UNITED STATES of America, Plaintiff-Appellee, v. John Odell CULVERHOUSE, also known as John Odel Culverhouse, Defendant-Appellant, 507 F.3d 888

EMILIO M. GARZA, Circuit Judge: John Odell Culverhouse (“Culverhouse”) appeals the district court’s denial of his motion to vacate, set aside, or correct his sentence pursuant to 28 U.S.C. § 2255. For the following reasons we vacate and remand to the district court. I In February 2002, a grand jury indicted Culverhouse on counts of manufacture and attempted manufacture of methamphetamine in violation of 21 U.S.C. §§ 841(a)(1) and 846. Count one charged manufacture on or about August 19, 2001 and count two charged attempted manufacture on or about December 10, 2001, both offenses occurred in or around Waco, Texas.

In March 2002, the court appointed Marlin Blackledge (“Blackledge”) to represent Culverhouse. In February, the court had also appointed Blackledge to represent Damen Henry (“Henry”), a man who had prior involvement with Culverhouse related to his dealings in methamphetamine. In April, Henry was debriefed by investigators with Blackledge present. Henry had engaged in methamphetamine dealings with Culverhouse in the past. Based on the record, we cannot determine what information Henry shared at his debriefing.

Aside from the debriefing, during the police’s prior investigations of Culverhouse’s manufacture, Henry told officers that he had personally observed Culverhouse with 800 ephedrine tablets on or about August 18, 2002. At his rearraignment in May, while represented by Blackledge, Culverhouse pleaded guilty to the manufacturing charge. As part of his plea agreement, the government did not pursue the attempt charge.

Culverhouse’s presentence investigation report (“PSR”) states that at the debriefing Henry told investigators that on two prior occasions in San Antonio, he had seen Culverhouse with four to five containers holding approximately seven ounces of methamphetamine each. The PSR also included Henry’s statements regarding the 800 tablets.

In the PSR, the probation officer recommended that Culverhouse be held accountable for the following drug amounts: 800 ephedrine pills seen by Henry in Culverhouse’s residence on or about August 18, 2001 2.66 grams of methamphetamine confiscated when Culverhouse’s residence was searched on August 19, 2001 0.65 gram of methamphetamine recovered at a neighboring residence where Culverhouse was thought to have “cooked” the methamphetamine 320 pseudoephedrine pills confiscated by law enforcement officials in a search of Culverhouse’s vehicle on December 10, 2001 4 containers each holding seven ounces of methamphetamine (793.8 grams) seen in Culverhouse’s possession in San Antonio, Texas, according to Henry.

In June 2002, Blackledge moved to withdraw from representation of Culverhouse after reading the PSR and noting the conflict between his clients, Culverhouse and Henry. The district court granted the motion and then appointed new counsel, Vik Deivanayagam (“Deivanayagam”), to represent Culverhouse at sentencing. Deivanayagam lodged two objections related to the amount of drugs attributed to Culverhouse for sentencing purposes. Between Culverhouse’s plea and sentencing, Henry admitted that he never saw Culver-house with 800 tablets. The first objection focused on this change in Henry’s story.

The government, now knowing the statement to be false, conceded this objection and the district court did not consider the 800 tablets in sentencing. The second objection dealt with Henry’s credibility and the use of Henry’s statements regarding the containers to attribute 793.8 grams of methamphetamine to Culverhouse for sentencing purposes. Henry testified at the sentencing hearing and recounted again his two prior meetings with Culverhouse in San Antonio.

In response to questioning, he noted that the transaction in which he saw the containers of methamphetamine occurred “about four years ago.” The judge, finding Henry credible as to his statements about the prior transaction, overruled the second objection. Based on a total offense level of 33, and a criminal history category of VI, the court sentenced Culverhouse to 235 months’ imprisonment, 3 years’ supervised release, a $4,000 fine and $100 mandatory assessment.

Culverhouse moved to vacate, set aside, or correct his sentence under 28 U.S.C. § 2255 based, in part, on ineffective assistance by both Blackledge and Deivanaya-gam. The district court denied his motion without holding an evidentiary hearing.

We granted a certificate of appealability (“COA”) on two issues raised by Culver-house: (1) whether Blackledge performed ineffectively by engaging in a conflict of interest during his representation of Cul-verhouse and; (2) whether Deivanayagam performed ineffectively by failing to object adequately to a four-year-old methamphetamine transaction that was used as relevant conduct for sentencing purposes. Following the district court’s denial of a § 2255 motion, we review the district court’s mixed factual and legal conclusions with regard to ineffective assistance of counsel de novo.

We review any factual findings under the clearly erroneous standard. See United States v. Molina-Uribe, 429 F.3d 514, 518-19 (5th Cir.2005). II A The Sixth Amendment right to counsel includes the “right to representation that is free from any conflict of interest.” United States v. Vaquero, 997 F.2d 78, 89 (5th Cir.1993). To prove a Sixth Amendment violation based on a conflict of interest arising from multiple representation, Cuyler v. Sullivan requires a defendant to show: (1) that his counsel was acting under the influence of an actual conflict that (2) adversely affected representation.

See Cuyler, 446 U.S. 335, 348-49, 100 S.Ct. 1708, 64 L.Ed.2d 333 (1980); see also Mickens v. Taylor, 535 U.S. 162, 169, 122 S.Ct. 1237, 152 L.Ed.2d 291 (2002) (clarifying the standard under Cuyler). The defendant need not show prejudice in the sense that the outcome of the proceeding would have been different absent the conflict because prejudice is presumed upon a showing of an actual conflict that adversely affected representation. See Perillo v. Johnson, 205 F.3d 775, 781-82 (5th Cir.2000). Multiple representation does not necessarily create an actual conflict of interest. See United States v.

Garcia-Jasso, 472 F.3d 239, 243 (5th Cir.2006). We have noted that the first prong of Cuyler requires a defendant to show something more than a speculative or potential conflict. See United States v. Infante, 404 F.3d 376, 391-92 & n. 13 (5th Cir.2005). As such, we have stated that a conflict will exist only when counsel is “compelled to compromise his or her duty of loyalty or zealous advocacy to the accused by choosing between or blending the divergent or competing interests of a former or current client.” Garcia-Jasso, 472 F.3d at 243 (5th Cir.2006) (internal quotation marks omitted).

Thus, there is an actual conflict when the attorney knows that his clients’ interests diverge and must then choose between the interests of multiple clients, or be compelled to compromise his duty of loyalty. This brings us to the crux of the conflict issue in this case, that being when Blackledge learned of his clients’ conflicting interests. Culverhouse claims that at the debriefing, Henry provided the statement regarding Culverhouse’s 800 ephedrine tablets, as well as information about the prior meetings between Henry and Culverhouse.

The PSR, partially consistent with Culver-house’s claim, states that at the debriefing Henry provided information about the earlier meetings. Culverhouse notes that Blackledge, as Henry’s attorney, was present for this debriefing. Thus, Culverhouse argues that Blackledge’s interests were diametrically opposed when he represented both Henry and Culverhouse while Henry provided information to investigators that was beneficial to Henry but damaging to Culverhouse. Culverhouse contends that Henry gave the investigators evidence to make Culverhouse’s plea or conviction more likely in return for lenience.

Culver-house also claims that, in order to secure lenience for Henry, Blackledge assured Culverhouse’s guilty plea by “diverting” Culverhouse’s own attempts to object to the factual basis of his plea. The government contends that Culver-house presents only a speculative conflict. They argue that Culverhouse cannot show that Blackledge made the necessary choice between divergent interests. The record contains an affidavit filed by Blackledge wherein he states that, based on his appearance at the debriefing he did not recognize any “cross-linking” of the defendants.

The government, consistent with Blackledge’s motion to withdraw, contends that Blackledge did not learn of the conflicting interests of his clients until he read the PSR. The district court, in denying Culver-house § 2255 relief, noted that Culver-house could not establish a conflict by Blackledge because Blackledge had withdrawn before Culverhouse’s sentencing. However, Blackledge could have suffered under a conflict during his representation of Culverhouse at the guilty plea stage.

Whether a conflict exists in this case ultimately depends on when Blackledge became aware that Henry had information damaging to Culverhouse’s case. The record presents a factual dispute as to what information was shared by Henry at the debriefing. If Henry shared damaging information about Culverhouse, then Black-ledge suffered under a conflict in representing Culverhouse. Upon learning of Henry’s information Blackledge’s loyalties would have become divided at that point he could have counseled Henry to use, as part of a plea deal, the information damaging to Culverhouse.

And as Culverhouse alleges, Blackledge could have pressured Culverhouse to plead guilty based on his desire to secure lenience for Henry. See Infante, 404 F.3d at 390 and 392 (Where attorney represented witnesses who testified against his client, a conflict existed, in part, because those witnesses could receive lenience for their cooperative testimony.).

However, if Blackledge did not learn about Henry’s information until he received the PSR, then he did not suffer under an actual conflict during his representation of Culverhouse at the rearraignment hearing for he could not have “made a choice between possible alternative courses of action ... [and] if he did not make such a choice the conflict remained hypothetical.” Garcia-Jasso, 472 F.3d at 243.

Because of a lack of necessary evidence as to when Blackledge knew that Henry had information harmful to Culverhouse, we cannot determine on this record whether Blackledge suffered under an actual conflict during his representation of Cul-verhouse. After receiving the PSR, Black-ledge recognized that representation of both Henry and Culverhouse created a conflict. We cannot tell whether Black-ledge suffered from the same conflict during his representation of Culverhouse. Therefore, we must remand to the district court to have it clarify the matter via an evidentiary hearing.

Upon remand, if the district court finds that an actual conflict existed during Blackledge’s representation at the pleading stage, the court should then proceed to determine whether the actual conflict adversely affected Black-ledge’s performance. See, Perillo, 205 F.3d at 807 (“An adverse effect ... may be shown with evidence that counsel’s judgment was actually fettered by concern over the effect of certain trial decisions on other clients.”) (internal quotation marks omitted).

B Culverhouse also contends that his sentencing counsel, Deivanayagam, performed ineffectively by failing to present a valid argument for exclusion of the prior drug transaction in San Antonio with Henry from the' drug-quantity determination. Culverhouse claims that the transaction, which occurred close to three years prior to his manufacturing conduct, could not be considered relevant conduct under the sentencing guidelines. Deivanayagam did object to the use of the prior meetings with Henry to attribute drugs to Culverhouse.

However, when explaining his objection at the sentencing hearing, Deivanayagam’s argument focused entirely on Henry’s credibility. Based on the fact that Henry lied about seeing Culverhouse with the 800 tablets, Deivanayagam argued that Henry could not be trusted to tell the truth about the containers of methamphetamine. The district court, in ruling on Culverhouse’s § 2255 motion, noted that the sentencing court found by a preponderance of the evidence that the quantity of drugs used in the calculation was appropriate, and further found that the time interval did not defeat the finding of relevant conduct.

Therefore, the district court determined that an objection would have been without merit. We cannot, on this record, determine whether the district court’s assessment is correct. To prevail on a claim of ineffective assistance of counsel without establishing a conflict, a movant must proceed under the traditional Strickland standard by showing (1) that his counsel’s performance was deficient in that it fell below an objective standard of reasonableness, and (2) that the deficient performance prejudiced his defense. Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

In order to determine whether Deivanayagam erred in not adequately objecting to the district court’s finding of relevant conduct, we must consider the correctness of the district court’s relevant conduct ruling. United States v. Franks, 230 F.3d 811, 813 (5th Cir.2000). The sentencing judge made no express findings to support his relevant conduct conclusion, presumably adopting the PSR. Therefore, we must evaluate the PSR and the sentencing hearing to evaluate whether the finding can be supported by the evidence necessary to establish relevant conduct under the sentencing guidelines.

A defendant convicted of a drug offense is sentenced based on the amount of drugs involved in the offense, with quantities of drugs from multiple transactions added together. See U.S. Sentencing Guidelines MaNual (“USSG”) § 2D1.1, application note 6 (2001). The guidelines provide that, in determining the amount of drugs to be attributed for a base offense level, the district court may consider other offenses than those underlying the offense of conviction as long as the offenses constitute relevant conduct as defined in the guidelines. See id. § 1B1.3.

The. guidelines define “relevant conduct” to include “all acts and omissions ... that were part of the same course of conduct or common scheme or plan as the offense of conviction.” Id. § lB1.3(a)(2). An offense need not have resulted in a conviction to constitute relevant conduct under the guidelines, but the conduct must be criminal. See United States v.

Anderson, 174 F.3d 515, 526 (5th Cir.1999). “For two or more offenses to constitute part of a common scheme or plan, they must be substantially connected to each other by at least one common factor, such as common victims, common accomplices, common purposes, or similar modus oper-andi.” USSG § IB 1.3, application note 9(A). The record contains no evidence to suggest that the prior transaction to which Henry testified was part of a larger scheme or plan including Culverhouse’s manufacture three years later. The first offense involved Culverhouse alone, working as a distributor, with no accomplices.

In the conviction offense, Culverhouse worked with others in manufacturing methamphetamine. The two offenses can be connected by only the most general of purposes, in that they both involved methamphetamine. See, United States v. Wall, 180 F.3d 641, 645 (5th Cir.1999) (finding no common scheme where offenses shared only “the common general purpose of importing marijuana to the United States.”). Finally, the record contains no evidence of similarities in modus operandi.

In determining whether offenses are part of a common course of conduct, the commentary to § 1B1.3 provides that offenses must be “sufficiently connected or related to each other to warrant the conclusion that they are part of ... [an] ongoing series of offenses.” USSG § 1B1.3, application note 9(B). In measuring the sufficiency of this connection a court may consider “the degree of similarity of the offenses, the regularity of the offenses, and the time interval between the offenses. When one of the abóve factors is absent, a stronger presence of at least one of the other factors is required.” Id.

In evaluating the ruling of the sentencing court, we consider whether any of the three guideline factors, temporal proximity, regularity, and similarity can support the court’s finding that the transaction with Henry constitutes relevant conduct as part of the common course of conduct. The transaction with Henry occurred roughly three years prior to the conduct underlying Culverhouse’s manufacture charge. Our precedent at the time of Cul-verhouse’s sentencing clearly shows that temporal proximity was lacking. See United States v.

Miller, 179 F.3d 961, 966 n. 10 (5th Cir.1999) (finding two offenses, where one occurred nearly four years prior and the other twenty one months prior, temporally remote, and noting that the longest time span previously recognized by the Fifth Circuit had been eighteen months); Wall, 180 F.3d at 645-46 (finding temporal proximity lacking for offenses occurring four and five years apart, and noting that offenses were separated by “an unprecedented lapse of time for a case involving drug distribution”). However, a failure in temporal proximity does not, by itself, prevent a finding of relevant conduct.

The guidelines state that a stronger presence of regularity or similarity can compensate for the absence of temporal proximity. See USSG § 1B1.3, application note 9(B). In evaluating similarity, the pertinent inquiry is whether “there are distinctive similarities between the offense of conviction and the remote conduct that signal that they are part of a course of conduct rather than isolated, unrelated events that happen only to be similar in kind.” Wall, 180 F.3d at 646 (quoting United States v. Maxwell, 34 F.3d 1006, 1011 (11th Cir.1994)).

Based on the evidence before the sentencing judge, these two occurrences had no distinct similarities. The fact that they both involved methamphetamine is not enough. See Miller, 179 F.3d at 967. The offenses took place in different cities more than 150 miles apart. The first involved Culver-house as a distributor, the second, Culver-house as a manufacturer. Nor is there any evidence that the methamphetamine in the two offenses shared a common source, supplier, destination, or that the two offenses involved a similar modus operandi.

See Wall, 180 F.3d at 646 (evaluating similarity of source, supplier, destination and modus operandi for two drug offenses). Based on our case law at the time, the evidence before the district court at sentencing could not have supported a finding of similarity. While we question whether a showing of regularity alone could support a finding of relevant conduct when coupled with an absence of temporal proximity and similarity, we assume arguendo that a showing of sufficient regularity could provide the necessary support.

To demonstrate the necessary regularity, the government points to portions of the PSR showing that Culverhouse was involved in three other methamphetamine-related incidents between 1987 and his 2002 indictment, as well as PSR evidence that Culverhouse was known to be dealing methamphetamine in 1997. These “offenses” are drawn from the PSR’s criminal history report, as well as PSR data pulled from police reports related to Culverhouse. The sentencing judge could not have relied on this evidence to establish the regularity required by § 1B1.3.

None of these offenses occurred between the prior transaction with Henry and Culverhouse’s later manufacturing. See Wall, 180 F.3d at 646 (“Nor is there a pattern of regularity in light of the lengthy separation between [the offenses] and the lack of evidence of any illicit conduct between the two.”).

Further, our case law at the time established that five offenses over the course of fifteen years, “separated as they are by time and circumstances, cannot be considered repetitious or regular conduct to a degree significant enough to constitute sufficient connection under the guidelines.” Miller, 179 F.3d at 967; see also United States v. Ocana, 204 F.3d 585, 591 (5th Cir.2000) (finding regularity present where Ocana recruited drug couriers for trips in July, September, and November of the same year); United States v.

Bethley, 973 F.2d 396, 401 (5th Cir.1992) (Where defendant’s cocaine distribution activities took place within six months of the offense for which he was convicted, were of a continuous nature, and involved similar quantities and types of drugs, the district court did not err in finding the conduct relevant.). The record provides insufficient evidence of regularity necessary to support a finding of relevant conduct in the absence of both temporal proximity and similarity between the offenses.

The offenses in this case lack the temporal proximity, similarity, and regularity necessary to prove a common course of conduct. Neither can they be shown to be part of a common scheme or plan. Based on the law of this circuit at the time of Culverhouse’s sentencing and the record, the requirements for a finding of relevant conduct under § 1B1.3 could not be met. While we recognize that an objection to the district court’s relevant conduct finding likely would have been meritorious, this does not necessarily mean that counsel performed deficiently in failing to raise the objection.

We have described Strickland ’s first prong as “requiring that counsel research relevant facts and law, or make an informed decision that certain avenues will not be fruitful.” United States v. Phillips, 210 F.3d 345, 348 (5th Cir.2000) (internal quotation marks and citation omitted). This means that .“[s]olid, meritorious arguments based on directly controlling precedent should be discovered and brought to the court’s attention,” Id.

However, in assessing whether an attorney’s performance was deficient the court “must indulge a strong presumption that counsel’s conduct falls within a wide range of reasonable professional assistance; that is ... [that] the challenged action might be considered sound trial strategy.” Strickland, 466 U.S. at 689, 104 S.Ct. 2052. Further, we must make our best effort “to eliminate the distorting effects of hindsight, to reconstruct the circumstances of counsel’s challenged conduct, and to evaluate the conduct form counsel’s perspective at the time.” Id.

A finding of relevant conduct was required to attribute drugs to Culverhouse based on the earlier transaction. A review of the pertinent guidelines provision makes clear the necessary elements for a finding of relevant conduct. The above analysis, conducted in the vacuum that is our record on appeal, suggests that the objection would have been meritorious. Despite the apparent merit of the relevant conduct objection, we have no way to analyze the potential strategy behind the decision made by counsel. Counsel may have had valid reasons not to raise the objection.

Without any evidence as to counsel’s strategy, we refuse to make hindsight guesses on the matter. Because of a lack of necessary record evidence, we cannot adequately review the district court’s judgment as to whether Deivanayagam performed unreasonably under Strickland ’s first prong. Because we make no judgment as to counsel’s performance, we do not reach Strickland’s prejudice prong. Ill We do not have the requisite evidence before us to decide either ineffective assistance issue raised in Culverhouse’s COA.

Rather than decide Culverhouse’s ineffective assistance claims based on scant evidence or unnecessary assumptions, “the better approach is to have the district court conduct an evidentiary hearing.” United States v. Herrera, 412 F.3d 577, 582 (5th Cir.2005). For the foregoing reasons we VACATE the judgment of the district court as to the two issues on which we granted a COA and REMAND for further proceedings consistent with this opinion. .

After converting these drugs into marijuana equivalency, the PSR recommended attributing 2,218.22 kilograms of marijuana to Cul-verhouse, the vast majority of which was based on the information given by Henry. . Henry’s testimony placed his prior transaction with Culverhouse "about four years" pri- or to sentencing. This would put the transaction roughly three years prior to the conduct underlying the charge to which Culverhouse pleaded guilty. . See also Hall v.

United States, 371 F.3d 969, 974 (7th Cir.2004) ("[A] defendant can establish ineffective assistance of counsel by showing that his attorney pressured him to plead guilty because of the attorney’s conflict of interest.”); Berry v. United States, 293 F.3d 501, 504 (8th Cir.2002) (analyzing whether a reasonable attorney would have advised defendant to plead, and whether there was an adverse effect based on attorney's pressure to plead); Pegg v. United States, 253 F.3d 1274, 1278 (11th Cir.2001) (asking whether attorney's conflict "adversely affected defendant's decision to plead guilty”); Thomas v.

Foltz, 818 F.2d 476, 480 (6th Cir.1987) (asking whether the conflict "adversely affected the voluntary nature of the guilty plea entered.”). We note that Culverhouse, in showing an adverse effect, need not show that but for Blackledge's conflict there was a reasonable probability he would have decided to go trial. This "but for” showing is required for Strickland prejudice, but we have recognized that "the adverse effect standard is set intentionally lower than Strickland's actual prejudice standard.” Perillo, 205 F.3d at 806; see also Ward v.

Dretke, 420 F.3d 479, 487 (5th Cir.2005) (stating Strickland prejudice requirements in guilty plea setting). Because Culver-house need not show that he would have gone trial absent the conflict, the fact that he failed to withdraw his guilty plea at sentencing does not necessarily prevent a finding of adverse effect under Cuyler. The facts necessary to a conflict of interest claim oftentimes do not become apparent to a defendant until after sentencing.

We question whether Culver-house, just a month after his original counsel's withdrawal, understood the ramifications of Blackledge’s multiple representations to the extent necessary to move for withdrawal of his plea during sentencing. . Neither the PSR nor the sentencing judge made any reference to "relevant conduct” or to § IB 1.3 of the guidelines. The PSR appears to have assumed that the prior conduct testified to by Henry could be used to attribute drugs to Culverhouse.

While the district court can certainly rely on the PSR, see Anderson, 174 F.3d at 526 n. 3, we encourage courts to make clear their reasons for treating separate offenses as relevant conduct under § 1B1.3, either by stating their own reasons on the record or making sure that the PSR upon which they rely contains sufficient explanation. . Application note 9(

B) provides that temporal proximity, regularity, and similarity are “factors to be considered.” The note goes on to state that "[w]hen one of the factors is absent, a stronger presence of at least one of the other factors is required.” The note does not discuss the required showing when two factors are absent. See United States v. Jackson, 161 F.3d 24, 28-29 (D.C.Cir.1998) (finding that nothing necessarily precludes a district court judge from finding a course of conduct based primarily on similarity). .

Along with the prior conduct testified to by Henry, which appears to have occurred sometime in 1998, and the manufacture in August 2001 to which Culverhouse pleaded guilty, the PSR includes the following. In September 1987 Culverhouse engaged in two large transactions, one for which he received $10,000, whereby he sold fake methamphetamine to undercover agents in San Antonio. As a result of these two transactions, Culverhouse pleaded guilty to possession of a firearm by a convicted felon.

In February 1994, during the execution of a traffic warrant, officers discovered three small bags of methamphetamine on Culverhouse’s person. No charges were brought. The PSR indicates that by 1997 Culverhouse was known to be dealing methamphetamine in the Waco, TX area. In December 2001, after the conduct constituting the charge in this case but prior to his indictment, the police pulled over Culver-house’s vehicle and found him to be in possession of a number of items indicative of methamphetamine manufacture, including an amount of drugs attributed to Culverhouse, without objection, at his sentencing.