dissenting.
If the facts regarding Boykin’s counsel’s performance were as the majority has stated, then I would likely concur. But the record indicates that defense counsel did much more than the opinion suggests to vindicate his client and that some of the majority’s conclusions about whether he represented conflicting interests are erroneous. While I certainly believe that it is inadvisable for a criminal defense attorney to represent co-defendants at trial, because defense counsel in this case did not labor under an actual conflict of interest and did not provide constitutionally ineffective assistance, I respectfully dissent from parts III(2)(B) and (3) of the lead opinion.
Joint representation of co-defendants in a criminal trial “does not constitute per se ineffective assistance of counsel.” Moss v. United States, 323 F.3d 445, 455 (6th Cir.2003). As the majority points out, when a defendant does not object at trial to the joint representation, we presume prejudice only if the defendant demonstrates that his attorney actively represented competing interests. Cuyler v. Sullivan, 446 U.S. 335, 350, 100 S.Ct. 1708, 64 L.Ed.2d 333 (1980). We have stated that Sullivan requires a defendant in a joint representation case to establish both an “actual conflict” and an adverse “effect on representation” in order to be entitled to relief under the Sixth Amendment. McFarland v. Yukins, 356 F.3d 688, 705 (6th Cir.2004). To demonstrate an adverse effect on representation, the defendant must show that the conflict caused the attorney’s choice that the defendant alleges was deficient. Id. Additionally, “counsel’s choice to forego a defense that would have been inconsistent with counsel’s duty to another client is evidence of adverse effect only if it is clear that the choice was not part of a legitimate trial strategy, judged under the deferential review of counsel’s performance prescribed in Strickland.” Id. at 706, 104 S.Ct. 2052.
I. Counsel’s Performance at Trial
The majority lists several reasons why it believes Boykin established that his and *649McElrath’s trial counsel had an actual conflict of interest and that he favored McEl-rath. It states that defense counsel should have: attacked McElrath as the shooter or getaway driver; attacked Fitts with his prior inconsistent statements; called rebuttal witnesses to attack Fitts’s testimony; attacked Officer Perry’s testimony; called certain witnesses to identify McEl-rath as the shooter; called Boykin’s and McElrath’s grandmother as an alibi witness; called McElrath and/or Everett to testify that Boykin did not participate in the crime; moved to sever the trial on behalf of Boykin; and discarded the “inexplicable” Fitts-as-shooter defense theory. Majority opinion at pp. 645-47. The majority relies on counsel’s supposed failure to take the above actions to conclude that he was laboring under an actual conflict of interest and that Boykin is entitled to ha-beas relief. I believe that conclusion is incorrect for several reasons.
First, we do not know whether McEl-rath or Everett would have testified favorably for Boykin. Even if McElrath and Everett would have agreed to testify,1 there is absolutely no evidence that they would have said that Boykin was not with them in Clinton, Kentucky, on the night of the shooting. McElrath and Everett admitted to police that they were together the night of the murder. While it is true that neither McElrath nor Everett told police that Boykin was with them that night, they never told police that Boykin was not with them. Second, the record does not contain any affidavits from McEl-rath, Everett, the grandmother, or the other witnesses the majority postulates would have testified favorably for Boykin. Those witnesses who gave statements to police could have been lying and thus been unwilling to take the stand and give sworn testimony under penalty of perjury. Moreover, the witnesses may not have been competent to testify under the Kentucky Rules of Evidence, KRE 601.2
The decision to call or not call certain witnesses is exactly the type of strategic decision that the courts expect attorneys to make. See United States v. Best, 426 F.3d 937, 945 (7th Cir.2005) (“[A] lawyer’s decision to call or not to call a witness is a strategic decision generally not subject to review. The Constitution does not oblige counsel to present each and every witness that is suggested to him.” (citations and internal quotations omitted)); United States v. Luciano, 158 F.3d 655, 660 (2d Cir.1998) (“The decision not to call a particular witness is typically a question of trial strategy that appellate courts are ill-suited to second-guess.”); Blanco v. Singletary, 943 F.2d 1477, 1495 (11th Cir.1991) (“The decision as to which witnesses to call is an aspect of trial tactics that is normally entrusted to counsel.”); Tosh v. Lockhart, 879 F.2d 412, 414 (8th Cir.1989) (“[T]he decision not to use alibi testimony may reflect the reasonable exercise of judgment in view of the attorney’s concern that the testimony would be conflicting ... or otherwise unfavorable.” (internal citations omitted)); cf. Strickland v. Washington, 466 U.S. 668, 689, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) (“[T]he defendant must overcome the presumption that, under the circumstances, the challenged action ‘might be considered sound trial strategy.’ ” (citation omitted)). We are in no position to determine the reasonableness of counsel’s decisions not to call these witnesses because we know not to what facts they would testify; we have merely a few *650unsworn statements to the police. Furthermore, we cannot say with any certainty that defense counsel failed to call these witnesses specifically because, even though they would have been beneficial to Boy-kin’s defense, they would have been detrimental to McElrath’s — both uncertain conclusions.3 Counsel’s decisions could very well have been part of a sound trial strategy and not evidence of a conflict of interest. See McFarland, 356 F.3d at 706.
Next, the majority mistakenly contends that defense counsel did not vigorously cross-examine Fitts or Officer Perry. The record is replete with instances where defense counsel submitted into evidence through cross-examination the very facts the majority claims he was deficient for not arguing. Specifically, counsel confronted Fitts about the following: whether Fitts told the police McElrath committed the shooting; whether Fitts could even identify the shooters because it was dark outside and the shooters were wearing scarves over their faces; and Fitts’s failure to tell Officer Perry on the night of the crime that Boykin was one of the shooters. Additionally, defense counsel adequately explained that Fitts had changed his story.
Similarly, defense counsel extensively cross-examined Officer Perry and elicited the following testimony: only Fitts placed Boykin in Clinton, Kentucky, on the night of the shooting; nobody told police that the gun used in the shooting belonged to Boykin and no physical evidence connected Boykin to the gun; Fitts did not identify Boykin as a shooter until three or four days after the shooting; Fitts needed to pick Boykin out of a lineup to identify him, even though he knew Boykin personally; and, Perry did not remember asking McElrath during questioning whether Boykin was with McElrath in Clinton, Kentucky, the night of the shooting.
The majority opinion also stretches other facts. For example, it states that police found the murder weapon “at McElrath’s (and Boykin’s) grandmother’s house,” but fails to mention that Boykin lived at that house4 and the police found the weapon in his bedroom closet. Moreover, the majority opinion flatly states that Boykin’s ballistics expert testified that Fitts could not have been the shooter, but the trial transcript reveals that the expert testified that she could neither say that Fitts had fired a weapon or that he had not. And the majority’s pointing out that defense counsel failed to call to the stand “every witness that identified McElrath as the shooter,” is somewhat misleading — only Fitts and Wilson’s five-year-old son identified McElrath, and Fitts did testify, albeit for the prosecution.5
Most importantly, the majority opinion fails to articulate adequately why any of defense counsel’s other actions or inactions demonstrate an actual conflict of interest. That is, the majority opinion simply concludes that counsel’s trial strategy benefit-ted McElrath to the detriment of Boykin. And its determination that Boykin’s best defense would be to paint McElrath as the shooter misses the mark. Boykin’s best defense was to contend that he was not in *651Clinton, Kentucky, on the night of the murder, and defense counsel presented that defense through his cross-examinations of Fitts and Officer Perry. We are required to evaluate counsel’s performance without the distorting effects of hindsight. See Strickland, 466 U.S. at 689, 104 S.Ct. 2052. Defense counsel’s method of introducing evidence favorable to Boykin may not meet the majority’s subjective standard of performance, but it certainly does not support the conclusion that there was an actual conflict of interest. McElrath did not dispute that he was in the area at the time of the shooting; Eric and Sammy Hunter testified that McElrath unexpectedly visited their parents’ house down the street from the shooting and that he left immediately after the shooting. Pointing to McElrath as the shooter would not necessarily benefit Boykin — as long as Boykin was not the shooter it does not matter if McElrath (or Everett) was. The propositions that McElrath did not kill Wilson and that Boykin did not kill Wilson are not mutually exclusive.
Finally, the majority incorrectly compares this ease to McFarland. Majority opinion at p. 646. In that case, the defense attorney representing co-defendants Reeves and McFarland indicated to the court on the day of the trial that he might have to raise antagonistic defenses and, upon questioning by the court, both of the co-defendants explicitly stated that they believed they should have separate attorneys. McFarland, 356 F.3d at 693-94. Instead of allowing the defendants to get new attorneys, however, the trial court severed the cases and a different judge tried each one. Id. at 694. What distinguishes McFarland from the case at bar is that the defendants in McFarland had obviously antagonistic defenses — the police found drugs in Reeves’s bedroom, but counsel refused to utilize that fact to exculpate the McFarland — and the attorney made ambiguous statements about certain events, appearing to inculpate McFarland even though the evidence inculpated Reeves. Id. at 705, 708-09. The main issue in McFarland was whether each defendant possessed the drugs, an element the prosecution struggled to prove. It was clear that McFarland, who did not live in the bedroom in which the police found the drugs and who was not named by witnesses as selling drugs, could have simply asserted that Reeves possessed the drugs, but the attorney did not advance that defense because of his duty of loyalty to Reeves.
The majority contends that McFarland is factually on point because defense counsel here advanced an inexplicable theory (here the Fitts-as-shooter theory; in McFarland the theory that neither of the defendants possessed the drugs but that the two male occupants of the house did) instead of an obviously strong defense (here the McElrath-as-shooter and Boy-kin-in-Tennessee theory; in McFarland that only Reeves- had possession of the drugs). But McFarland is inapposite to the instant case because the majority’s proposed defense strategy (McElrath as the shooter) is not obviously beneficial to Boykin. Boykin could not exculpate himself simply by pointing to McElrath as the shooter because: (1) there were two shooters — theoretically both Boykin and McEl-rath could have been shooters, while in McFarland only Reeves lived in the bedroom in which the police found the drugs and witnesses identified only Reeves as being involved with the drugs; and (2) the Hunters’ testimony places McElrath down the street at the time of the crime, so labeling him as the shooter would have forced defense counsel to attack the Hunters. In response to the fact of the Hunters’ testimonies, Boykin contends that his attorney should have at least advanced the position that McElrath was the getaway *652driver. But again, that defense gets Boy-kin nowhere; he could have been a shooter and McElrath could have been the getaway driver. The Fitts-as-shooter joint defense theory may not have been wise given the evidence, but it certainly was not antagonistic to Boykin’s interests.
Thus, we are left to decide whether defense counsel’s seeking severance of the trial on behalf of McElrath but failing to seek a severance on behalf of Boykin establishes an actual conflict of interest. Defense counsel filed the motion to sever on behalf of McElrath because McElrath was charged only with complicity in the crime.6 Perhaps defense counsel should have filed motions to sever on behalf of both Boykin and McElrath, but the failure to do so alone does not demonstrate a conflict of interest. Once the court denied the motion to sever, it was reasonable to assume that it would also deny a similar motion on behalf of Boykin.
Because Boykin’s defense attorney did not labor under an actual conflict of interest, Boykin cannot meet the Sullivan Standard. Consequently, Boykin must prove the traditional Strickland elements: (1) that his attorney’s performance was constitutionally deficient; and (2) that the attorney’s deficient performance prejudiced Boykin. Strickland, 466 U.S. at 687, 104 S.Ct. 2052. Boykin has not satisfactorily established either prong of the Strickland standard. Therefore, I would hold that Boykin’s attorney did not provide ineffective assistance of counsel at the trial.
II. Counsel’s Performance on Direct Appeal
The majority relies upon the same facts and analysis that led to its conclusion that counsel had an actual conflict of interest at trial to determine that counsel had an actual conflict of interest on direct appeal as well. Based on the analysis in the preceding section, I disagree with the majority’s conclusion.
CONCLUSION
Respectfully, I cannot concur in the majority’s holding that Boykin’s attorney labored under an actual conflict of interest and therefore provided ineffective assistance of counsel. Therefore, I would AFFIRM the district court’s denial of Boy-kin’s petition for a writ of habeas corpus.
. As defendants in the trial, McElrath and Everett could not be compelled to testify.
. This would have been a particularly relevant consideration when deciding whether to put Wilson's five-year-old son on the stand.
. For example, it is not at all detrimental to McElrath for his and Boykin’s grandmother to testify that Boykin was at home in Tennessee the night of the murder.
. McElrath lived at a different residence.
. It is probable that Fitts’s initial statement to the police officer Morrison that “Treon done it” was a conclusion stemming from Fitts’s having seen McElrath drive by the house earlier, and not truly an identification of McEl-rath as one of the actual shooters. Nevertheless, defense counsel solicited from Officer Morrison the testimony about Fitts’s initial, conflicting statement.
. The motion reads:
Comes the Defendant, TREON McELRATH, and moves this Court, pursuant to CR 9.16, to grant said Defendant a separate trial from the other Defendants in that Treon McElrath is only charged with complicity and to do otherwise he may be found guilty by association, therefore would be prejudiced by trial with the Defendants charged with murder.