United States v. Ofray-Campos

KEENAN, District Judge. These five consolidated appeals arise from the convictions of the defendants, after a jury trial, for their participation in a multi-drug conspiracy. On October 4, 2002, a federal grand jury, sitting in the District of Puerto Rico, returned a two-count Indictment, charging forty-three defendants with conspiracy to distribute five kilograms or more of cocaine, fifty grams or more of crack, and one kilogram or more of heroin, in violation of 21 U.S.C. §§ 841 and 846. 1 The charges stemmed from the defendants’ alleged participation in Las Avispas, a heroin, powder cocaine, *11 crack, and marijuana ring that operated drug distribution centers, or drug “points,” throughout neighborhoods in the Guayama and Salinas regions of Puerto Rico, from April 1993 to September 2002. Of the forty-three defendants who were indicted, thirty-seven pleaded guilty. The present Defendants-Appellants, Mizaury López-Soto (“López-Soto”), Heriberto Ofray-Campos (“Ofray”), Pedro José Diaz-Cla-vell (“Diaz-Clavell”), Dennys Cruz-Pereira (“Cruz-Pereira”), and Modesto Zaragoza-Lasa (“Zaragoza-Lasa”) (collectively, the “Appellants”), opted to go to trial, along with an additional defendant, Carlos Esco-bar-Figueroa, whose appeal proceeded separately. Trial began on August 5, 2003 and concluded on September 29, 2003. The jury found the Appellants guilty and indicated by a special verdict form that the charged conspiracy involved the threshold amounts of the narcotics described in the Indictment. For the reasons that follow, we vacate the convictions of Diaz-Clavell and Zara-goza-Lasa and remand for new trials. We affirm the convictions of Cruz-Pereira and López-Soto but vacate their sentences and remand for re-sentencing. We affirm the conviction and sentence of Ofray. BACKGROUND In setting forth the background of this case, we present the facts in a light that is most favorable to the Government’s case and thus supportive of the jury’s verdict. We provide additional facts where they are relevant to the legal analysis of specific issues. See United States v. Rodriguez-Marrero, 390 F.3d 1, 6 (1st Cir.2004). The Government’s evidence at trial consisted, among other things, of the testimony of law enforcement agents and cooperating witnesses; audio recordings of drug transactions; and physical evidence, including guns and narcotics, recovered during the course of the investigation. The evidence established that, from 1993 to 2002, Las Avispas, a gang with approximately forty members under the leadership of José Dávila-López (a/k/a José Ca-bezón), owned and operated drug points in Guayama and Salmas from which members of the organization sold large quantities of cocaine, crack, heroin, and marijuana. The drug points operated by Las Avispas included the points at Las Vias and La Plumita, in the Borinquen ward of Guaya-ma. For several years during the time period at issue, Las Avispas was engaged in a violent war over territory in the Bor-inquen ward with a smaller gang, Los Jibaritos, that operated a drug point at Las Ruinas, in close proximity to the Las Vias and La Plumita drug points. The Las Vias point carried out a thriving narcotics trade, operating around the clock with nearly constant demand from street users. Francisco Rivera Cardona, a former Avispas street dealer who sold crack and cocaine at Las Vias from 1996 to 1999, testified that he typically sold between 400 and 500 capsules of crack per eight-hour shift, as well as approximately twenty-five $10 bags of cocaine per shift. Members of Las Avispas and Los Jibaritos regularly carried weapons and frequently engaged in shoot-outs with each other and with members of other rival gangs in the area. Las Avispas, Los Jibaritos, and other neighboring drug rings operated under similar principles. Drug point owners, such as José Cabezón, were rarely seen at the actual drug points and had limited interaction with the gangs’ lower level members. In addition to drug point owners and managers, workers in the drug rings included suppliers, who sold large quantities of narcotics on a regular basis to several different gangs; sellers, who sold the packaged drugs on the street to end users; runners, who kept the points *12 stocked with drugs; and enforcers, who often carried firearms and used violence to protect the gangs' members, drugs, and drug proceeds from the violent encroachment of rivals. Often, members of the drug gangs occupied overlapping roles. The Government’s witnesses also described how different drugs were packaged. Cocaine and marijuana were sold in plastic bags, in $5 and $10 quantities. Crack was sold in small plastic vials, at $3 or $3.50 per vial. Heroin was packaged in aluminum foil of different colors. The heroin sold by Las Avispas at Las Vias generally was packaged in violet-colored foil. The cooperating witnesses also described how the cash proceeds were safeguarded and how the cash was spent, often on luxury items, to disguise its illicit source. Much of the testimony focused on the narcotics activity and violent conduct of members of Las Avispas who were named in the Indictment but who were not on trial. The Government also offered testimony about the seizures of drugs and firearms that were made from many of the absent co-defendants. Although the Government’s witnesses provided a comprehensive description of Las Avispas’ structure and activities, and much testimony was presented relating to the narcotics activities of the thirty-seven absent co-defendants, the quantity and quality of the proof adduced against each of the present Appellants varied markedly. The evidence offered against each Appellant was as follows. Ofray Ofray owned two bars, Rumba’s Pub and La Cota Rota, 2 in Guayama, as well as an upholstery store in Salinas, from which he sold cocaine in “wholesale” quantities, ranging from eighths of a kilogram to full kilograms. Abdul Mendoza-Lebrón (“Mendoza-Lebrón”), a former member of Los Jibaritos who acted as second-in-command of their Las Ruinas drug point but who also sold drugs to Las Avispas in 1997 and 1998, testified that Ofray was a major cocaine supplier in Guayama. Mendoza-Lebrón purchased crack from Ofray for re-sale at Las Ruinas, usually from Ofray’s upholstery store or from Ofray’s residence in Puente Jobo, on a weekly basis from 1994 until Mendoza-Lebrón’s arrest in 1999. Mendoza-Lebrón testified that Of-ray sold large amounts of drugs to Las Avispas, Los Jibaritos and, indeed, “to anybody who would come to buy from him.” Juan Rivera-Rivera (“Rivera-Rivera”), the former leader of the rival Las Jibaritos gang during much of the time period at issue, further testified that Ofray supplied cocaine to drug points run by Las Avispas and that Rivera-Rivera personally purchased cocaine from. Ofray on one occasion, in 1997 or 1998. Carlos Collazo (“Collazo”), another cooperating witness who operated a drug point in the San Felipe ward, between Guayama and Salinas, and frequently associated with and supplied guns to Avispas members, testified that he learned from López-Soto that Ofray was a crack supplier. 3 Angel Villodas (“Villodas”), a cooperating witness who began working as a confidential informant for the FBI in 2000, testified that, on June 8, 2001, he arranged with Ofray to purchase approximately two ounces of cocaine for $1200. Villodas made the purchase on June 14, 2001, at Rumba’s Pub. Although Villodas conferred with Ofray when he first arrived at Rumba’s Pub to complete the transaction, Villo- *13 das testified that it was Ofray’s part-time employee, Appellant Diaz-Clavell, who delivered the cocaine to Villodas in the bathroom of Rumba’s. Villodas also testified that Ofray frequently carried an automatic weapon. In addition, law enforcement officers testified about Ofray’s possession of two handguns. Specifically, in the early morning of July 14, 2002, police officers recovered a .9mm pistol that Ofray deposited in a parked van outside of Rumba’s Pub. Later that day, while arresting Ofray, officers seized a Colt .45 pistol that Ofray carried in a duffel bag. Diaz-Clavell Diaz-Clavell had a full-time job with the Puerto Rico Department of Public Works, but also worked part-time at Ofray’s establishments, Rumba’s Pub and La Copa Rota. Mendoza-Lebrón testified that Diaz-Clavell was very frequently in Ofray’s company and often present when Mendoza-Lebrón and Ofray conducted drug deals. Mendoza-Lebrón also testified that Diaz-Clavell stashed cocaine and cash for Ofray and, on several occasions when Mendoza-Lebrón purchased cocaine from Of-ray, Diaz-Clavell brought the drugs to the apartment above Ofray’s upholstery shop, where the deals were consummated. As stated above, Villodas, who was Diaz-Clavell’s second cousin, testified that, on June 14, 2001, when Villodas went to Rumba’s Pub to execute the cocaine deal that he had pre-arranged with Ofray, it was Diaz-Clavell who delivered the two ounces of cocaine to Villodas, in the bar’s bathroom. López-Soto The Government’s cooperating witnesses, including Mendoza-Lebrón, Rivera-Rivera, and Collazo, testified at length about López-Soto’s control of several drug points, including the Olimpo point in Bor-inquen, from which López-Soto sold cocaine and crack from 1998 to 1998. Mendoza-Lebrón testified that López-Soto regularly carried weapons and acted as an enforcer for Las Avispas from 1993 through 1998, and shot one of the men working for Mendoza-Lebrón around Christmas of 1998. López-Soto conducted frequent drug deals with Las Avispas’ members at drug points in Borinquen, which remained open twenty-four hours each day. López-Soto was both a buyer and seller of narcotics; Rivera-Rivera testified that, on a particular occasion in 1996 or 1997, he sold López-Soto 100 vials of crack, and in 1998, bought one kilogram of cocaine from López-Soto. Collazo testified that he supplied automatic weapons and silencers to López-Soto in 1996 and 1997. López-Soto also was arrested in possession of a .9mm pistol in March 1998. In addition, Collazo testified that López-Soto carried out the murder of Santito, a rival drug dealer, in March 1997. Rivera-Rivera stated that López-Soto admitted to Rivera-Rivera that he had murdered San-tito. Cruz-Pereira Mendoza-Lebrón testified that Cruz-Pereira regularly purchased crack from him and his cousin, Wilson Mendoza-Vás-quez (“Wilson”) — a member of Las Avis-pas and one of the defendants named in the Indictment. Mendoza-Lebrón sold crack to Cruz-Pereira until Mendoza-Le-brón’s arrest in 1999, on those occasions when Wilson did not have crack to sell. In addition to selling crack, Mendoza-Lebrón bought “eighths” of crack from Cruz-Per-eira. The Government also presented evidence of three recorded drug purchases that were made from Cruz-Pereira. On June 21, 2001, Villodas, who had known Cruz-Pereira since their youth, went to Cruz-Pereira’s house on Santa Ana Street *14 and arranged to purchase crack. Cruz-Pereira left in his truck and returned approximately thirty or forty minutes later with approximately fifty vials of crack, which he sold to Villodas in exchange for cash. Agent Edwin Rosa-Ferrer (“Agent Rosa-Ferrer”), an undercover officer working for the Puerto Rico Police Department, was introduced to Cruz-Pereira through an individual named “Christopher.” On November 17, 1998, Agent Rosa-Ferrer went to Cruz-Pereira’s house and arranged to purchase one gram of cocaine for $40. Agent Rosa-Ferrer later went to La Ponderosa, a bar owned by Cruz-Pereira, to complete the transaction. At La Ponderosa, Cruz-Pereira received the cocaine from Cruz-Pereira and handed the $40 to “Christopher.” On December 3, 1998, again at La Ponderosa, Agent Rosa-Ferrer purchased five grams of cocaine from Cruz-Pereira for $150. Zaragozctr-Lasa Zaragoza-Lasa, who was known by the nickname “El Viejo Mode,” was alleged to be a major heroin and cocaine supplier for Las Avispas. Cardona testified that he saw Zaragoza-Lasa, whom he knew only by the name “El Viejo Mode,” at Las Vias on “several occasions” speaking with Javier Báez, the manager of the drug point. On one occasion, in 1998, Cardona saw Zaragoza-Lasa hand to Baez a brick-shaped object wrapped in paper. Cardona testified that he believed the package contained drugs because a runner later brought to Cardona what appeared to be the package’s wrapping, and Cardona was able to make crack out of cocaine residue that was on the paper. Cardona also testified that when drug supplies at Las Vias became depleted, the stock of drugs would increase following “El Viejo Mode” ’s visit. Olga Lebrón-Ortiz (“Lebrón-Ortiz”), a cooperating witness who regularly sold marijuana and cocaine in the Borinquen ward in Guayama in 1994 and 1995 and later made undercover drug purchases for the FBI, saw Lasa — whom she also knew only by the name “El Viejo Mode” — in Borinquen regularly from 1995 through 2002. Zaragoza-Lasa appeared in Borin-quen up to twice a week. Lebrón-Ortiz stated that she saw Zaragoza-Lasa deliver to Javier Baez a transparent bag containing violet-colored aluminum foil packages. Lebrón-Ortiz corroborated other testimony that established that heroin was regularly packaged for sale by Las Avispas at Las Vias in violet-colored aluminum foil. Lebrón-Ortiz also testified that, on two occasions in 2002, she saw Zaragoza-Lasa give free samples, or “tastes,” of heroin to Borinquen street junkies. “Overview” Testimony In addition to the aforementioned testimony from cooperating witnesses and an undercover officer, the Government offered the testimony of two lead case agents, Agent Ricardo Rivera and Agent José Tirado, who were both members of the Puerto Rico Police Department assigned to the joint FBI task force investigating narcotics and weapons activity in Guayama and Salinas. Agent Rivera, the Government’s first witness, provided a general overview of the pervasive, violent drug activity in the Guayama and Salinas region, identifying eleven different drug points that were controlled by seven gangs. In addition to testifying about the general hierarchy among the workers at the drug points and the way in which drug points typically operated, Rivera-over defense counsel’s frequent objections-summarized the role played by each of the five Appellants in Las Avispas. Agent Rivera stated that Ofray dealt large quantities of drugs from his bars and upholstery shop, carried weapons regularly, and was a friend and drug supplier to José Cabezón, *15 Las Avispas’ leader. Agent Rivera described Díaz-Clavell as Ofray’s assistant and runner, who regularly delivered drugs to Ofray. Rivera also discussed López-Soto’s role as an enforcer for Las Avispas, and his frequent carriage of weapons. Agent Rivera identified Cruz-Pereira as the leader of the Santa Ana drug point in the Puente Jobo ward. Finally, Agent Rivera stated that Zaragoza-Lasa delivered drugs regularly to Las Avispas and, in particular, supplied heroin to the Las Vias drug point. Agent Rivera did not, however, testify that he had any personal knowledge of the Appellants’ alleged narcotics activity. Rather, he derived his conclusions “through information that we obtained from informants who collaborated in the investigation and informants of the Puerto Rico Police Department and state policemen as well.” Agent Tirado provided similar “overview” testimony over defense counsel’s objections later in the trial. Like Agent Rivera, Agent Tirado explained the role each Appellant played in the Las Avispas drug organization. Agent Tirado described Ofray as a Las Avispas cocaine supplier who operated the San Felipe drug point in Salinas (that is, Ofray’s upholstery store); Díaz-Clavell as Ofray’s runner; López-Soto as a Las Avispas “hit man” and owner of a drug point; Cruz-Pereira as the manager of the Santa Ana drug point in the Puente Jobo ward; and Zara-goza-Lasa as one of Las Avispas’ suppliers, and “the main supplier of heroin at the Pales Matos drug points in the Borin-quen ward, in Guayama and other drug points” who was known to supply heroin in violet-colored aluminum foil packets. While testifying about each Appellant’s role in the conspiracy, Agent Tirado referred to a chart that contained the names and alleged positions occupied in the conspiracy by the Appellants and the other defendants named in the indictment. Like Agent Rivera, Agent Tirado conceded that his information was obtained solely through second-hand information given to him by informants and “through the investigation that we conducted.” DISCUSSION The Appellants assert numerous individual claims on appeal. One claim, which initially was raised only by Cruz-Pereira, arises from the district court’s allegedly improper handling of a jury note submitted after the close of evidence. In its written response to the note, the court imparted to the jury information that was extrinsic to the evidence presented at trial and potentially prejudicial to all five Appellants. Because the district court’s response to the jury note arguably affected the jury’s verdict as to all five Appellants, we deal first with this issue before addressing Appellants’ many individual challenges. The Jury Note On September 29, 2003, after the jury received final instructions and began deliberations, the district court told both the prosecutors and defense counsel that, if the court received a note from the jury, it would notify all counsel, and that counsel were required to arrive within ten minutes after being notified. The trial judge warned that if counsel failed to appear within ten minutes of the notification, he would respond to the jury’s note without input from anyone who was absent. The record indicates that the court subsequently received four jury notes, including the final note, which stated that the jury had reached a verdict. Two notes, the contents of which are irrelevant for purposes of this appeal, were marked as “Jury Note # 1” and “Jury Note # 3”, and were signed by the foreperson, respectively, at 5:55 p.m. and 6:00 p.m. The note *16 which concerns us was marked as “Juror Note # 2” (“Note # 2”), and was signed by the foreperson at 6:15 p.m. 4 Note # 2 was written in Spanish and headed with the single, underlined word “Duda,” which is translated as “Doubt.” The district court responded to the note, in English, as follows: Dear Jury: The question you posed, Those that are imprisoned, that is the others in the Indictment, are they in jail for a conspiracy? The answer is yes. s/Juan M. Pér-ez-Giménez, U.S.D.J., 9-29-03. This written answer was then sent to the jury. The final note from the jury, stating that the jury had reached a verdict, was signed by the foreperson at 7:15 p.m. As stated, Cruz-Pereira was the sole Appellant to assert a claim relating to the jury note. Cruz-Pereira claimed that the trial court’s handling of the note violated Rule 43(a) of the Federal Rules of Criminal Procedure, which requires the presence of the defendant at every stage of the trial, and that the substance of the district court’s response to the note resulted in great prejudice. The Government argued, first, that the district court’s divergence in this case from the standard practice of handling the jury note was due to defense counsel’s failure to follow the judge’s reasonable ten-minute rule; second, that the record indicated that defense counsel were in fact present when the district court drafted its proposed answer to Note #2 and did not object to the answer; and third, even if the court improperly handled the note, that the error was harmless in light of the overwhelming evidence against Cruz-Pereira. Because it is undisputed that the district court informed the jury of a fact that was never offered in evidence — namely, that thirty-seven non-testifying co-defendants were incarcerated for a conspiracy 5 — and because that fact had the potential of tainting the jury’s verdict with respect to all of the Appellants, we exercised our discretion to review the jury note issue as it applied to all five Appellants, even though only Cruz-Pereira initially raised the claim on appeal. See United States v. Riverar-Rosario, 300 F.3d 1, 10 n. 1 (1st Cir.2002) (“In exceptional cases, especially in criminal cases, appellate courts, in the public interest, may, of their own motion, notice errors to which no exception has been taken, if the errors are obvious, or if they otherwise affect the fairness, integrity or public reputation of judicial proceedings.”) (quoting United States v. Atkinson, 297 U.S. 157, 160, 56 S.Ct. 391, 80 L.Ed. 555 (1936) (internal quotation marks omitted)). Our initial review of the procedural component of the district court’s handling of the jury note was somewhat hampered by the state of the trial transcript. As all the parties agree, the portion of the transcript that relates to the district judge’s receipt of the jury notes and his drafting of the answer to Note # 2, in particular, contains gaps and omissions. Because of the incomplete transcription, it was unclear which defense counsel, if any, were present when the judge discussed his proposed response to Note # 2. It was equally un *17 clear whether defense counsel were given an opportunity to object to the proposed response or whether counsel in fact voiced any objections. Unsurprisingly, the briefs submitted by the Government and Appellant Cruz-Pereira contained differing interpretations of the incomplete record. Counsel’s statements at oral argument did not help to clarify the circumstances under which the district court responded to Note #2. Accordingly, we directed the parties to submit supplemental briefing on the issue of whether the district court’s handling of the jury note was erroneous, as well as affidavits from the trial attorneys, setting forth relevant facts related to the trial judge’s handling of the jury note, in an attempt to reconstruct the record. In addition, we directed the parties to address the substantive issue of prejudice that may have resulted, as to each Appellant, from the court’s answer to Note # 2. The affidavits and supplemental briefing submitted by the parties lead us to conclude that the procedure followed by the court in handling the jury note was not erroneous. In United States v. Maraj, 947 F.2d 520 (1st Cir.1991), we set forth the proper procedure for a trial court’s handling of a jury note: The preferred practice for handling a jury message should include these steps: (1) the jury’s communique should be reduced to writing; (2) the note should be marked as an exhibit for identification; (3) it should be shown, or read fully, to counsel; and (4) counsel should be given an opportunity to suggest an appropriate rejoinder. If the note requires a response ore tenus, the jury should then be recalled, the note read into the record or summarized by the court, the supplemental instructions given, and counsel afforded an opportunity to object at side-bar. If, however, the note is to be answered in writing, the court’s reply should be marked as an exhibit for identification, the judge should read both the jury’s note and the reply into the record, and counsel should be afforded an opportunity to register objections before the reply is transmitted to the jury. Id. at 525. In addition, “it is also helpful for the judge to inform counsel of the substance of [the] proposed response.... ” Id. (quoting United States v. Ronder, 639 F.2d 931, 934 (2d Cir.1981) (internal quotation marks omitted)). The rules for handling a jury note that are set forth in Maraj are well-settled. See United States v. Parent, 954 F.2d 23, 25 (1st Cir.1992) (stating that Maraj drew upon “long-settled precedent” and “stands foursquare for the proposition that messages from a deliberating jury, pertaining to ongoing deliberations, ought to be fully disclosed to the lawyers when received, so that the latter may be heard before the judge implements a course of action”). A judge’s responding to a jury note outside the presence of counsel and defendant also violates Rule 43 of the Federal Rules of Criminal Procedure, which states that the stages of a trial at which the defendant must be present include “every trial stage, including jury impanelment and the return of the verdict[.]” Fed.R.Crim.P. 43(a)(2). The affidavits submitted by the parties lead us to conclude that all parties were present at the time that the district court drafted a proposed answer to Note #2 and that the note was discussed with counsel. The affidavits received from trial counsel for López-Soto (Attorney Inserni), Zaragoza-Lasa (Attorney Cruz), and Cruz-Pereira (Attorney Dolz) all state that the attorneys do not recall whether they were present when the trial judge drafted his answer to Note # 2. The attorney who represented Ofray and Diaz-Clavell jointly at trial (Attorney Diaz) did not submit an *18 affidavit. By contrast, the Government’s lead prosecutor at trial, Assistant United States Attorney (“AUSA”) Irene Feldman, offers a clear recollection of the events at issue. She states in her affidavit that she “quite definitely remember[s] the presence of all attorneys in the crowded room when the notes were read and the answers were discussed and drafted by the court.” In particular, AUSA Feldman states that she is “absolutely certain that the second question [Note # 2] was discussed in the presence of all parties prior to the court’s having drafted and filed a response.” In light of defense counsel’s collective uncertainty and the Government’s definitive recollection that all parties were present when the district court drafted its response to Note # 2, we find that there was no procedural error in the district court’s handling of the jury note. It is evident from the record that Note #2 was reduced to writing, and equally evident from the Government’s affidavit that the note was shown to and/or discussed with defense counsel, and that counsel thus had the opportunity to object to the proposed answer. 6 We turn next to the issue of whether the trial judge’s response to Note #2, resulting in the jury’s exposure to potentially prejudicial extrinsic information, constituted substantive error. The Sixth Amendment requires that the jury’s verdict must be based solely upon the evidence developed at trial. See Turner v. Louisiana, 379 U.S. 466, 472, 85 S.Ct. 546, 13 L.Ed.2d 424 (1965). “The theory of our system is that the conclusions to be reached in a case will be induced only by evidence and argument in open court, and not by any outside influence, whether of private talk or public print.” Patterson v. Colorado, 205 U.S. 454, 462, 27 S.Ct. 556, 51 L.Ed. 879 (1907) (Holmes, J.). “[Exposure to extrinsic information deprives a criminal defendant of the protections of the Sixth Amendment, including his right of confrontation, of cross-examination, and of counsel.” United States v. Santana, 175 F.3d 57, 65 (1st Cir.1999) (citations and footnote omitted); see also id. (finding that “the judge should not have intervened by allowing the jury to observe [extrinsic evidence] after the close of evidence and without the standard safeguards of a criminal trial” and vacating conviction); Lacy v. Gardino, 791 F.2d 980, 983 (1st Cir.1986) (finding that defendant’s Sixth Amendment rights to confrontation and cross-examination were violated when juror peeled tape off exhibits unmasking information concerning defendant’s prior criminal record); United States v. Hans, 738 F.2d 88, 93 (3d Cir.1984) (trial court committed reversible error by granting jury’s request to examine objects not in evidence, after deliberations had begun). The jury’s exposure to extrinsic facts is especially troubling when the trial judge is the source of the information. A district court must use extreme caution in answering questions from juries so as not to usurp the jury’s fact finding role. As we have noted on many occasions “undeniably inherent in the constitutional guarantee of trial by jury is the principle that a court may not step in and direct a finding of contested fact in favor of the prosecution regardless of how overwhelmingly the evidence may point in that direction.” *19 United States v. Sabetta, 373 F.3d 75, 80 (1st Cir.2004) (quoting United States v. Rivera-Santiago, 107 F.3d 960, 965 (1st Cir.1997)). “The Constitution casts judge and jury in mutually supporting—yet nevertheless distinct—roles.... ‘The trial judge is ... barred from attempting to override or interfere with the jurors’ independent judgment in a manner contrary to the interests of the accused.’ ” United States v. Argentine, 814 F.2d 783, 788 (1st Cir.1987) (quoting United States v. Martin Linen Supply Co., 430 U.S. 564, 573, 97 S.Ct. 1349, 51 L.Ed.2d 642 (1977)). Here, the jury requested, and the court affirmatively provided, undisputably extrinsic information. There was simply no evidence presented at trial relating to the current status of any of the thirty-seven absent co-defendants named in the Indictment. None of the thirty-seven co-defendants testified at trial, and no mention was made of whether any of the thirty-seven had been imprisoned or found guilty of the charges in the indictment. Thus, the jury’s note itself suggests that the jury had been exposed to the extrinsic “fact” that the absent co-defendants were incarcerated. More crucially, the judge’s response to Note #2, in which he confirmed for the jury that the thirty-seven co-defendants who did not appear at trial were in prison for their participation in a conspiracy, constituted new evidence, delivered to the jury from the bench rather than the witness stand, and unaccompanied by any of the safeguards of a criminal trial, in violation of Appellants’ Sixth Amendment rights. See Turner, 379 U.S. at 472-73, 85 S.Ct. 546 (“Trial by jury in a criminal case necessarily implies at the very least that the ‘evidence developed’ against a defendant shall come from the witness stand in a public courtroom where there is full judicial protection of the defendant’s right of confrontation, of cross-examination, and of counsel.”). Thus, the jury’s exposure to extrinsic information amounts to an error of constitutional dimension. Having found error, we must next select the appropriate standard for our review. Normally, in the absence of any objection to the answer that the district court proposed and submitted to the jury, we would review for plain error. The gaps in the trial transcript prevent us, however, from ascertaining whether defense counsel asserted timely objections to the court’s proposed response. The affidavits submitted in supplemental briefing do not provide much illumination. The trial attorneys for López-Soto, Zaragoza-Lasa, and Cruz-Pereira all state that, although they are unsure whether they were present when the note was discussed, they would have certainly objected if the trial judge had stated that he intended to issue a written affirmative response to Note #2. As Attorney Cruz states, “I don’t remember any type of discussion in relation to the answer to jury note number 2. I am familiar with the experience of the defense attorneys involved in this case and I believe at least one, if not all of us, would have objected to the wording contained in the answer to jury note number 2 and would have requested that the jury be advised that they were only to concern themselves with those defendants who were at trial.” AUSA Feldman’s affidavit contains only the oblique suggestion that defense counsel failed to voice objections. Regarding defense counsel’s reaction to the court’s proposed answer to Note #2, Feldman states, “I recall that defense counsel responded to the jury note and proposed response as if it were favorable to the defendants.” Feldman does not affirmatively state that she recalls that there were no objections asserted. Moreover, the Government does not claim, in its supplemental brief, that defense counsel failed to *20 object to the district court’s proposed response. Further, Feldman’s recollection that defense counsel responded to the district court’s answer as if it were “favorable” is contradicted by the record. The transcript, though incomplete, indicates that at least one attorney raised a concern about the trial court’s answer. After the court read aloud, in English, both the question posed in Note # 2 and the answer that was given by the court, the transcript indicates that Attorney Diaz stated, “But they should know that they didn’t go to trial.” The court responded as follows: “I didn’t tell them that they didn’t go to trial. The question you posed, those that are in prison; that is the others that are in the indictment, are they in jail for conspiracy? The answer is yes.” Thus, it appears that Attorney Diaz (who, as noted, did not provide an affidavit) did go on record to express a reservation about the propriety of the court’s response. In sum, the affidavits submitted by counsel and our review of the incomplete transcript do not lead us to find that defense counsel failed to object to the trial court’s proposed response to the jury note. Rather, the record indicates that some objection was in fact asserted and thus weighs against plain error review. Ordinarily, “wherever material uncertainties result from an incomplete or indecipherable record and impede or affect our decision, we resolve such uncertainties against appellants.” Credit Francais International, S.A. v. Bio-Vita, Ltd., 78 F.3d 698, 700-01 (1st Cir.1996); see also Real v. Hogan, 828 F.2d 58, 60 (1st Cir.1987) (“[I]t is the appellant who must bear the brunt of an insufficient record on appeal.”). The rule, however, is applied typically in cases where an appellant has failed to provide an adequate record. In this instance, however, Appellants are not to blame for any ambiguities or omissions in the record. We have been provided with the necessary portions of the trial transcript. It is the transcript itself that is deficient. Mindful that “[a] criminal defendant has a right to a record on appeal which includes a complete transcript of the proceedings at trial,” United States v. Carrillo, 902 F.2d 1405, 1409 (9th Cir.1990) (citing Hardy v. United States, 375 U.S. 277, 279-82, 84 S.Ct. 424, 11 L.Ed.2d 331 (1964)), we will not resolve the transcript’s inadequacies against the Appellants. Where, as here, we deem a claim of jury contamination to be preserved, we are faced with two potential standards of review: that of abuse-of-discretion or that of harmless error. In the majority of our cases that have involved claims that a jury was improperly exposed to extrinsic information, our review has been for abuse of discretion. These cases involve the jury’s accidental exposure to potentially prejudicial material that was not offered in evidence at trial. See, e.g., United States v. Bradshaw, 281 F.3d 278 (1st Cir.2002) (un-redacted version of indictment accidentally left in jury room during deliberations); United States v. Gomes, 177 F.3d 76 (1st Cir.1999) (indictment from prior trial of defendant left in jury room); United States v. Boylan, 898 F.2d 230 (1st Cir.1990) (magazine containing potentially prejudicial article about defense counsel accidentally left in jury room). In each of these cases, the colorable claim of jury contamination was brought to light either before or after the rendering of a verdict, the trial judge conducted an inquiry to discern whether the jurors were in fact prejudiced by their exposure to outside material, and where appropriate the judge employed remedial measures, such as issuing curative instructions, see Bradshaw, 281 F.3d at 291, or dismissing potentially prejudiced jurors, see Gomes, 177 F.3d at *21 82. In Bradshaw, we explained that a trial court, when confronted with a claim of jury contamination, has broad discretion to “fashion an appropriate procedure for assessing whether the jury has been exposed to substantively damaging information, and if so, whether cognizable prejudice is an inevitable and ineradicable concomitant” of the jury’s exposure to improper extrinsic information. 281 F.3d at 290. Where the jury’s contact with outside information is inadvertent and not accompanied by “egregious circumstances,” and the trial judge responds to the claim of contamination by conducting an inquiry and employing remedial measures, we do not presume that the jury’s exposure to extrinsic material resulted in prejudice. Id. at 288. Instead, we review the trial judge’s actions for abuse of discretion. See United States v. Yeje-Cabrera, 430 F.3d 1, 10 (1st Cir.2005). In Bradshaw, however, we declined to hold that a presumption of prejudice could never be applied: “We leave for another day the question of whether a jury’s exposure to substantively damaging information may sometimes occur under circumstances so aggravated as to warrant the application of the ... presumption [of prejudice] even without deliberate misconduct (and if so, what those circumstances might comprise). That question simply is not presented here.” Bradshaw, 281 F.3d at 288 n. 5. Bradshaw recognized that Santana was one of those eases. In Santana, 175 F.3d at 57, we held that the jury’s exposure to outside information required the application of a presumption of prejudice. In Santana, as in this case, we confronted the “unusual” situation where the jury was exposed to extrinsic information after the close of evidence and during its deliberations, not as a result of inadvertence or juror misconduct, but rather because of the trial judge’s “approval of the jury’s request to consider information outside the record.” Id. at 65. In Santana, the trial judge permitted the jury to file back into the courtroom after the close of evidence and observe the defendant’s ears, which had remained covered throughout the trial by headphones used for the Spanish translation. A contested issue at trial was whether the defendant was in fact the individual who had participated in a drug deal that had been surveilled by a government agent; in identifying the defendant as a participant in the deal, the agent remarked on the fact that the defendant’s ears were oddly protuberant. After observing the defendant without headphones, the jury returned to its deliberations and subsequently returned a guilty verdict. There, as here, the trial judge did not conduct any inquiry into possible prejudice resulting from the jury’s exposure to extra-record evidence, nor did the court issue curative instructions or undertake any other remedial measures. Under those circumstances, we found that “the court’s decision to allow the jury to consider extrinsic information is .... subject to de novo review and it is error per se.” Id. We find that review for abuse of discretion is inappropriate in this case. Here, the jury’s exposure to extrinsic material did not occur inadvertently: the undisput-ably extrinsic information was supplied to the jury by the trial court itself, in response to the jury’s note, and thus essentially was offered to the jury as evidence. The judge, who served as the source of the extrinsic information, conducted no inquiry as to the prejudice that may have resulted from the jury’s receipt of the off-record fact and did not undertake any remedial measures. Thus, there was no action undertaken by the court that we can now review. Abuse of discretion is simply a poor fit. *22 We conclude instead that the presumption of prejudice that we applied in Santana is also appropriate in this case. Here, as in Santana, the jury actively sought and received extrinsic information from the trial judge. Here, too, as discussed below, the information was prejudicial, was probably used by the jury as evidence of the Appellants’ guilt, and was unaccompanied by an instruction or any other curative undertaking by the district court. Accordingly, this case, like Santana, presents sufficiently aggravated circumstances that a presumption of prejudice is warranted. Because the jury’s exposure to extrinsic factual information in this case raises a presumption of prejudice, “the government bears the burden of showing beyond a reasonable doubt that the extrinsic information did not contribute to the conviction.” Id. at 66 (citations omitted); see also Lacy, 791 F.2d at 983. An error will be deemed harmless if “the beneficiary of ... [the] constitutional error [can] prove beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained.” Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967); see also Santana, 175 F.3d at 66. In Sullivan v. Louisiana, 508 U.S. 275, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993), the Supreme Court explained that the harmless error inquiry is “not whether, in a trial that occurred without the error, a guilty verdict surely would have been rendered, but whether the guilty verdict actually rendered in this trial was surely unattributable to the error.” Id. at 279, 113 S.Ct. 2078. To determine whether the jury’s exposure to extrinsic evidence was harmless, a reviewing court must “assess the record as a whole to determine the impact of the improper evidence upon the jury-The prejudicial effect of the improper evidence must be weighed against the weight of the properly admitted evidence.” Lacy, 791 F.2d at 986 (quoting Morgan v. Hall, 569 F.2d 1161, 1166 (1st Cir.1978) (internal quotation marks omitted)); see also United States v. Weiss, 752 F.2d 777, 783 (2d Cir.1985) (possibility of prejudice is assessed “by reviewing the entire record, analyzing the substance of the extrinsic evidence, and comparing it to that information of which the jurors were properly aware”). The inquiry regarding harmless error analysis “is case-specific” and requires consideration, among other factors, of “the centrality of the tainted evidence, its uniqueness, its prejudicial impact, the use to which the evidence was put, and the relative strength of the parties’ cases.” United States v. Garcia-Morales, 382 F.3d 12, 17 (1st Cir.2004). We first examine “the nature of the extrinsic information to which the jury was exposed to determine its potential prejudicial effect.” Santana, 175 F.3d at 66. Second, we analyze the weight of the trial evidence properly adduced against each of the five Appellants. See id. Nature of the Prejudicial Information Both the information conveyed to the jury and the manner of its conveyance were prejudicial. The district court’s confirmation that the thirty-seven defendants named in the Indictment who did not appear at trial were all incarcerated for participation in the conspiracy was tantamount to an announcement to the jury that the thirty-seven absent defendants had been convicted of the charges contained in the Indictment. The Government concedes, as it must, that “[t]he fact that other co-defendants are in prison for the same conspiracy” is “irrelevant to the guilt of those on trial.” Courts in this and other circuits have strongly cautioned against the admission of such evidence. “[A] defendant is entitled to have the question of his guilt determined upon the evidence *23 against him, not on whether a codefendant or government witness has been convicted of the same charge.” United States v. Dworken, 855 F.2d 12, 30 (1st Cir.1988) (quoting United States v. Miranda, 593 F.2d 590, 594 (5th Cir.1979) (internal quotation marks omitted)). The potential for prejudice is present where evidence of a co-conspirator’s conviction is admitted for substantive purposes. See United States v. Blevins, 960 F.2d 1252, 1260-62 (4th Cir.1992). Under such circumstances, the jury may abdicate its duty and “regard the issue of the remaining defendant’s guilt as settled and the trial as a mere formality.” United States v. Griffin, 778 F.2d 707, 711 (11th Cir.1985). As the Fifth Circuit, reversing a conviction on the grounds that the jury was told that a non-testifying co-defendant had pleaded guilty, explained in United States v. Hansen, 544 F.2d 778 (5th Cir.1977): “There is no need to advise the jury or its prospective members that some one not in court, not on trial, and not to be tried, has pleaded guilty. The prejudice to the remaining parties who are charged with complicity in the acts of the self-confessed guilty participant is obvious.” Id. at 780. Regardless of whether an absent co-defendant has pleaded guilty or been convicted after trial, the admission of such evidence not only results in the danger that the jury will improperly infer guilt by association, it also “significantly undercuts the defendant’s right to have a jury’s verdict based only upon evidence that is presented in open court and is thereby subject to scrutiny by the defendant.” Blevins, 960 F.2d at 1260. Thus, “where a missing co-defendant does not testify, ‘it is generally accepted that absent agreement, courts and prosecutors generally are forbidden from mentioning that a co-defendant has either pled guilty or been convicted.’ ” United States v. Carraway, 108 F.3d 745, 756 (7th Cir.1997) (quoting United States v. Johnson, 26 F.3d 669, 677 (7th Cir.1994) (internal quotation marks omitted)). Where, as here, the defendants are being tried for their participation in an alleged conspiracy that took place over the course of nearly a decade, the danger that the defendants will be found guilty by sheer association with guilty non-testifying co-defendants is great. See United States v. Izzi 613 F.2d 1205, 1210 (1st Cir.1980) (“Guilt by association is one of the ever present dangers in a conspiracy count that covers an extended period.”) (citing Kotteakos v. United States, 328 U.S. 750, 774-75, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)). Justice Jackson, in his well-known concurrence in Krulewitch v. United States, 336 U.S. 440, 69 S.Ct. 716, 93 L.Ed. 790 (1949), reflected on the abnormally high risk to co-defendants in a conspiracy trial of being found guilty by association: A co-defendant in a conspiracy trial occupies an uneasy seat. There generally will be evidence of wrongdoing by somebody. It is difficult for the individual to make his own case stand on its own merits in the minds of jurors who are ready to believe that birds of a feather are flocked together. Id. at 454, 69 S.Ct. 716 (Jackson, J., concurring). Although a jury “may wonder what happened to the co-defendants whose names have been mentioned in the indictment and in the course of the trial but who have not appeared before them,” the preferred course is simply to instruct the jury “not to concern itself with that question.” Carraway, 108 F.3d at 756; see also Edward J. Devitt, Charles B. Blackmar, et al., Federal Jury Practice and Instructions § 5.02, at 106 (1992). The district court’s announcement to the jury that thirty-seven absent co-defendants named in the Indictment were *24 incarcerated for the charged conspiracy increased by an order of magnitude the risk that the Appellants would be found guilty because of their association with the absent alleged co-conspirators, rather than because of the evidence offered at trial. Quite simply, the extrinsic information allowed the jury to draw the prejudicial inference that, if thirty-seven non-testifying co-defendants had been found guilty of the crimes described in the Indictment, the Appellants also must be guilty. 7 Moreover, the court did not instruct the jury that the incarcerated status of the thirty-seven absent co-defendants, or the fact that they may have been found guilty, should have no bearing on the jury’s verdict with respect to each of the defendants on trial. See United States v. Riverar-Santiago, 872 F.2d 1073, 1083 (1st Cir. 1989). Absent the safeguard of a proper instruction, the prejudice arising from the court’s response to the note is even more serious. The Government claims that the district court’s answer was not especially prejudicial because “the jury note itself illustrates the jury already had knowledge that the other co-defendants were imprisoned, as it begins with the phrase, ‘those that are imprisoned.’ ” The Government’s argument is unavailing. To the extent that the note reveals that the jury knew that the absent co-defendants were imprisoned, as discussed above, such awareness was not gleaned from evidence presented at trial. The fact that the jury demonstrated its knowledge of an undisputably extrinsic fact does not cut against our evaluation of the prejudicial impact of the jury note and the district court’s response. As stated, the Sixth Amendment requires that a jury’s verdict must be based solely upon the evidence developed at trial, see Turner, 379 U.S. at 472, 85 S.Ct. 546, and “[t]he jury’s exposure during its deliberations to extrinsic information, whatever its source, is an error of constitutional proportions .... ” Santana, 175 F.3d at 65 (emphasis added). Moreover, the Government cannot, and does not, attempt to deny that the most prejudicial extrinsic information in fact came from the trial judge — namely, confirmation that the thirty-seven alleged co-conspirators were imprisoned for the conspiracy. It was this information that most readily permitted the jury to draw the impermissible inference of guilt by association. We also find that the manner in which the jury requested the extrinsic information suggests that a direct connection existed between the extrinsic evidence and the jury’s verdict. The fact that the note was headed by the word “Duda,” or “Doubt,” strongly indicates that the jury may have sustained doubt about some or all of the Appellants’ involvement in the charged conspiracy and was looking to the trial judge to resolve that doubt. See Rivera-Santiago, 107 F.3d at 966 (finding it “significant,” in determining that district court had exposed jury to extrinsic information, that jury was asking for judge to provide portion of record to “ ‘clarify some doubts’ ”). That the jury actively sought the information that it received, rather than obtained it accidentally or in an otherwise unsolicited fashion, further indicates that the answer was important to its *25 verdict. See Santana, 175 F.3d at 67 (“[B]eeause the jurors specifically asked to observe Santana without his headphones, they obviously deemed such evidence important to their deliberations.”). It is also important, in considering prejudice, that the answer to the jury’s question was supplied by the trial judge, and thus stamped with the imprimatur of the court, rather than by comparatively less authoritative sources, such as prosecutorial comment, juror misconduct, or the inadvertent admission of extrinsic evidence into the jury room. See id.; Argentine, 814 F.2d at 788. Finally, the fact that the jury returned with a verdict a mere forty-five minutes after receiving the answer to Note # 2, which was the last note the jury sent out prior to the announcement that it had reached a verdict, supports the inference that the jury attributed weight to the trial judge’s response, and indeed considered the court’s response to be important, if not critical, in arriving at the verdict. See Rogers v. United States, 422 U.S. 35, 40, 95 S.Ct. 2091, 45 L.Ed.2d 1 (1975) (noting that jury’s returning with a verdict “five minutes” after receiving judge’s response to query weighed in favor of finding that judicial response to note was harmful error); Rivera-Santiago, 107 F.3d at 967 (drawing inference that judge’s improper response to jury influenced verdict where jury returned verdict two hours after receiving judge’s answer and noting that timing of jury’s response to judge’s answer was factor to consider in harmless error analysis). Thus, although the “fact” of the absent co-defendants’ incarceration was neither raised nor disputed at trial, and although it was not proper for the jury to consider such a fact in reaching a verdict, it was clearly information that was material to the jury’s verdict. See Santana, 175 F.3d at 67 (finding that “the connection between the extrinsic information at issue here—the appearance of Santana’s ears—and an issue material to and disputed throughout the trial—the identity of the [supplier of] crack cocaine—is unmistakable”). In sum, the nature of the extrinsic information received by the jury was prejudicial to the Appellants. Weight of the Evidence Because the relative strength of the Government’s case varied as to each Appellant, we consider each of the Appellants in turn, in light of the court’s response to Note # 2. López-Soto The error was harmless as to Ló-pez-Soto. The evidence clearly established that López-Soto was a member of the Avispas conspiracy and, more particularly, an enforcer for the organization who regularly carried guns and engaged in violence. Juan Rivera-Rivera, who lived near López-Soto in Marín and frequently associated with members of Las Avispas before the Avispas and Los Jibaritos became embroiled in a war over territory, provided ample, detailed testimony about López-Soto’s ownership of drug points at Olimpo, Las Palmas de Arroyo, Marin, and El Flamboyan. The evidence showed that López-Soto sold crack and cocaine from Olimpo from 1993 through 1998. Rivera-Rivera testified about specific drug deals in which he engaged with López-Soto, including the sale of 100 vials of crack to López-Soto in 1996 or 1997 and the purchase from López-Soto of one kilogram of cocaine in 1998. Rivera-Rivera identified specific individuals who assisted López-Soto in drug dealing, including Freddie El Agente and “Cabe,” in Olimpo. Further testimony established that Ló-pez-Soto acted as an enforcer for Las Avispas who regularly carried and used guns. Testimony was presented about Ló-pez-Soto’s participation in specific acts of *26 violence, including his participation in an attempted contract “hit” of a rival drug dealer, with Rivera-Rivera, and his murder of rival dealer Santito in 1997. Collazo also provided detailed testimony about Ló-pez-Soto’s narcotics activity, including his purchase of large quantities of cocaine which López-Soto subsequently cooked into crack. In addition, Collazo testified about specific weapons, including automatic weapons equipped with silencers, that he sold to Zaragoza-Lasa, thus supporting the Government’s theory that López-Soto acted as an enforcer for Las Avispas. In light of the Government’s strong evidence of López-Soto’s participation in the charged conspiracy, and the comparative weakness of the defense’s case, we conclude that the guilty verdict was “surely unattributable” to the district court’s answer to the jury note. Sullivan, 508 U.S. at 279,113 S.Ct. 2078. Ofray The Government’s case against Ofray was overwhelming. Mendoza-Le-brón testified at length that Ofray supplied large-unit quantities of cocaine to Las Av-ispas on a weekly basis from 1994 through 1999. There was ample testimony establishing that Ofray frequently sold narcotics from his businesses in Salinas and Guaya-ma, which included an upholstery store, Rumba’s Pub, and La Copa Rota. Mendoza-Lebrón testified that he frequently contacted Ofray by cellular phone to arrange purchases and designate a meeting place, and that Ofray would subsequently meet Mendoza-Lebrón at the designated place and there receive drugs from Ofray in exchange for cash. Rivera-Rivera also testified about Ofray’s sale of cocaine to Las Avispas and his purchase from Ofray of nearly one kilogram of cocaine in 1997 or 1998. In addition, Villodas testified in detail about the purchase of cocaine that he prearranged with Ofray on June 8, 2001, in which he agreed to buy two ounces of cocaine for $1200. Villodas provided further testimony about the execution of the sale on June 14, 2001, in which Villodas— while wearing a monitoring device — went to La Copa Rota, as arranged, met with Ofray, and shortly thereafter received the cocaine. Finally, law enforcement testimony established that a ,9mm pistol belonging to Ofray was recovered from outside Rumba’s Pub and a Colt .45 pistol was seized from Ofray’s duffle bag at the time of his arrest on July 14, 2002. As with López-Soto, the strength of the government’s case leads us to conclude that the extrinsic information to which the jury was exposed did not sway the verdict as to López-Soto and was thus harmless beyond a reasonable doubt. Cruz-Pereira The district court’s error also was harmless beyond a reasonable doubt as to Cruz-Pereira. There was convincing evidence of Cruz-Pereira’s participation in three pre-arranged drug buys. Agent Rosa-Ferrer purchased cocaine on two separate occasions from Cruz-Pereira, on November 17, 1998 and December 3, 1998, at Cruz-Pereira’s bar, La Ponderosa. Angel Villodas, wearing a monitoring device, purchased vials of crack from Cruz-Per-eira on June 21, 2001, in the backyard of Cruz-Pereira’s residence in Santa Ana. Mendoza-Lebrón also testified that Cruz-Pereira purchased crack from Mendoza-Lebrón’s cousin, Wilson — a member of Las Avispas — and from Mendoza-Le-brón himself in large quantities and on -a weekly basis until Mendoza-Lebrón’s arrest in 1999. Mendoza-Lebrón also stated that he purchased crack from Cruz-Per-eira. *27 Because the evidence of Cruz-Pereira’s drug dealing was strong, if not overwhelming, and was supported by additional evidence to show that Cruz-Pereira engaged in narcotics activity with non-government actors, including Mendoza-Lebrón and Wilson, we find that the district court’s erroneous answer to the jury note was harmless beyond a reasonable doubt as to Cruz-Pereira. Diaz-Clavell By contrast, the evidence presented against Diaz-Clavell, while legally sufficient to support his conviction for participation in the conspiracy, was not strong enough to persuade us beyond any reasonable doubt that the jury’s verdict of guilt was surely unattributable to the court’s response to Note # 2. See Santana, 175 F.3d at 67. The case against Diaz-Clavell relied exclusively on the testimony of cooperating witnesses. See United States v. Bosch, 584 F.2d 1113, 1123 (1st Cir.1978) (finding government’s case “not overwhelming” and constitutional error harmful, in part, because “[t]he government’s case consisted primarily of the testimony of admitted accomplices, whose credibility was attacked”). Although it was undisputed that Diaz-Clavell worked as a part-time employee at Ofray’s businesses, and the Government presented substantial evidence to establish that Ofray was a major cocaine supplier to Avispas, proof of Diaz-Clavell’s narcotics activity was relatively scant. Inculpating testimony was limited to the testimony of Mendoza-Lebrón, who stated that Diaz-Clavell assisted Ofray in his drug dealing by stashing cocaine and cash; and the testimony of Villodas, who testified that Diaz-Clavell delivered approximately two ounces of cocaine to Villo-das on one occasion, on June 14, 2001, in the bathroom of La Copa Rota, after Villo-das had arranged several days earlier to buy cocaine from Ofray. Mendoza-Lebrón initially stated only that he often saw Diaz-Clavell present during drug deals that Ofray conducted and that “whenever Eric [Ofray] gave things to me, Pepe [Diaz-Clavell] would always be present....” When asked whether Diaz-Clavell did “anything in relation to the drug deal?”, Mendoza-Lebrón answered as follows, “Well, most of the time. I can’t say all of the time because not all of the time was he [Diaz-Clavell] with Eric but he was present during several transactions that I did with Eric.” It is well-settled that such testimony, establishing nothing beyond the fact that Diaz-Clavell was present during drug transactions, without more, would not make him culpable of conspiracy. See, e.g., United States v. Guerrero, 114 F.3d 332, 342 (1st Cir.1997) (“[P]roof of sufficient participation in the crime, as well as knowledge of it, is required to convict; the defendant’s ‘mere presence’ at the scene of criminal activity is not enough.”); United States v. Ocampo, 964 F.2d 80, 82 (1st Cir.1992) (although defendant knew that her residence was used by co-defendant for drug dealing, government must also prove defendant’s participation); United States v. Hyson, 721 F.2d 856, 862-63 (1st Cir.1983). On re-direct examination, Mendoza-Lebrón did state that Diaz-Clavell assisted Ofray “several times” in the latter’s drug dealing by helping Ofray to stash cocaine. However, Mendoza-Le-brón’s testimony lacked detail. This evidence, while certainly enough to justify affirmance on review for legal sufficiency, was not so conclusive as to convince us that the jury note error was harmless beyond any reasonable doubt. Mendoza-Lebrón’s testimony about Diaz-Clavell’s conduct stood alone. No other Government witness observed Diaz-Clavell assist Ofray in any narcotics transaction. No physical evidence linked Diaz- *28 Clavell to Ofray’s drug dealing. Villodas’ testimony was similarly uncorroborated. There was no testimony by any Government agent or any other witness to support Villodas’ account of the June 14, 2001 cocaine deal in La Copa Rota, and there was no physical evidence that tied Diaz-Clavell to the drug deal. In addition, the recording that Villodas made of the cocaine transaction, which revealed the voices of both Villodas and Diaz-Clavell, did not contain any statements relating to the drug deal. The recording merely captured Villodas and Diaz-Clavell (Villodas’ second cousin) conversing about family members, which was consistent with Diaz-Clavell’s own testimony about his encounter with Villodas at La Copa Rota on June 14, 2001. The testimonies of Agent Rivera and Agent Tirado, portions of which the Government cites in support of its assertion that evidence of Diaz-Clavell’s guilt was overwhelming, provided no corroboration. Although Agent Rivera testified that Diaz-Clavell acted as a drug runner for Ofray at La Copa Rota, the statement was based not on his first-hand knowledge but rather on information gleaned from the Government’s cooperating witnesses — namely, Mendoza-Lebrón and Villodas. In other words, Agent Rivera’s statement constituted “overview” testimony delivered at the outset of the Government’s case, in which the case agent merely summarized the evidence that would later be provided by Villodas and Mendoza-Lebrón. Similarly, Agent Tirado’s testimony that, “[t]hrough the investigation that we did,” Diaz-Clavell assisted Ofray in the drug trade did not serve to corroborate the testimonies of the cooperators. Like Agent Rivera’s testimony, Agent Tirado’s testimony — which purportedly provided additional support for the accounts of the cooperating witnesses-in fact derived entirely from and thus did no more than summarize the testimonies of Mendoza-Lebrón and Villodas. Agent Rivera’s and Agent Tirado’s conclusions regarding Diaz-ClavelPs role in the conspiracy served merely to bolster the testimony of the cooperating witnesses. Although Agent Tirado also testified that, following the completion of the June 14, 2001 cocaine transaction at La Copa Rota, Villodas stated to Tirado that Diaz-Clavell had given the drugs to Villo-das, this testimony cannot be deemed to be corroborative. In considering the strength of the Government’s case for the purposes of a harmless error analysis, we consider only evidence that was properly admitted. See, e.g., United States v. Zanghi, 189 F.3d 71, 83 (1st Cir.1999) (weighing only government’s admissible evidence in conducting harmless error analysis); United States v. Ferreira, 821 F.2d 1, 8 (1st Cir.1987) (finding error not to be harmless where “strength of the admissible, circumstantial evidence is not ‘overwhelming’ ”) (emphasis added). In assessing the relative strength of the Government’s case, it is also significant that Diaz-Clavell testified on his own behalf. Diaz-Clavell stated that, although he spoke with his cousin, Villodas, on June 14, 2001 at La Copa Rota, where he worked and frequently visited even when not working, he did not at that time or any other time deliver drugs to Villodas. Diaz-Clavell also testified, more generally, that although he frequented La Copa Rota and Rumba’s Pub and associated with Of-ray, both socially and as a part-time employee, he did not assist Ofray in conducting drug deals. On cross-examination, the Government did not substantively impeach Diaz-Clavell’s testimony. While the jury chose to credit the accounts of the cooperating witnesses over the admittedly self-serving testimony of the defendant, Diaz-Clavell’s countervailing testimony on his *29 own behalf is a factor in conducting the harmless error analysis. Cf. Garcia-Morales, 382 F.3d at 18 (finding error harmless where government’s case was strong but “defense was relatively weak, consisting entirely of cross-examination of the prosecution’s witnesses to challenge their credibility”); United States v. Sepulveda, 15 F.3d 1161, 1182 (1st Cir.1993) (finding error harmless where defendant’s evidence was “a drop in the proverbial bucket”). It is also instructive to contrast the weight of the Government’s evidence in this case with the prosecution’s proof in Santana. There, we found that the evidence against Santana, though legally sufficient, was not strong enough to render harmless the admission of prejudicial extrinsic evidence. In Santana, the evidence of the defendant’s participation in the alleged narcotics activity consisted of the agent’s eyewitness identification of Santana as a drug supplier as well as the following facts: (1) the November 8 drug transaction took place in a vehicle registered in [Santana’s] name, and law enforcement officers observed Santana twenty minutes after the transaction standing next to the vehicle in front of his business; (2) on February 23, 1995, [Santana’s non-testifying co-defendant] called, then visited, Santana’s business shortly after an unexpected drug order was placed; and (3) on March 23, 1995, [the non-testifying co-defendant] told [the government’s confidential informant] that the crack cocaine would be of good quality because his supplier was the “same guy, red truck.” Santana, 175 F.3d at 67. The evidence against Diaz-Clavell was no more persuasive. As discussed, there was no admissible direct evidence to corroborate the accounts of the two cooperating witnesses. Finally, the extrinsic information received by the jury was especially prejudicial to Diaz-Clavell. The risk that Diaz-Clavell, a comparatively minor participant in the widespread nine-year conspiracy, would be found guilty merely because of his association with his co-defendants, against most of whom far more evidence was adduced, was particularly high. Although a relative paucity of evidence was presented against Diaz-Clavell, the district court’s answer to Note #2 was likely to cause the jury to disregard weaknesses in the Government’s case and result in a finding of guilt simply by virtue of his association with the numerous other individuals named in the Indictment. Thus, in Diaz-Clavell’s case, there was more than a reasonable possibility that the extrinsic information “tipped the scales in favor of the government’s theory.” Rivera-Santiago, 107 F.3d at 967. After considering the prejudicial nature of the extrinsic information, as well as the relative strength of the Government’s case against Diaz-Clavell, “it is impossible to conclude beyond a reasonable doubt that the court’s error” in informing the jury that thirty-seven absent co-defendants were in prison for the conspiracy “did not contribute to the verdict.” Santana, 175 F.3d at 67. Accordingly, we vacate Diaz-Clavell’s conviction and remand for a new trial. 8 Zaragoza-Lasa The evidence offered against Zar-agoza-Lasa, while also legally sufficient to support the conviction for drug conspiracy, was not so overwhelming as to persuade us that the jury note error was harmless in *30 his case. Despite the Government’s contention that overwhelming evidence established that Zaragoza-Lasa was “a drug supplier to Las Avispas,” comparatively little evidence linked Zaragoza-Lasa to any drug transactions. Zaragoza-Lasa was purported to be a major supplier of heroin to the Avispas drug point at Las Vias. However, the only admissible evidence of Zaragoza-Lasa’s involvement in any narcotics activity was provided by the brief and uncorroborated testimony of two cooperating witnesses, Francisco Cardona and Olga Lebrón-Ortiz, both of whom identified Zaragoza-Lasa in court but knew him only by his nickname “El Viejo Mode.” Cardona, who worked regular shifts as a Las Avispas street dealer at Las Vias from 1996 through 1999, testified only that he saw Zaragoza-Lasa on “several occasions,” or “three or four times,” arrive at Las Vias, speak with the drug point’s manager, Javier Baez, and on one occasion hand Baez a brick-shaped package wrapped in paper. Cardona testified that a runner brought him what appeared to be the paper that had wrapped the package and that cocaine residue was present on the paper. Cardona also testified that the drug supplies at Las Vias increased following Zaragoza-Lasa’s visits. Lebrón-Ortiz also testified that she witnessed Zaragoza-Lasa frequently in Bor-inquen from 1998 to 2002 and witnessed Zaragoza-Lasa hand to Baez on one occasion a transparent bag filled with violet-colored aluminum foil parcels. Lebrón-Ortiz also provided the only direct evidence of Zaragoza-Lasa’s involvement with narcotics: namely, that on two occasions she witnessed him giving tastes of heroin to street junkies in Borinquen. The above-mentioned testimony permits the inference that Zaragoza-Lasa supplied heroin to Las Vias. It is quite another matter, however, to conclude that this evidence rendered harmless beyond any reasonable doubt the prejudicial extrinsic information to which the jury was exposed. As with Diaz-Clavell, the mere fact that Zaragoza-Lasa was present in a known drug area (Las Vias) and associating with an Avispas member (Baez) is not enough. Nor is there much heft in the fact that both cooperating witnesses observed Zara-goza-Lasa hand suspicious packages to Baez. Although an inference can be drawn that Zaragoza-Lasa delivered heroin to Baez, because the violet-colored aluminum packaging was consistent with the type of packaging Las Avispas typically used for heroin, Lebrón-Ortiz did not actually observe Zaragoza-Lasa possess or distribute heroin. Similarly, Cardona’s testimony about Zaragoza-Lasa’s handing Baez a suspicious package on one occasion and Cardona’s statement that more drugs were available at Las Vias after Zaragoza-Lasa’s visits do not amount to conclusive evidence of Zaragoza-Lasa’s guilt. The direct evidence of Zaragoza-Lasa’s drug activity, consisting only of the fact that Zaragoza-Lasa on two occasions provided heroin samples to Borinquen junkies, also was less than overwhelming. While this evidence established that Zaragoza-Lasa had some involvement in narcotics activity, it simply does not qualify as strong evidence of Zaragoza-Lasa’s participation in the charged narcotics conspiracy- Moreover, as with Diaz-Clavell, there was no physical evidence to corroborate the cooperating witnesses’ testimony. Zaragoza-Lasa was not recorded on any monitoring device. There was no physical evidence seized from him. He was not alleged to have taken part in any Government-arranged undercover drug purchase. Although Agents Rivera and Tirado testified confidently about Zaragoza-Lasa’s role as a major heroin supplier to Las *31 Avispas drug points, including Las Vias, their testimonies once again derived entirely from, in the words of Agent Rivera, “information that we obtained from informants who collaborated in the investigation and informants of the Puerto Rico Police Department....” Again, as discussed above, this purportedly corroborating evidence amounts to nothing more than case agent “overview” testimony that merely summarizes the first-hand accounts of the cooperating witnesses. As both Agents Rivera and Tirado conceded on cross-examination, their conclusions about Zaragoza-Lasa’s culpability were not based on their personal knowledge. Again, our decision in Santana provides illumination. The Government’s case against Zaragoza-Lasa, as against Diaz-Clavell, was not as strong as the case presented against Santana, where we found harmful error. Unlike Santana, the Government’s proof against Zaragoza-Lasa did not include a government agent’s eyewitness identification or circumstantial evidence that closely linked the defendant to an undercover drug purchase. Here, again, the evidence consisted solely of the uncorroborated testimony of two cooperating witnesses. Further, the Government offered more ample, direct evidence against López-Soto, Ofray, Cruz-Pereira, and absent members of the Avispas conspiracy named in the Indictment. As discussed above, the danger that a defendant, against whom comparatively less evidence was introduced, would be found guilty by association with more evidently culpable co-defendants was high, and was made correspondingly higher by the prejudicial nature of the extrinsic information imparted to the jury. As with Diaz-Clavell, in light of the prejudicial nature of the extrinsic information given to the jury and the relative strength of the Government’s ease against Zaragoza-Lasa, we cannot conclude with any assurance that the extrinsic information did not contribute to the jury’s determination that Zaragoza-Lasa was guilty. Accordingly, we vacate Zaragoza-Lasa’s conviction and remand for a new trial. 9 Individual Claims I. Insufficiency of Evidence of Conspiracy Appellant Cruz-Pereira claims that the district court erred in denying his motion for a judgment of acquittal pursuant to Rule 29 of the Federal Rules of Criminal Procedure because the Government presented insufficient evidence of his participation in the charged conspiracy. Review of a district court’s denial of a Rule 29 motion is de novo. United States v. Hernandez, 218 F.3d 58, 64 (1st Cir.2000). “Challenges to the sufficiency of the evidence and to the denial of the motion for judgment of acquittal raise a single is-sue_” Id. at 64 n. 4 (quoting United States v. Morillo, 158 F.3d 18, 22 (1st Cir.1998) (internal quotation marks omitted)). A district court’s denial of a motion for acquittal must be affirmed “unless the evidence, viewed in the light most favorable to the government, could not have persuaded any trier of fact of the defendant’s guilt beyond a reasonable doubt.” United States v. Paradis, 802 F.2d 553, 559 (1st Cir.1986). The reviewing court cannot weigh evidence or make credibility judgments. See United States v. Ortiz, 966 F.2d 707, 711 (1st Cir.1992). The court must reject only “those evidentiary interpretations ... that are unreasonable, *32 insupportable, or overly speculative, and must uphold any verdict that is supported by a plausible rendition of the record.” Hernandez, 218 F.3d at 64 (quoting United States v. Spinney, 65 F.3d 231, 234 (1st Cir.1995) (internal quotation marks omitted)). To prove Cruz-Pereira’s culpability of the charged conspiracy, the Government was required to show “the existence of a conspiracy, the defendant’s knowledge of the conspiracy, and the defendant’s voluntary participation in the conspiracy.” United States v. Gomez-Pabon, 911 F.2d 847, 852 (1st Cir.1990). To establish that Cruz-Pereira belonged to and participated in the drug conspiracy, the Government must show two kinds of intent: “intent to agree and intent to commit the substantive offense.” Id. at 853 (quoting Rivera-Santiago, 872 F.2d at 1079) (internal quotation marks omitted). The Government was not required to prove that each co-conspirator knew about or had contact with all other members of the conspiracy or knew details about and participated in every act in furtherance of the conspiracy. United States v. Martinez-Medina, 279 F.3d 105, 113 (1st Cir.2002); see also Rivera-Santiago, 872 F.2d at 1079 (“A defendant may culpably join a drug-trafficking conspiracy without knowing the full extent of the enterprise or the identities of all the coconspirators.”). Cruz-Pereira argues that, although the Government presented evidence of his drug sales to an undercover agent and an FBI informant, there was no evidence presented that connected Cruz-Pereira with Las Avispas or showed that he conspired with anyone (other than Government agents) to sell drugs. Thus, the defense claims that because the evidence proved only that Cruz-Pereira “worked alone,” there was insufficient evidence to convict him of conspiracy. As discussed above, the Government presented evidence that Cruz-Pereira not only sold drugs to Agent Rosa-Ferrer and Informant Villodas, but also that Cruz-Pereira engaged in crack transactions with Las Avispas. Specifically, Mendoza-Le-brón testified that he and his cousin, Wilson — an established Avispas dealer — sold crack to both Cruz-Pereira and Cruz-Per-eira’s co-defendant, Carlos Escobar Figueroa, at the drug points Cruz-Pereira and Escobar operated at Puente Los Jo-bos, from 1993 until Mendoza-Lebrón’s arrest in 1999. Mendoza-Lebrón also testified that he sold drugs to José Cabezón, the leader of Las Avispas. Cruz-Pereira’s regular drug purchases from Wilson, an established Avispas member, provided the jury with legally sufficient proof of Cruz-Pereira’s participation in the charged conspiracy. In addition, the jury reasonably could have inferred a link between Cruz-Pereira and Las Avispas arising from the fact that Cruz-Pereira bought crack from a dealer (Mendoza-Lebrón) who acted as a common supplier to other members of Las Avispas-namely, Escobar Figueroa and Cabezón. In sum, the Government presented evidence legally sufficient to support the jury’s finding that Cruz-Pereira was guilty of the charged conspiracy. II. Purported Judicial Bias López-Soto claims that the trial judge’s open hostility and bias against his trial counsel, Attorney Inserni, rendered his trial fundamentally unfair. As evidence of bias, López-Soto points out that the trial court, among other things, interrupted In-serni during opening and closing statements; told Inserni to “shut up” during a conference at sidebar and made demeaning comments about Inserni’s performance, including his use of notes; asked several leading questions of witnesses; and re *33 fused to pause trial proceedings while In-serni left court to use the restroom. “It is well-established that a judge is not a mere umpire; he is ‘the governor of the trial for the purpose of assuring its proper conduct,’ and has a perfect right—albeit a right that should be exercised with care—to participate actively in the trial proper.” Logue v. Dore, 103 F.3d 1040, 1045 (1st Cir.1997) (quoting Quercia v. United States, 289 U.S. 466, 469, 53 S.Ct. 698, 77 L.Ed. 1321 (1933)). Where, as here, an appellant claims that the trial judge overstepped his bounds and displayed judicial bias to the extent that the trial was rendered fundamentally unfair, the appellant must show that the judge’s actions resulted in “serious prejudice.” United States v. Cunan, 152 F.3d 29, 37 (1st Cir.1998). In reviewing the record for alleged judicial bias, we “consider [] ‘isolated incidents in light of the entire transcript so as to guard against magnification on appeal of instances which were of little importance in their setting.’ ” United States v. Candelaria-Silva, 166 F.3d 19, 35 (1st Cir.1999) (quoting United States v. Montas, 41 F.3d 775, 779 (1st Cir.1994)). The Supreme Court has held that “expressions of impatience, dissatisfaction, annoyance, and even anger ... are within the bounds of what imperfect men and women ... sometimes display. A judge’s ordinary efforts at courtroom administration—even a stern and short-tempered judge’s ordinary efforts at courtroom administration—remain Immune” to claims of judicial bias. Liteky v. United States, 510 U.S. 540, 555-56, 114 S.Ct. 1147, 127 L.Ed.2d 474 (1994); see also Morales v. INS, 208 F.3d 323, 327 (1st Cir.2000). Here, López-Soto has identified several moments during the trial which, at most, demonstrate the court’s occasional impatience or frustration with Attorney Inserni. The isolated instances cited by López-Soto do not amount to judicial bias, and López-Soto has not come close to showing how the purportedly improper conduct of the trial court resulted in serious prejudice. Although the trial judge did interrupt Inserni during opening and closing arguments, the interruptions were almost invariably interposed for a proper purpose. For example, during Inserni’s opening statement, when Inserni stated that his client was being charged with the murder of Santito, the district court quite rightly broke in to instruct the jury that López-Soto was not on trial for murder. To the extent that the trial judge expressed frustration with Inserni or rebuked him, almost all of the remarks at issue—including the court’s imploring In-serni to “shut up and let me finish”—were spoken at sidebar, out of the jury’s presence. See Candelaria-Silva, 166 F.3d at 35 (noting that “[o]n several occasions, this Court has held that a trial judge’s frustration displayed at sidebar does not deprive a defendant of a fair trial” and collecting cases). López-Soto has not identified any statement that the court made to Inserni in the jury’s presence, including the court’s remark about Inserni’s notes, that was so hostile or demeaning as to rise to the level of judicial bias. See United States v. Rodriguez-Rivera, 473 F.3d 21, 27-28 (1st Cir.2007). The questions that the district court asked certain witnesses also did not constitute judicial testifying or blatant bias against López-Soto; rather, on the relatively few occasions when the district court questioned a witness, it was usually to clarify a witness’s answer regarding the time frame or location of events, for example, about which the witness was testifying. A trial court’s sua sponte questioning of witnesses, to clarify counsel’s questions or the witness’s answer “is generally an *34 appropriate role for the judge to play.” Id. at 27. Finally, although the trial court’s refusal to continue proceedings for five minutes so that Inserni could use the restroom was not especially accommodating, the failure to grant the brief break did not result in even slight prejudice to Ló-pez-Soto, because the witness testifying at that time was not providing any evidence that was relevant to López-Soto. In sum, the few instances of the trial court’s frustration, cherry-picked from the voluminous record of a nearly two month-long trial, do not reveal judicial bias, let alone the serious prejudice against the Appellant that he must establish to prevail on a claim of judicial bias. López-Soto’s claim of judicial bias is without merit. III. Ineffective Assistance of Counsel López-Soto claims that he received ineffective assistance of counsel because his trial counsel, Attorney Inserni, revealed to the jury in opening argument that López-Soto was being prosecuted in the local court for the murder of Santito, a rival drug dealer. López-Soto asserts that In-serni’s commentary about the murder, which was neither charged in the Indictment nor listed as an overt act, opened the door to permit the government to introduce highly prejudicial testimony about the murder. “We have held with a regularity bordering on the monotonous that fact-specific claims of ineffective assistance cannot make their debut on direct review of criminal convictions, but, rather, must originally be presented to, and acted upon by, the trial court.” United States v. Mala, 7 F.3d 1058, 1063 (1st Cir.1993). This is because an appellate court usually is ill-equipped to handle the fact-specific inquiry that such claims often require. See id. “In addition, the insights of the trier, who has seen and heard the witnesses at first hand and watched the dynamics of the trial unfold, are often of great assistance.” United States v. Moran, 393 F.3d 1, 10 (1st Cir.2004). In the exceptional case, however, where the record is sufficiently developed, and critical facts are not in dispute, such claims may be reviewed. United States v. Torres-Rosario, 447 F.3d 61, 64 (1st Cir.2006). This is not a case that presents an exception to the well-settled rule. No claim of ineffective assistance was raised in the trial court. Although there was ample colloquy relating to Attorney Inserni’s decision to discuss the fact that local charges were pending against López-Soto for the murder of Santito, we cannot tell from the record whether Inserni’s decision to discuss the uncharged murder was a legitimate tactical decision at the time that it was made or deficient performance in violation of López-Soto’s right to effective assistance of counsel. See Moran, 393 F.3d at 10-11 (“Although hindsight is always 20/20, we cannot tell from this record whether [trial counsel’s decision], when made, was a calculated stratagem or a mere oversight.”). In addition, it is the trial court, rather than the appellate court, that is in the best position to assess whether Inserni’s decision, if it was in fact constitutionally deficient, resulted in prejudice to López-Soto’s substantial rights, as required under Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). See Moran, 393 F.3d at 11. Thus, we decline to review López-Soto’s claim. Instead, we dismiss the claim without prejudice to its reassertion, should the Appellant so choose, in a proceeding under 28 U.S.C. § 2255. See id. TV. Evidence of Santito’s Murder Cruz-Pereira argues that he was prejudiced by evidence that López-Soto confessed to the murder of Santito. He pro *35 tests that evidence of this murder was unrelated to the conspiracy, not charged in the Indictment, and not related to Cruz-Pereira, and caused a “spillover” effect that was highly prejudicial. Cruz-Pereira asserts that admission of evidence of San-tito’s murder violated Rules 403 and 404(b) of the Federal Rules of Evidence. We review the district court’s decision whether to admit evidence pursuant to Rule 404(b) for an abuse of discretion, and “will reverse a district court’s Rule 403 balancing ‘only in exceptional circumstances.’ ” United States v. Manning, 79 F.3d 212, 217 (1st Cir.1996) (quoting United States v. Garcia, 983 F.2d 1160, 1173 (1st Cir.1993)). Further, the court will treat any error in admitting Rule 404(b) evidence as harmless if it determines that the disputed evidence did not contribute to the verdict. United States v. Levy-Cordero, 156 F.3d 244 (1st Cir.1998). The court did not abuse its discretion in allowing evidence of Santito’s murder. First, Rule 404(b) is not implicated in this case. As the Government notes, the Indictment charged that the defendants engaged in violence to carry out the drug conspiracy. The Government presented evidence of the drug wars between Las Avispas and Las Jibaritos and other rival gangs, as well as evidence that a number of individuals acted as enforcers for the drug organizations. In particular, the Government charged in the Indictment that López-Soto acted as an enforcer and presented evidence at trial to establish his role as a “hit man” for Las Avispas. Thus, evidence that López-Soto murdered the leader of a rival drug gang is direct proof of the means used to carry out the conspiracy, and is not “other act” evidence that is subject to Rule 404(b). See, e.g., United States v. McGuire, 389 F.3d 225, 229 (1st Cir.2004) (noting that evidence that defendant pistol-whipped drug customer whom defendant thought had “set him up” was direct proof of defendant’s participation in conspiracy and thus was not other act evidence that required analysis under 404(b)). Under the Rule 403 analysis, the probative value of the evidence was not substantially outweighed by the danger of unfair prejudice. As discussed, the evidence was probative of López-Soto’s participation in the conspiracy and of the means used by the members of Las Avis-pas to further the drug ring’s goals. Further, the trial court instructed the jury that, although it could consider evidence of Santito’s murder “as part of the manner or means used by members of the conspiracy to carry out the objects of the conspiracy,” the jury could not consider the murder as evidence that any of the Appellants, apart from López-Soto, participated in the conspiracy. The trial court’s limiting instruction helped to safeguard the jury’s consideration of the evidence against the claimed “spillover” effect. See United States v. Sabatino, 943 F.2d 94, 96-97 (1st Cir.1991) (trial court’s limiting instructions may safeguard against prejudicial spillover effect). In sum, the probative value of the evidence was not substantially outweighed by its prejudicial effect. Admission of evidence relating to Santito’s murder was proper. V Restriction of Ofray’s Cross Examination of Govt. Witness Ofray claims that the district court erred by (a) not granting his Rule 33 motion for a new trial on the ground that the court improperly precluded Ofray from introducing evidence of a Government witness’s bias and motive; and (b) limiting Ofray at trial from cross-examining the Government’s witness as to bias and motive. *36 The Government’s witness, Agent Candelaria, testified that, in the early-morning of July 14, 2002, he was sleeping at his residence when he was awakened by a detonation that sounded like a gunshot. Candelaria called the police and then, from his window, observed Ofray on the street outside Rumba’s Pub, holding a weapon. Candelaria then saw Ofray walk up to a parked van and place the weapon inside. Based on the information provided by Candelaria, police later recovered the weapon, an automatic pistol, from the van. Ofray was arrested hours later, and at the time of his arrest, a second weapon was recovered from his duffel bag. During cross-examination, Ofray’s counsel sought to question Candelaria about alleged family ties and hostile relations that existed between him and Ofray. Specifically, Ofray wished to show that Cande-laria had known Ofray a long time because of the latter’s ownership of Rumba’s Pub and that the two men were in fact related, namely because Candelaria’s brother had a son with Ofray’s sister—in other words, that the two men shared a nephew. Ofray wished to introduce “evidence of family troubles between [Ofray’s] sister and the witness’ brother to impeach the witness as to bias and motivation to cause harm to [Ofray] by fabricating a case against him.” After hearing the proffer of testimony to be elicited from Candelaria, the district court precluded Ofray’s counsel from questioning Candelaria about the allegedly hostile family relationship. In addition to restricting the scope of Ofray’s cross examination of Candelaria, the trial court precluded the defense from calling in its case-in-chief Agent Irving Ofray, who Of-ray’s counsel proffered “was being called to testify about bias and motivation on the part of the Government agents to induce witnesses to fabricate evidence against the defendant-appellant.” Denial of a Rule 33 motion for a new trial is reviewed for abuse of discretion. United States v. Díaz, 300 F.3d 66, 78 (1st Cir.2002). As a threshold matter, the Government points out that the district court never ruled on Ofray’s Rule 33 motion. Accordingly, we remand Ofray’s Rule 33 motion to the district court. Thus, the remaining claim is that the court erred by restricting the scope of cross-examination. A court’s limitation of the scope of cross-examination is reviewed de novo to determine whether the defendant “was afforded a reasonable opportunity to impeach adverse witnesses” in conformity with the Confrontation Clause. United States v. Martinez-Vives, 475 F.3d 48, 53 (1st Cir.2007). Under the Confrontation Clause, a defendant has the right to cross-examine the government’s witness about his bias against the defendant and his motive for testifying. See Delaware v. Van Arsdall, 475 U.S. 673, 679, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986). Nevertheless, the right to cross-examination is not without limits, and “trial judges retain wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on such cross-examination based on concerns about, among other things, harassment, prejudice, confusion of the issues, the witness’ safety, or interrogation that is repetitive or only marginally relevant.” Id. “The trial court’s latitude in shaping such restrictions is ‘wide.’ ” United States v. Vega Molina, 407 F.3d 511, 523 (1st Cir.2005) (quoting Van Arsdall, 475 U.S. at 679, 106 S.Ct. 1431). “[Restrictions on cross-examination regarding bias are erroneous only if they are manifestly unreasonable or overbroad.” United States v. Callipari, 368 F.3d 22, 36 (1st Cir.2004), vacated on other grounds, 543 U.S. 1098, 125 S.Ct. 985, 160 L.Ed.2d 998 (2005) (quoting United States v. Gonzalez-Vazquez, 219 F.3d 37, 45 (1st Cir.2000) *37 (internal quotation marks omitted)). “To establish that the district court has abused its discretion, the defendant must show that the limitations imposed were clearly prejudicial.” United States v. Williams, 985 F.2d 634, 639 (1st Cir.1993). “It follows logically, therefore, that should an error be revealed, we may affirm the conviction if we are confident that it was harmless beyond a reasonable doubt.” Martinez-Vives, 475 F.3d at 53. Here, the district court precluded the proffered cross examination primarily on the ground that it was “collateral matter.” The court stated that cross examination as to Candelaria’s bias would be irrelevant, in light of the fact that other evidence corroborated Candelaria’s account, namely the testimony of a police officer who recovered a weapon in the van parked outside of Rumba’s Pub several hours after Candelaria claimed to have witnessed Ofray deposit the weapon in the van. The fact that Candelaria’s testimony was partially corroborated by the subsequent testimony of another officer, however, did not provide an adequate basis for the court’s preclusion of the cross examination sought by Ofray. While it is true that police recovered a firearm from the exact place where Candelaria testified the firearm was deposited, there was no corroboration of Candelaria’s testimony that it was Ofray, rather than another individual, who placed the weapon in the van. In other words, cross examination of Candelaria as to his bias against Ofray could have allowed the jury to infer that Candelaria observed an individual other than Ofray possess a gun and then — because of hostility between him and Ofray — -pinned the act of firearm possession on Ofray. Although cross examination of Candelaria would indeed have required a certain amount of delving into collateral matter, the issue of Candelaria’s possible bias against Ofray was a relevant fact for the jury to consider in deciding whether to credit Candelaria’s testimony. It was thus error for the trial judge to preclude the defense from attempting to show that there was a familial tie between Candela-ria and Ofray, and that bad blood existed between the two men. This was clearly evidence of Candelaria’s motive and bias for providing inculpating testimony. 10 Any error, however, was harmless, because the jury was confronted with other testimony that established that Of-ray sold large quantities of drugs and regularly carried weapons in connection with his role as the operator of several drug points. Although Ofray’s possession of a weapon on July 14, 2002, was an overt act charged in the Indictment, the jury could have chosen to disbelieve all of Candela-ria’s testimony about the gun possession, and even have disbelieved the testimony of the police officer who discovered the gun in the van, and still found Ofray to be guilty of the charged drug conspiracy. As discussed above, the Government’s proof of Ofray’s participation in the charged conspiracy was overwhelming. This was not a case where the success of the Government’s proof turned on the jury’s determination of Candelaria’s credibility. Cf. United States v. Mulinelli-Navas, 111 F.3d 983, 993 (1st Cir.1997) (finding erroneous limitation of cross examination not *38 harmless where proof of charges relied solely on government witness’s testimony and allowing cross examination to proceed could have allowed jury to discredit witness’s testimony). Here, Ofray’s theory of the case — urged in cross examination of most of the Government’s witnesses, in opening statement, and in closing argument — was that the Government had failed to present credible evidence of any connection between Ofray and Las Avispas. Of-ray’s theory of defense, although ultimately unsuccessful, did not suffer as a result of his inability to impeach Candelaria’s testimony regarding Ofray’s possession of a firearm on a single occasion. See United States v. Cunan, 152 F.3d 29, 38 (1st Cir.1998) (finding no error in court’s restriction of cross examination where restrictions did not impede assertion of theory of defense). In sum, the trial court’s decision to limit the cross examination of Candelaria, even if erroneous, was harmless in light of overwhelming evidence of Ofray’s guilt. VI. Sentencing Errors (A) Ofray At Ofray’s sentencing hearing on February 25, 2005, the district court, after calculating the applicable sentencing guidelines range as 188 to 235 months, imposed a “middle-of-the-Guidelines” sentence of 200 months. Ofray in essence asserts a two-part, procedural challenge to his sentence. First, he claims that the district court erroneously treated the sentencing guidelines as mandatory, rather than advisory. Second, Ofray maintains that the court failed to consider the relevant sentencing factors set forth in 18 U.S.C. § 3553(a). Where, as here, an appellant challenges the procedural aspects of his sentencing, our review is for abuse of discretion. See Gall v. United States, — U.S.-, 128 S.Ct. 586, 597, 169 L.Ed.2d 445 (2007). We determine whether the court made errors “such as failing to calculate (or improperly calculating) the Guidelines range, treating the Guidelines as mandatory, failing to consider the [18 U.S.C.] § 3553(a) factors, selecting a sentence based on clearly erroneous facts, or failing to adequately explain the chosen sentence- — -including an explanation for any deviation from the Guidelines range.” United States v. Politano, 522 F.3d 69, 72 (1st Cir.2008) (quoting Gall, 128 S.Ct. at 597) (internal quotation marks omitted). Ofray’s claim that the district court erroneously treated the guidelines as mandatory, rather than advisory, can be disposed of quickly. Here, it is evident from the sentencing transcript that the district court, and all parties, were well aware that the sentencing guidelines were only advisory. As the Government points out, Of-ray’s own counsel stated on the record at the sentencing hearing that “the guidelines are only advisory for this court.” The district court also stated as follows: “The court is using the guidelines as advisory....” There is no doubt, therefore, that the district court imposed the sentence under the correct understanding that the sentencing guidelines are advisory, not mandatory. Similarly, it is evident that the district court did not abuse its discretion in failing to consider the relevant sentencing factors set forth in § 3553(a). While the court did not expressly refer to the statute or recite the factors, it provided a reasoned and adequate explanation for the sentence that was imposed. See United States v. Garcia-Carrasquillo, 483 F.3d 124, 132 (1st Cir.2007) (“An important prerequisite to our reasonableness analysis is the district court’s reasoned explanation for the sentence imposed.... ”). We “allow a good deal of leeway” in reviewing the *39 adequacy of a district court’s explanation. United States v. Gilman, 478 F.3d 440, 446 (1st Cir.2007). Ultimately, we defer to “the district court’s sentence as long as the court has provided a plausible explanation, and the overall result is defensible.” United States v. Dixon, 449 F.3d 194, 204 (1st Cir.2006). Here, the district court properly began the sentencing procedure by determining Ofray’s adjusted offense level. The court then stated that, although it was aware of the advisory nature of the guidelines, “nevertheless I believe that the guidelines do provide the factors which the court should consider in this sentence, in which [sic] also provides a reasonable sentence in view of the evidence in this case.” The court then found that the applicable sentencing guidelines range for Ofray, which is uncontested on this appeal, was a term of imprisonment of 188 to 235 months and a range for a monetary fine of $20,000 to $4 million. Before imposing the 200-month prison term and the below-guidelines fine of $10,000, the district court gave the following explanation for the sentence: The court notes that the defendant[’]s substantial participation in furtherance of an extensive and violent drug trafficking enterprise and its detrimental consequences to society, which [sic] warrants a sentence at the middle of the guideline range. There is no indication that the district court failed to consider, or accord sufficient weight, to the relevant sentencing factors. The court emphasized the scope and violent nature of the conspiracy, Ofray’s extensive role within it, and the “detrimental consequences to society.” The court also stated that the reasonableness of a sentence at the middle of the guidelines range was justified by the evidence presented at trial. The reasonableness of the within-guidelines sentence is strongly supported by our review of the overall record. As discussed above, the evidence clearly established that Ofray sold large quantities of narcotics from various drug points that he owned and operated over the course of many years, and that he regularly carried weapons in connection with his narcotics activity. While the district court’s explanation for a “middle of the guideline range” sentence could have been more comprehensive, the court’s sentence was supported by a plausible explanation and constituted an overall defensible result. Accordingly, we find that Ofray “has not carried the heavy burden of proving that his within-the-range sentence was unreasonable or an abuse of discretion.” United States v. Innarelli, 524 F.3d 286, 292 (1st Cir.2008). (B) Cruz-Pereira Cruz-Pereira argues that the evidence was insufficient to support the district court’s imposition of a two-level upward adjustment to his base offense level for his role as an organizer, leader, manager, or supervisor of other participants in the drug-selling enterprise. See U.S.S.G. § 3Bl.l(c). We review a sentencing court’s findings of fact for clear error, while questions of law involved in sentencing determinations are reviewed de novo. See United States v. Ramos-Paulino, 488 F.3d 459, 463 (1st Cir.2007). “A question about whether the evidence is sufficient to support a particular guideline determination is a question of law and, therefore, engenders de novo review.” Id. “Where the undisputed facts support more than one plausible inference, the sentencing court’s choice among supportable alternatives cannot be clearly erroneous.” United States v. Garcia, 34 F.3d 6,10 (1st Cir.1994). *40 The sentencing guidelines provide for an upward adjustment to a defendant’s base offense level due to the relative importance of his role in the offense. United States v. Cruz, 120 F.3d 1, 3 (1st Cir.1997) (en banc) (citing U.S.S.G. § 3B1.1(c)). For a sentencing court to apply an upward adjustment pursuant to section 3Bl.l(c), the evidence must show that the defendant “exercised control over, organized, or was otherwise responsible for superintending the activities of’ at least one other participant in a criminal activity on at least one occasion. Garcia-Morales, 382 F.3d at 19 (quoting Cruz, 120 F.3d at 3) (internal quotation marks omitted); accord United States v. Voccola, 99 F.3d 37, 44 (1st Cir.1996) (finding that single directed transaction is sufficient to support imposition of enhancement). It is not enough, however, that the defendant merely controlled, organized, or managed criminal activities; rather, he must instead control, organize, or manage criminal actors. Ramos-Paulino, 488 F.3d at 464; see also United States v. Thiongo, 344 F.3d 55, 63 (1st Cir.2003) (“Section 3B1.1 does not apply to a defendant who merely organizes or supervises a criminal activity that is executed without the aid of others.”). Cruz-Pereira argues that there was no evidence in the record that permitted the district court to find that he supervised, managed, organized, led or otherwise exercised control over any non-government actor in the narcotics activity at issue. He maintains that the evidence established only that he acted alone in operating the Santa Ana drug point, located at his house, and the drug point located at his bar, La Ponderosa. The Government counters that the court’s finding was based on legitimate inferences. Specifically, the Government points to the fact that Informant Villodas testified that other people were present in the backyard of Cruz-Pereira’s house when Villodas made the controlled purchase of crack, and that the audio recording of the transaction, though of poor quality, revealed another voice or voices in the background. This evidence, the Government contends, permitted the inference that one or more drug dealers, apart from Cruz-Pereira, sold drugs out of Cruz-Per-eira’s house, over whom Cruz-Pereira necessarily exerted control. The Government also cites testimony to the effect that the drug point was open twenty-four hours a day, which also permitted the court to infer that others assisted Cruzr-Pereira in his narcotics activity. Finally, the Government claims that Cruz-Pereira is eligible for the enhancement solely on the basis of the fact that he exerted control over Las Avispas’ property, by virtue of his “management responsibility over the Las Avis-pas Santa Ana drug point,” from which Cruz-Pereira made the sales to the undercover agent and FBI informant. There was no sufficient basis for the court’s imposition of a two-level role-in-offense enhancement under section 3Bl.l(c). Where, as here, the basis for an enhancement pursuant to section 3Bl.l(c) is not “apparent from the record, ... the sentencing court, in order to apply such an enhancement, must make a specific finding which identifies those being managed ‘with enough particularity to give credence to the upward adjustment.’ ” United States v. Medina, 167 F.3d 77, 80 (1st Cir.1999) (quoting United States v. McDowell, 918 F.2d 1004, 1011 (1st Cir.1990)). Here, the district court at sentencing determined that Cruz-Pereira exercised control over others in conducting drug activity because the drug point at Santa Ana operated twenty-four hours each day and because the audio recording made by Informant Villodas during his controlled purchase of *41 crack in Cruz-Pereira’s house revealed another voice or voices. The district court drew the plausible inference, based on the audio recording made by Villodas and the fact of the drug point’s round-the-clock operation, that others must have been present while Cruz-Pereira sold drugs or that at least one other person played a role in the operation of the drug point. However, the record is devoid of any evidence to show that Cruz-Pereira exercised control over any individual. Even if the district court reasonably inferred that Cruz-Pereira associated with others in operating the drug point, there was no basis for the finding that the Appellant oversaw their activities. See Ramos-Paulino, 488 F.3d at 464. In Ramos-Paulino, where we remanded for re-sentencing, there was testimony that the defendant — who was convicted of alien smuggling — acted with the assistance of an individual named “Domingo,” who helped the defendant collect payment and information from, and provide false documents for, the aliens. We concluded that this testimony, which afforded the sole arguable basis for the sentencing court’s finding that the defendant exercised control over the criminal activity of another individual, provided an insufficient basis for the imposition of an upward enhancement. We explained that “[although [the defendant] worked hand in glove with the mysterious Domingo, there is nothing to show either that he was her subordinate in the chain of command or that she oversaw his activities.” Id. Here, too, there was no basis for the district court’s finding that Cruz-Pereira exercised any control over or oversaw the activities of any individual who participated in narcotics activity at the Santa Ana drug point. There was no evidence, for example, of any communication between Cruz-Pereira and any other individual from which Cruz-Pereira’s supervisory role could be inferred. Nor was there any evidence that any other person assisted Cruz-Pereira in the undercover purchases made by Villodas or Agent Rosa-Ferrer, or that any other individual performed any acts at Cruz-Pereira’s express or implied direction. The mere fact that Cruz-Per-eira was characterized as an owner of the Santa Ana drug point, without more, does not justify the imposition of the role-in-the-offense enhancement. See Medina, 167 F.3d at 80 (citing United States v. Graham, 162 F.3d 1180 (D.C.Cir.1998)); United States v. Sostre, 967 F.2d 728, 733 (1st Cir.1992) (a finding that defendant played an “essential role” in drug conspiracy insufficient to warrant enhancement under section 3B1.1). The Government’s argument that Cruz-Pereira qualified for the upward enhancement on the basis of his control over the “the property, assets or activities of a criminal organization,” is squarely foreclosed by our holding in Ramos-Paulino. 11 As we explained in that case: [T]he management of criminal activities (as opposed to the management of criminal actors) may ground an upward departure but not an upward role-in-the-offense adjustment. Although both may lead to similar outcomes, there is an important structural distinction between sentencing enhancements and sentencing departures. For present purposes, then, we are constrained by the unam *42 biguous ease law holding that management of criminal activities, standing alone, does not constitute a basis for a role-in-the-offense enhancement under section 8B1.1. Ramos-Paulino, 488 F.3d at 464 (citations omitted). Thus, the two-level role-in-the-offense enhancement imposed by the district court cannot be based on a finding that Cruz-Pereira managed property belonging to Las Avispas. Accordingly, because the district court’s imposition of a two-level upward enhancement pursuant to section 3Bl.l(c) was based on legally insufficient evidence, we vacate Cruz-Pereira’s sentence and remand for re-sentencing. Here, however, as in Ramos-Paulino, we leave open the full gamut of possibilities—for example, the district court, if it can identify a participant or participants under the [Cruz-Pereira’s] sway, may reimpose the managerial role enhancement; or it may essay an upward departure for management of criminal activities; or it may simply eschew any further embellishments and impose what it deems to be a reasonable sentence. We take no view either as to the course to be followed or as to the duration of the sentence to be imposed. In the first instance, these are matters for the sentencing court. Id. (C) López-Soto López-Soto claims that his sentence of forty years (480 months), which constituted an upward variance of greater than two and one half times the maximum recommended sentence under the applicable guidelines range, was unreasonable; that he was deprived of his right to due process by the retroactive application of Booker1 s “remedial” holding; and that the court failed to give him the required notice of the upward variance under Rule 32(h) of the Federal Rules of Criminal Procedure. “We review a district court’s sentence for reasonableness, which involves a procedural as well as a substantive inquiry.” Politano, 522 F.3d at 72 (citing Gall, 128 S.Ct. at 597). In reviewing the reasonableness of a particular sentence, we afford the district court broad discretion. “[A]fter the court has calculated the [applicable guidelines range], ‘sentencing becomes a judgment call, and a variant sentence may be constructed based on a complex of factors whose interplay and precise weight cannot even be precisely described.’ ” Id. (quoting United States v. Martin, 520 F.3d 87, 92 (1st Cir.2008)). “Assuming a plausible explanation and a defensible overall result, sentencing is the responsibility of the district court.” United States v. Jimenez-Beltre, 440 F.3d 514, 519 (1st Cir.2006). Here, the court gave the following explanation for its imposition of a sentence of 480 months: In this case we believe that the guidelines do not reflect the seriousness of the offense and do not provide reasonable and adequate deterrence and punishment. Based on evidence presented during the trial, the defendant was a trigger-man who possessed powerful weapons to facilitate the instant offense. Moreover, he was involved in violence at controlled drug points where he sold cocaine, crack cocaine, heroin and marijuana during different periods. Therefore, to provide just punishment in light of the seriousness of the offense and to protect the community, the Court will sentence the defendant according to the statute. López-Soto received a sentence that was two and one half times greater—and more *43 than twenty-four years longer — than the top of the recommended guidelines range. In such a case, the district court must offer an especially compelling reason for its sentence. See United States v. Smith, 445 F.3d 1, 4 (1st Cir.2006) (“The farther the judge’s sentence departs from the guidelines sentence ... the more compelling the justification based on factors in section 3553(a) that the judge must offer in order to enable the court of appeals to assess the reasonableness of the sentence imposed.”) (quoting United States v. Dean, 414 F.3d 725, 729 (7th Cir.2005) (Posner, J.) (Internal quotation marks omitted)); see also Martin, 520 F.3d at 91 (“[T]he guidelines are the starting point for the fashioning of an individualized sentence, so a major deviation from them must ‘be supported by a more significant justification than a minor one.’ ”) (quoting Gall, 128 S.Ct. at 597). The court’s stated grounds for the imposition of a sentence that so markedly exceeded the recommended guidelines range were neither sufficiently particularized nor compelling to survive our review for reasonableness. The district court based its upward variance on two factors: (1) López-Soto’s possession of “powerful weapons” as a “triggerman,” and (2) his involvement in violence in connection with the narcotics activity. The court linked these facts to several relevant factors set forth in § 3553(a), namely the “need for the sentence to reflect the seriousness of the offense,” “deterrence,” “to provide just punishment,” and “to protect the public.” 18 U.S.C. § 3553(a)(2)(A), (B), (C). While the factors identified by the court may have justified a substantial upward variance, they simply do not support the imposition of a statutory maximum sentence of forty years, that is so far above the guidelines range. López-Soto’s use of powerful weapons and his engagement in violence were not so unusually egregious as to justify the imposition of the most severe possible sentence. See, e.g., United States v. Poynter, 495 F.3d 349, 354 (6th Cir.2007) (vacating sixty year statutory maximum sentence, which was more than three times higher than guidelines maximum, because grounds for upward variance, while supporting imposition of severe sentence, did not sufficiently distinguish defendant from other offenders guilty of similar crimes). The court’s reference to López-Soto’s possession of weapons as a triggerman lacks compelling force, in part because firearm possession had already been considered, and accounted for, in the two-level enhancement applied in the calculation of Appellant’s adjusted offense level. See U.S.S.G. § 2Dl.l(b)(l). “When a factor is already included in the calculation of the guidelines sentencing range, a judge who wishes to rely on that same factor to impose a sentence above or below the range must articulate specifically the reasons that this particular defendant’s situation is different from the ordinary situation covered by the guidelines calculation.” United States v. Zapete-Garcia, 447 F.3d 57, 60 (1st Cir.2006). It was not necessarily unreasonable for the sentencing court to have found that a variance was warranted because López-Soto possessed more than one weapon in his role as an enforcer for Las Avispas and that the weapons were “powerful.” But these factors were not so distinct from the firearm possession that was incorporated into the guidelines calculation as to justify a variance of such magnitude. See id. Similarly, the court’s reference to Ló-pez-Soto’s engagement in violence was not sufficiently compelling to support the degree of increase over the top of the recommended guidelines range. The court expressly stated, at the beginning of the sentencing hearing, that it would not consider López-Soto’s alleged involvement in *44 the murder of Santito in imposing sentence. Further, the court concluded that, for purposes of assessing López-Soto’s conduct in determining the appropriate sentence, the presentence report “contains all the essential information I need.... I think it’s complete.” The pre-sentence report, in turn, did not refer to any specific acts of violence committed by López-Soto; rather, it described the role of enforcers within the drug ring, who would use weapons to protect the organization’s leadership and operations from rivals, and identified López-Soto as a “hitman” within the organization. If the court had, in fact, weighed the murder as a factor in sentencing, the imposition of such a severe sentence might have been justified. Absent the murder, however, the generic reference to “violence” — while supporting a sentence in excess of the guidelines — did not justify the imposition of the most severe sentence allowable by law. As deplorable as López-Soto’s conduct was, not all enforcers of drug organizations who carry automatic pistols and engage in generically described acts of violence deserve the statutory maximum sentence. “While there is no yardstick perfectly calibrated to measure one crime and one criminal from the next crime and the next criminal, there are certainly measurable differences between [López-Soto’s] situation and the situation of offenders who might warrant the statutory maximum or something approaching it.” Poynter, 495 F.3d 349, 354 (6th Cir.2007). By imposing the maximum sentence on López-Soto, “the district court left us little room to distinguish” between him and more culpable narcotics conspirators who occupy the role of enforcer within a drug ring. Id. An enforcer for a drug organization who, for example, is found by the sentencing judge to have committed one or more uncharged murders or other specific egregious acts of violence, or whose criminal history category “substantially under-represents the seriousness of the defendant’s criminal history or the likelihood that the defendant will commit other crimes,” U.S.S.G. § 4A1.3(a)(l), would be substantially more culpable than López-Soto. In sum, the district court’s description of López-Soto’s conduct, while justifying an upward variance, was not sufficiently compelling to support a statutory sentence of more than double the maximum of the applicable guidelines range. There was ample room for a variance above the guidelines and below the statutory maximum to accomplish the trial judge’s stated purposes in sentencing López-Soto. Although “we emphasize that we do not reject the sentence imposed below solely because of the magnitude of its deviation from the guideline-recommended range,” Zapete-Garcia, 447 F.3d at 61, the statutory maximum forty-year (480-month) sentence simply does not stem from a plausible explanation, does not constitute a defensible result, and therefore cannot survive our review for reasonableness. See Jimenez-Beltre, 440 F.3d at 519. Accordingly, without expressing any opinion on what sentence should be imposed on remand, we vacate López-Soto’s sentence and remand for re-sentencing consistent with this opinion. 12 *45 CONCLUSION For the foregoing reasons, we vacate the convictions of Díaz-Clavell and Zaragoza-Lasa and remand both cases for new trials. We affirm the convictions of Cruz-Pereira and López-Soto, but vacate both sentences and remand for re-sentencing consistent with this opinion. We affirm both the conviction and sentence of Ofray. Footnotes [1] The second count, which was charged against only a few of the defendants, was a narcotics forfeiture count and is not relevant to any issue raised in the present appeals. [2] The bar was also known, and at times referred to at trial, by the name "La Copa Deportiva.” [3] Collazo's testimony about López-Soto's statement was admitted, over objection, as a co-conspirator statement. [4] It is unclear why Note # 2 — which was signed by the foreperson and thus presumably submitted fifteen minutes after “Jury Note # 3” — was not in fact marked as the third (rather than the second) note. [5] Although the trial judge translated Note # 2 into English to read that the “others in the Indictment” were in prison for “a conspiracy,” the Spanish words "por conspiración” are more aptly translated as "for conspiracy.” In any event, given the context of Note # 2, as discussed below, it is evident that the jury was referring to the conspiracy charged in the Indictment. [6] Because we have determined that the parties and defense counsel were, in fact, present when the district court drafted its answer to Note # 2, we need not decide whether counsel’s failure to follow the court’s ten-minute rule operated to waive or forfeit an appellate challenge to the manner in which the court handled the note or the substance of the court’s answer. [7] In considering the question of prejudice, it is also worth noting that AUSA Feldman, after being apprised of the trial court’s answer to Note # 2, expressed reservations about the answer, stating to the court, “Unless you think [the answer to the note] would be prejudicing them.” The incomplete transcript prevents us from determining exactly when AUSA Feldman made this remark. What is clear, however, is that the Government was aware that the court's answer could prejudice the defendants. [8] Because we vacate Diaz-Clavell’s conviction, and because the other issues raised by Diaz-Clavell on this appeal are unlikely to be implicated on re-trial, we need not consider the other claims he has asserted. [9] Having vacated Zaragoza-Lasa’s conviction, we do not address the rest of his claims, which—like the additional claims asserted by Diaz-Clavell—are unlikely to be implicated in a new trial. [10] The Government seeks to justify the court’s decision to preclude on the ground that the proffered cross examination was "inherently speculative.” But it is clear from the record of the proffer that Ofray’s counsel had an adequate basis from which to assert his claim that Candelaria had a bias against Ofray stemming from hostile familial relations. It is worth noting that the district court did not preclude the requested cross examination on the ground that the line of questioning was inherently speculative. [11] The note to section 3Bl.l(c), which the Government cites, provides, in relevant part, that a sentencing court may impose an upward departure "in the case of a defendant who did not organize, lead, manage, or supervise another participant, but who nevertheless exercised management responsibility over the property, assets, or activities of a criminal organization.” U.S.S.G. § 3B1.1, cmt. n. 2. [12] Because we have disposed of López-Soto’s sentencing argument on reasonableness grounds, we need not address his argument that his sentence created an ex post facto effect that deprived him of due process. See Zapete-Garcia, 447 F.3d at 59 n. 3 (citing United States v. Lata, 415 F.3d 107, 112 (1st Cir.2005)). Similarly, we need not rule on Appellant's challenge to his sentence on the grounds that proper notice of the upward variance was not provided pursuant to Rule 32(h) of the Federal Rules of Criminal Procedure. We note, however, that under the Supreme Court’s recent decision in Irizarry v. *45 United States, - U.S. -, 128 S.Ct. 2198, 2200, 171 L.Ed.2d 28 (2008), which was reached during the pendency of the present appeals, Rule 32(h) does not apply to a variance from a recommended guidelines sentencing range.