). He is convinced that the traffic stop that led to the search was a pretext to allow the police to rummage inside the auto without a warrant, violating his reasonable expectation of privacy. And he is also convinced that the judge should have held an evidentiary hearing on the issue. The story behind all this is easily told.
Guzmán had rented a Chevy Malibu but had let three others (coconspirators, it turns out) use it on the day of the search. The police arrested the trio after pulling the car over on a highway for a traffic violation and spotting what appeared to be heroin in the auto’s “interior” (exactly where the record does not say). Guzmán was not in the car. According to an officer at the scene, his department’s policy is that once law enforcers take custody of a car they must itemize its contents in front of the arrestees. And an inventory search of this auto turned up over 5,000 baggies of heroin in the car’s trunk, along with other drugrelated paraphernalia.
Calling the search and seizure “illegal,” Guzmán argued below that none of this evidence could be used against him. He also insisted that his car-renter status meant that he had a reasonable expectation of privacy at stake in the search. He offered no affidavit or equivalent evidentia-ry material to support his suppression theory, however. And the judge denied the suppression motion, without holding a hearing or issuing a written decision. Guzmán moved the judge to reconsider, but again failed to provide an affidavit. The judge denied that motion too. Six months later, Guzmán moved once more for reconsideration, this time supplying an affidavit swearing that he had loaned the car to his three “friends” and had told them not to open the glove compartment or the trunk and that “[u]pon information and belief no traffic violation” had occurred while his friends were in the rented auto. Motion denied, the judge ruled.
In this venue, Guzmán still champions (as we just said) his theories that the stop was pretextual and that the search compromised his reasonable expectation of privacy. Nothing that he says on this score matters, however, and our de novo review, see United States v. De Jesús-Viera, 655 F.3d 52, 58 n. 3 (1st Cir.2011), leads straight to affirmance. The reason why is clear.
Having arrested Guzmán’s “friends” after seeing a baggie of suspected heroin in the car’s interior, officers could search the auto under two traditional exceptions to the warrant requirement: the-search-incident-to-arrest exception and the inventory-search exception. The first exception permits a warrantless search if “it is reasonable to believe” that the car has “evidence of the offense of arrest.” Arizona v. Gant, 556 U.S. 332, 351, 129 S.Ct. 1710, 173 L.Ed.2d 485 (2009); see also Polanco, 634 F.3d at 42 (stressing how “Gant clarified” that an auto search may come within this exception “only in two very specific situations: when the ar-restee is unsecured and within reaching distance of the passenger compartment at the time of the search (the officer-safety exception), or when it is reasonable to believe evidence relevant to the crime of the arrest might be found” in the auto) (internal quotation marks omitted). During oral argument before us Guzmán’s lawyer tried to parry this point by challenging the lawfulness of the trio’s arrest. But he has no standing to do so, given that he himself was not one of the arrestees. See generally Alderman v. United States, 394 U.S. 165, 174, 89 S.Ct. 961, 22 L.Ed.2d 176 (1969) (noting that “Fourth Amendment rights are personal rights which, like some other constitutional rights, may not be vicariously asserted”). Indeed, we asked *207Guzmán’s lawyer for authority to back up her surmise. She had none. And our research turned up none.
The other exception holds that if the arrests are legal, then the police can take the car back to the barrack and search it pursuant to standard inventory procedures&emdash;provided also that they do not “act[] in bad faith or for the sole purpose of investigation.” Colorado v. Bertie, 479 U.S. 367, 371, 372, 107 S.Ct. 738, 93 L.Ed.2d 739 (1987). Once again Guzmán’s lawyer wrangled with us over the legality of his friends’ arrests, without citing any pertinent authority. And once again Guzmán’s argument falls apart on standing grounds, given that he himself was not among those arrested. So he cannot challenge the propriety of the inventory search. See Alderman, 394 U.S. at 174, 89 S.Ct. 961.
But wait, says Guzmán, surely the police did not have to take the car&emdash;they could have called him to come get it instead. This line of argument is a dead end, however, because “the existence of alternative means of dealing with the automobile, even less intrusive means, does not illegitimate the constables’ decision” to remove it. United States v. Rodriguez-Morales, 929 F.2d 780, 786 (1st Cir.1991).
As for his faulting the judge for not convening an evidentiary hearing&emdash;an issue we review for abuse of discretion, see, e.g., United States v. Allen, 573 F.3d 42, 50-51 (1st Cir.2009)-we note that Guz-mán had to satisfy an entry-level burden of showing that the police’s warrantless search did not come within “any” recognized warrant exception, id. at 51 (quoting United States v. Calderon, 77 F.3d 6, 9 (1st Cir.1996)), including the two we just mentioned. And this he did not do, for he alleged nothing “definite, specific, detailed, and nonconjectural” that could defeat application of either exception here. See id. (quoting Calderon, 77 F.3d at 9). No surprise, then, that we see no abuse of discretion.
With that said, we turn to his next challenge.
(2)
Sufficiency of the Evidence
Guzmán asks us to declare the evidence inadequate to support his gun-count conviction. Basically, his gripe is that nothing shows that he had taken some step to help another use, carry, or possess a firearm. The evidence was hardly overwhelming, as the government concedes. But viewed in the proper light, see Polanco, 634 F.3d at 45, we believe that there was enough to sink Guzmán’s argument.
Here is why. The jury heard from cooperating-conspirator Roberto Pizarro Orta (“Pizarro”) about the time Combo-leader Guzmán took his brother Alberto to the house of a man named “Willie.” Willie was supposed to give the brothers Guzmán the cash from that week’s heroin sales. Like his brother, Alberto was a Combo member. And it was no secret that Alberto had a gun on him. Pizarro “saw” him with a “.38 revolver,” actually. Anyway, a suspicious car was in the area. And Guzmán and the others asked Pizarro to check it out. “I told them I was not going to check any car, empty”&emdash;and by “empty” he meant “unarmed.” So Pizarro asked for and got Alberto’s pistol.17 Thin this evidence may be. But reading everything together with all reasonable inferences in the light most *208favorable to the government, we think there was enough for a levelheaded jury to find that Guzmán had his pistol-packing brother tag along to make sure that he got his hands on the drug money without incident — particularly when one remembers the other evidence showing how Combo leaders used armed enforcers to help push their dirty business to the max. And so there was sufficient evidence of Guzmán’s facilitation. See Badamo, 1999 WL 1338076, at *2.
(3)
Jury Instructions
Guzmán next contends that the judge misinstructed the jury on the gun count. But because he never objected below, Guz-mán must run the gauntlet of plain-error review. See, e.g., United States v. Griffin, 524 F.3d 71, 76 (1st Cir.2008). That requires him to show an error that was “obvious and clear” and that affected his substantial rights. Id. Even then, we need not reverse unless the error also seriously undermined the fairness, integrity, or public reputation of judicial proceedings. See Kinsella, 622 F.3d at 83.
Moving to the particulars of the present case, we see that the judge started off talking about the elements of the drug-conspiracy count, instructing the jury on the law of conspiracy, for example. Then he charged the jury on the elements of the gun count, touching on how the law prohibits anyone from aiding and abetting the use or carrying of a firearm during and in relation to a drug crime or possessing a firearm in furtherance of one. Because the drug crime here was conspiracy to possess and distribute illegal drugs, the judge talked about the law of conspiracy some more. And then the judge explained the meaning of aiding and abetting.
For his first argument, Guzmán says that the judge never told the jury that it could consider “whether defendant aided and abetted a member in using or carrying firearms in relation” to a drug crime only if it first determined that defendants had conspired to commit a drug crime. Nonsense. The judge explicitly charged the jury that to find Guzmán guilty on the gun count it would “first” have to find that “the defendants committed the crime of conspiring to possess with intent to distribute” the illegal drugs listed in the indictment. Hardly the stuff of plain error, by any stretch of the imagination.
Guzmán also calls out the judge for not telling the jury that to find “Guzmán” guilty on the gun count it had to find that he had “performed some affirmative act, had some stake or power over any firearm.” Reduced to essentials, what the judge told the jury was aiding and abetting “means intentionally to help someone else commit the crime charged.” And, he added, a defendant is guilty of aiding and abetting if “someone else committed the charged crime” (i.e., using or carrying a firearm during and in relation to a drug crime or possessing a firearm in furtherance of one) and the “defendant[] consciously shared the other person’s knowledge of the charged crime,” “intended to help” him, and “took part in the endeavor seeking to mak[e] it succeed.” There is no etched-in-stone way to convey the aiding- and-abetting idea to the jury, and what is necessary might turn on the circumstances of the case. See United States v. Urciuoli, 513 F.3d 290, 300 (1st Cir.2008); see also United States v. Gonzalez, 570 F.3d 16, 29 (1st Cir.2009) (illustrating that point by noting that “we have explicitly declined to require the ‘shared’ intent language found in some of our opinions”). But the judge’s discussion essentially mirrors the ones found in our cases. See, e.g., United States v. Rodríguez-Lozada, 558 F.3d 29, 41 (1st Cir.2009); United States v. Garcia-*209Carrasquillo, 483 F.3d 124, 130 (1st Cir.2007); United States v. Rosario-Diaz, 202 F.3d 54, 62-63 (1st Cir.2000). And Guz-mán has not shown why the particulars of his case required the judge to phrase the aiding-and-abetting concept any differently. So when you get right down to it, what the judge said does not come within shouting distance of plain error.
As a last-ditch effort, Guzmán lodges a double complaint against the judge’s conspiracy charge. For openers, he calls them unduly repetitive, noting that the judge discussed general conspiracy principles three times. He then sounds off about the charge’s placement, groaning that the second set of conspiracy instructions were read too close to the aiding and abetting charge. And all of this, he says, confused jurors into thinking that they could convict him of conspiring to possess firearms, rather than of aiding and abetting a firearm’s use or carry during and in relation to, or possession in furtherance of, a drug crime. We think not.
Notably for present purposes, Guzmán’s counsel candidly conceded at oral argument that the judge did not misstate the law of conspiracy. And having read the entire jury charge with care, see Griffin, 524 F.3d at 76 (telling us that this is what we must do), we cannot buy the idea that the repetition must have — by its sheer weight — coaxed the jury to find against Guzmán on the gun count. Nor did we detect any defect with the instructions’ placement that might have befuddled the jury in deciding his fate on that count. Nothing remotely resembling plain error here, to put it mildly.
(4)
Napue Violations
Convinced that prosecutors violated the constitutional rule against knowingly using perjured testimony, see Napue v. Illinois, 360 U.S. 264, 269, 79 S.Ct. 1173, 3 L.Ed.2d 1217 (1959), and that there is a “reasonable likelihood” that that testimony affected the jury’s judgment, see United States v. Mangual-Garcia, 505 F.3d 1, 10 (1st Cir.2007), Guzmán blisters the judge for not granting him a new trial — an issue that usually invites abuse-of-discretion review, see González-González, 258 F.3d at 20. The Napue-violation point that he pushes here pivots on his belief that prosecutors let some cooperating witnesses lie on the stand about the benefits that they would get after testifying under plea and cooperation agreements. What makes this argument a no-go is that it is not the one that he had presented below.
Guzmán’s new-trial motion accused only cooperating witness Garcia of telling whoppers about potential benefits. A former heroin seller at the Dr. Pila Housing Project, Garcia testified that he would give money from his sales to “Willie,” who would in turn give it to Guzmán. García explained that he had confessed to the police that he had participated in the Combo-drug conspiracy, a crime with which he was never charged. He also said that he was cooperating with prosecutors because he wanted to turn his life around. And he denied any agreement for prosecutorial leniency. Guzmán’s new-trial motion speculated that Garcia must have lied because it was “highly unlikely” that anyone would cooperate without an agreement. The judge found no Napue violation. But Guz-mán has now shifted the focus of his attack from Garcia to “Lespier” and others “whose testimony, produced under a plea and cooperation agreement,” suggested that prosecutors had offered “ ‘no benefits’ ” to secure their help. Having switched tactics this way so late in the game, Guzmán has waived the argument that he now seeks to pursue. See United States v. Slade, 980 F.2d 27, 31 (1st Cir.*2101992) (emphasizing that “a party is not at liberty to articulate specific arguments for the first time on appeal simply because the general issue was before the district court”); United States v. Dietz, 950 F.2d 50, 55 (1st Cir.1991) (holding that a criminal defendant, unhappy with a judge’s ruling “yet persuaded that his original arguments lacked merit, cannot switch horses mid-stream in hopes of locating a swifter steed”); see also United States v. Charles, 213 F.3d 10, 21 (1st Cir.2000) (echoing points made by Slade and Dietz).
One last issue, and we are done.
(5)
Statutory Sentencing Enhancement
Because of two prior drug felonies, Guzman got an automatic life sentence for his conviction on the drug-conspiracy count. See 21 U.S.C. § 841(b)(1)(A) (enhancing the mandatory sentence for a controlled-substance offense to life imprisonment if the defendant has two or more final convictions for a felony-drug offense at the time he committed the crime for which he is being sentenced). Not willing to take that sentence without a fight, Guzmán argues (as he did below) that the two priors — resulting from a pair signed plea agreements between him and the government — were not separate criminal episodes, as § 841(b)(1)(A) has been read to require. See, e.g., United States v. De Jesus Mateo, 373 F.3d 70, 74 (1st Cir.2004); Martinez-Medina, 279 F.3d at 123. In reviewing the judge’s decision, we decide de novo any underlying legal issues, see United States v. Rivera-Rodríguez, 617 F.3d 581, 608 (1st Cir.2010), but check any fact findings for clear error, see United States v. Fink, 499 F.3d 81, 88 (1st Cir.2007).
At sentencing the judge read each plea agreement’s statement of facts into the record. The first one said:
On or about 1994 to 1996 and October 1997 the defendant Danny Guzman Cor-rea was the supplier of heroin for the drug point located at the Portugués and Bélgica Wards in Ponce, owned by co-defendant Orlando Rosa Rodriguez.
The second one said:
On or about 1996 and October 1997 the defendant Danny Guzman Correa was one of the suppliers of heroin, cocaine and marihuana to co-defendants Orlando Maldonado Orengo, Samuel Arce Leon and Ignacio Mendoza Diaz who would later provide the same to co-defendant Ramona Nieves Santiago for further distribution at her drug point located at the Rosaly Residential in Ponce, Puerto Rico.
The judge deemed the two separate criminal incidents. We see no error. Sure, the pair overlapped a bit in time. But the second involved different coconspirators, a different locale, and a different mix of drugs than the first. As we have said, time and time again, “an ongoing course of criminal conduct such as narcotics trafficking may involve many such criminal episodes, each a discrete occurrence.” United States v. Lino, 493 F.3d 41, 43 (1st Cir.2007) (quoting De Jesus Mateo, 373 F.3d at 74, in turn quoting Martinez-Medina, 279 F.3d at 123). Plus, “[t]he fact that all are related, part of a series, or part of a continuous course of criminal dealing, does not necessarily render them a single criminal episode.” Id. (again quoting De Jesus Mateo, 373 F.3d at 74, in turn quoting Martinez-Medina, 279 F.3d at 123, again). A contrary holding, we have stressed, “would insulate the very career criminals the statute is designed to reach — those continuously engaged in criminal conduct.” Martínez-Medina, 279 *211F.3d at 123 (quoting United States v. Maxey, 989 F.2d 303, 307 (9th Cir.1993)). So it is here. Consequently, the judge did not err in treating these convictions as distinct.
In a last-gasp bid to undo his life sentence, Guzman says that certain drug-quantity calculations in the presentence report “are far from reliable.” But the jury’s decision to hold him responsible beyond a reasonable doubt for at least 1 kilogram of heroin was enough to trigger the life sentence, when combined with his two priors, of course. And that means that we need not take up this issue. See, e.g., United States v. Rivera-Ruiz, 244 F.3d 263, 272-73 (1st Cir.2001) (relying on United States v. Tavano, 12 F.3d 301, 307 (1st Cir.1993)).
Final Words
For the reasons arrayed above, we conclude that our defendants were lawfully tried, convicted, and sentenced. Affirmed.
. The Jencks Act takes its name from Jencks v. United States, 353 U.S. 657, 77 S.Ct. 1007, 1 L.Ed.2d 1103 (1957). See United States v. Dupont, 15 F.3d 5, 7 n. 2 (1st Cir.1994) (noting that the statute "codif[ies] Jencks ”).
. Rule 12.1(d) provides that a "court may” excuse noncompliance with the Rule "[f]or good cause” shown.
.The Sixth Amendment does not literally give the accused the right to present witnesses. It gives him the right "to have compulsory process” to "obtain[]” them. See U.S. amend. VI. But the right to compulsory process includes the right to present witnesses. See Taylor v. Illinois, 484 U.S. 400, 409, 108 S.Ct. 646, 98 L.Ed.2d 798 (1988).
. A bundle of heroin, Serrano explained, contains “25 little bags of heroin,” all "wrapped up in a bag.”
. We say "generally” because the high court explicitly declined to revisit "the narrow exception to this general rule for the fact of a prior conviction.” Id. at 2160 n. 1.
. Alleyne applies to cases like this one that were on direct appeal when it was released. See, e.g., Schriro v. Summerlin, 542 U.S. 348, 351, 124 S.Ct. 2519, 159 L.Ed.2d 442 (2004).
.A judge normally applies the guidelines in vogue at the time of sentencing unless doing so "would violate the ex post facto clause,” in which case the judge must use the version "in effect on the date” the defendant committed the offense. See U.S.S.G. § 1B1.11; see also Peugh v. United States, - U.S. -, 133 S.Ct. 2072, 2078-88, 186 L.Ed.2d 84 (2013) (fleshing out the ex post facto analysis).
. Burrito later worked at El Mesón Bypass for a few weeks in December 2005, another witness noted.
. As the government points out, the record strongly suggests that Fournier's counsel well knew the notes’ details before Vidro’s testimony. During his cross-examination of Vidró, for example, Fournier’s lawyer asked, "And isn’t it true that you told an agent that decking was done for Mr. Fournier’s son?” After the prosecutor objected, counsel replied that "[i]t’s a prior statement of the witness.” And at side bar counsel said twice that he had “read it somewhere,” adding that Vidró "said that to an agent pursuant to a document I saw.” If true, that would be a problem for Fournier. See Ellsworth v. Warden, 333 F.3d 1, 6 (1st Cir.2003) (holding that even exculpatory "[ejvidence is not suppressed if the defendant either knew, or should have known[,] of the essential facts permitting him to take advantage of any exculpatory evidence”) (internal quotation marks omitted). On the next trial day, however, Fournier’s attorney did a 180, telling the judge that he had not known *197about "those statements.” We need not referee this duel, because Fournier’s argument fails for the reasons recorded above.
. In a supplementary pro se brief, Fournier contends for the first time that the evidence at trial materially varied from the allegations in the indictment. As he sees it, the evidence shows that he may have violated some gun laws but did not prove that he had joined the charged drug conspiracy. His argument has no bite, however, because (as we just explained) sufficient evidence supports his drug-conspiracy conviction.
. As any crime-drama enthusiast knows, Miranda v. Arizona requires that law-enforcement officers advise a custodial suspect of four rights before interrogating him:
[A suspect] must be warned ... [1] that he has the right to remain silent, [2] that anything he says can be used against him in a court of law, [3] that he has the right to the presence of an attorney, and [4] that if he cannot afford an attorney one will be appointed for him prior to any questioning.
384 U.S. 436, 479, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966). Officers need not use any particular magic words, see id. at 476, 86 S.Ct. 1602 — any words that clearly inform the suspect of his rights will do, see Florida v. Powell, *200559 U.S. 50, 60, 130 S.Ct. 1195, 175 L.Ed.2d 1009 (2010).
. It may seem strange to talk about plain error, given how Alleyne came down after our five defendants argued these consolidated appeals to us. But to preserve a claim, a litigant must press the point below, even if the caselaw is against him — otherwise our review is (at best) limited to plain error. See United States v. Harakaly, 734 F.3d 88, 93-95 (1st Cir.2013).
. Somewhat relatedly, Rodriguez grouses that because the judge let jurors have a copy of the indictment in the jury room during deliberations, they had at their fingertips the *203"unproved” allegations about the other drug points, which, he adds, without citing a single case, caused "unfair prejudice.” But the judge instructed them that the indictment itself is not evidence of guilt. He also told them that the defendants are presumed innocent until proven guilty beyond a reasonable doubt. We presume that juries follow instructions. See, e.g., United States v. Venti, 687 F.3d 501, 504 (1st Cir.2012). And those instructions are sufficient "to guard against the jury’s using the indictment as evidence,” United States v. McFarlane, 491 F.3d 53, 60 (1st Cir.2007), which takes all the wind out of Rodriguez's sail on this issue.
. As an aside, we note that an interesting issue now before the Supreme Court is
[wjhether the offense of aiding and abetting the use of a firearm during and in relation to a crime of violence or drug trafficking crime, in violation of 18 U.S.C. §§ 924(c)(1)(A) and 2, requires proof of (i) intentional facilitation or encouragement of the use of the firearm, as held by the First, Second, Third, Fifth, Seventh, Eighth, Ninth, and Eleventh Circuits, or (ii) simple knowledge that the principal used a firearm during a crime of violence or drug trafficking crime in which the defendant also participated, as held by the Sixth, Tenth, and District of Columbia Circuits.
Petition for Writ of Certiorari at I, Rosemond v. United States, No. 12-895 (U.S. Jan. 16, 2013); see also Rosemond v. United States, - U.S. -, 133 S.Ct. 2734, 186 L.Ed.2d 191 (2013) (granting certiorari).
. We can and do rely on Badamo as persuasive authority. See 1st Cir. R. 32.1.0(a).
. For sentencing purposes, judges must convert diverse drugs into "marijuana equivalents” so that they are adding apples to apples in computing the defendant’s offense level. See Kinsella, 622 F.3d at 82.
. Alberto, by the way, was also indicted in this case but pled guilty to the drug-conspiracy count long before trial.