dissenting.
In Dellosantos the government charged defendants with participating in a single conspiracy (the so-called “overarching conspiracy” covering two states and two illicit drugs). The court in that case unequivocally stated: “we conclude that there was insufficient evidence to support the finding of a single conspiracy.” 649 F.3d at 119. My colleagues must therefore agree (they would say so if they did not), that the Fifth Amendment bars the government from re-indicting defendants for that single, over*45arching conspiracy charged but not proven in Dellosantos. Indeed, it would appear that even the government agrees with that unremarkable proposition. The government instead re-indicted defendants on a narrower, cocaine-only portion of the previously charged single, overarching conspiracy. So we have a classic double jeopardy question: Is the new charge the “same offense” as the offense for which defendants were previously indicted and acquitted?
When the offenses to be compared are each conspiracy offenses, the comparison can sometimes be tricky. See United States v. Laguna-Estela, 394 F.3d 54, 57 (1st Cir.2005) (employing a “nuaneed,” five-factor test). Here, though, the government concedes that there is no element of the second charged offense that was not an element of the first charged offense. Indeed, there will not even be any new evidence. Under basic double jeopardy rules, it therefore follows that the two offenses are the same. See Brown v. Ohio, 432 U.S. 161, 168, 97 S.Ct. 2221, 53 L.Ed.2d 187 (1977) (“[T]wo offenses are the same unless each requires proof that the other does not.”).
The majority misses this straightforward answer by failing to compare the offense charged in the new indictment (the Massachusetts-based, cocaine-only conspiracy) with the offense for which defendants were previously acquitted (the single overarching conspiracy). Instead, the majority devotes most of its effort to comparing the offense charged in the new indictment with only the unproven portion of the overarching conspiracy charged in Dellosantos. Noting that Dellosantos found that the Massachusetts-based, cocaine-only conspiracy and the Maine-based, cocaine and marijuana conspiracy were “distinct,” 649 F.3d at 119, the majority interprets that finding to mean that, for double jeopardy purposes, the newly charged offense is different than the offense for which defendants were acquitted.
That is simply the wrong comparison. As the majority avoids acknowledging, but cannot claim to dispute, Dellosantos plainly held that the evidence was insufficient to prove the single, overarching conspiracy that was charged. Id. So, that is the offense for which re-prosecution is barred. And that is therefore necessarily the offense to which this new indictment must be compared to see if it alleges a different offense (i.e., one with a new element).
To support a contrary and entirely unprecedented approach, my colleagues declare that the single conspiracy charged in Dellosantos “did not exist.” Slip Op. at 16. And to make that declaration binding on defendants, their two opinions audition a trio of preclusion doctrines for roles these .doctrines have never played before in any cited precedent. For example, to script a role for the law of the case doctrine, the concurrence must pretend that this case is actually the same case as Dellosantos. See, e.g., United States v. Moran, 393 F.3d 1, 7 (1st Cir.2004) (The law of the case doctrine governs “the same issues in subsequent stages in the same case ”) (emphasis added). As the docket numbers and common sense evidence, it clearly is not.13 More importantly, in applying each of the estoppel doctrines mentioned in their opinions, my colleagues make three fundamental mistakes.
First, their description of the holding in Dellosantos that they say now binds these *46defendants in this subsequent case materially errs in its lack of precision. The precise question at issue in Dellosantos was whether the government had proven beyond a reasonable doubt that there was an overarching, two-state conspiracy to distribute both cocaine and marijuana. The holding in Dellosantos that the evidence was insufficient to prove such a single conspiracy required acquittal on the charge. Whether there was in fact no such single conspiracy is something that the court could not have known and the defendants need not have proved (and actually did not argue). For that simple reason there is no basis at all for applying issue preclusion to establish that the single overarching conspiracy never existed. See, e.g., Global NAPs, Inc., v. Verizon New Eng., 603 F.3d 71, 95 (1st Cir.2010) (Issue preclusion requires that resolution of the issue was essential to judgment on the merits in the first case).
Second, defendants are not engaged in any inconsistent advocacy regarding the existence of the overarching conspiracy. They argued, first, that the government failed to prove such a conspiracy; they argue, now, that the government is still trying to convict them of a lesser included version of that exact same conspiracy. The lack of any inconsistency between these two positions renders judicial estoppel inapplicable. See, e.g., Perry v. Blum, 629 F.3d 1, 8-9 (1st Cir.2010) (Judicial estoppel requires, among other things, that the party’s earlier and later positions are “clearly inconsistent.”).
Third, and most tellingly, even if one were to accept the proposition that these defendants are now bound to accept as a fact the non-existence of the single overarching conspiracy for which they were acquitted in Dellosantos, so what? Such a proposition leaves untouched the pertinent double jeopardy inquiries: Can the prosecution re-indict on that offense? Clearly it cannot. And does this newly charged offense have any element that was not included in that prior acquitted offense? As the government concedes, it does not. Under Brown v. Ohio, double jeopardy therefore bars this second prosecution. 432 U.S. at 168, 97 S.Ct. 2221.
This conclusion is hardly surprising. By definition, most acquittals occur precisely because the government fails to prove a part of its case. If that failure therefore were to mean that we no longer compare the newly charged offense to the prior acquitted offense because the greater crime charged “did not exist,” then Fifth Amendment protections would shrink markedly. Surely the majority does not mean to suggest that anytime the government overcharges and a court finds that it fails to prove the aggravating element, the government may re-indict on a lesser included offense merely by pointing out that it failed to prove a portion of the previously acquitted, greater offense? Brown v. Ohio plainly prohibits such a gambit. Id. Yet that is exactly what the majority says the government can do here.14
Nor does it matter that the government may have proved part of its case the first *47time around. That happens in most prosecutions that end in acquittal. See, e.g., United States v. Glenn, 828 F.2d 855, 858-60 (1st Cir.1987). Nevertheless, as long as the disposition “represents a resolution, correct or not, of some ... of the factual elements of the offense charged,” the double jeopardy clause bars re-prosecution for that same offense. United States v. Martin Linen Supply Co., 430 U.S. 564, 571, 97 S.Ct. 1349, 51 L.Ed.2d 642 (1977). And, as I must stress, that “same offense” includes any offense that does not contain a new element. Brown, 432 U.S. at 168, 97 S.Ct. 2221.
The majority cannot pass its holding off as a one-time only, idiosyncratic mulligan for the prosecution. Prosecutors not infrequently stretch ambitiously in charging the scope of conspiracies so as to rope in more drugs, more guns, and thus higher minimum and maximum sentences. See, e.g., Glenn, 828 F.2d at 858-60. And jurors not infrequently buy such claims, leading to judicial reversals for insufficient evidence. In Glenn, for example, we reversed a conviction for conspiracy to distribute both hashish and marijuana because the evidence showed only that the defendant was involved in a conspiracy to distribute hashish. Id.; see also United States v. Franco-Santiago, 681 F.3d 1, 4 (1st Cir.2012) (insufficient evidence to show that defendant conspired to rob multiple banks, though sufficient evidence to show he conspired to rob one bank); United States v. Valerio, 48 F.3d 58, 63-65 (1st Cir.1995) (insufficient evidence to show that defendant conspired to possess cocaine with intent to distribute, though arguably sufficient evidence that she conspired to possess cocaine); United States v. Hernandez, 625 F.2d 2, 3-4 (1st Cir. 1980)’ (insufficient evidence to show that defendant conspired to distribute eight pounds of cocaine, though arguably sufficient evidence that he conspired to sell one ounce). After today, in this circuit (but fortunately in no others yet), the government need not worry itself too much over losses of that type; rather, it can recharge on the smaller, subsumed conspiracy, using even (as here) only evidence from.the first unsuccessful prosecution.15
With some reason, my colleagues may wonder in hindsight whether Dellosantos was correctly decided.16 But the Fifth Amendment prevents us from revisiting how to decide Dellosantos. Martin Linen Supply Co., 430 U.S. at 571, 97 S.Ct. *481349 (“A verdict of acquittal could not be reviewed on error or otherwise without putting (a defendant) twice in jeopardy.”) (internal citations and quotation marks omitted). The only relevant finding from Dellosantos now is that it issued a “substantive ruling that went to the ultimate question of guilt or innocence,” Slip Op. at 11, indicating acquittal for insufficient evidence on the charged offense. Dellosantos, 649 F.3d at 126.
The government could have avoided all of this by pleading alternative counts in Dellosantos. See United States v. Calder-one, 982 F.2d 42, 48 (2d Cir.1992). Prosecutors are well aware of the double-edged sword presented by tiered, lesser included offenses. Indicting a defendant in the alternative on both greater and lesser included versions of a crime likely increases the chances of a guilty verdict but also decreases the chances of a guilty verdict on the greater offense. Here, the government was overly ambitious, charging defendants with only the most extended characterization of the conspiracy that involved the greatest drug quantity. It then failed to prove that charged conspiracy. Allowing the government to take a second' shot, trying defendants for an offense entirely subsumed in the prior acquitted offense, materially shifts the balance in favor of the government in a way that our founders sought to avoid. And because prosecutors frequently charge conspiracy counts and stack drug quantity offenses, others who may actually be innocent of any crime may pay a dear price.
I therefore respectfully dissent from the majority’s decision to permit a post-acquittal prosecution for a lesser included version of the prior charged offense.
. Were it the same case, then I would presume no additional appointments of counsel under the Criminal Justice Act would have been required, Speedy Trial Act deadlines would be calculated accordingly, limitations periods would not have continued running after the acquittal, and so on.
. Implicitly recognizing that double jeopardy would bar the prosecution here if Dellosantos resulted in an acquittal for insufficient evidence, the concurrence argues that Dellosantos vacated defendants' convictions on purely procedural grounds, like a defect in the charging instrument, and thus double jeopardy does not bar this prosecution. See Montana v. Hall, 481 U.S. 400, 404, 107 S.Ct. 1825, 95 L.Ed.2d 354 (1987). But Dellosantos plainly stated: "the government failed to prove beyond a reasonable doubt ... the conspiracy that was charged.” 649 F.3d at 126 n. 18 (emphasis in original). There was no defect in the charging instrument except to the extent that the government charged an offense that it could not prove.
. The concurrence, like the majority, offers no counter to the argument that its approach allows re-prosecution for uncharged lesser included offenses in a broad array of cases of which the foregoing is simply a sample.
. Under Griffin v. United States, 502 U.S. 46, 112 S.Ct. 466, 116 L.Ed.2d 371 (1991), the indictment might arguably have been read as charging a dual object conspiracy that could arguably be proven merely by supporting a conviction on one of the two objects. 649 F.3d at 126-27 (Howard, J., concurring in part and dissenting in part). Or one might arguably have read the indictment as simply a charge of conspiracy to distribute and possess controlled substances, with the dual product allegations serving as details that would lead only to a mere variance analysis. See, e.g., Martin v. Kassulke, 970 F.2d 1539, 1545-46 (6th Cir. 1992); cf. Marshall v. Bristol Superior Court, 753 F.3d 10, 18 (1st Cir.2014).
Apparently, though, the prosecution did not press these arguments. 649 F.3d at 126 n. 18 (“[T]he government argues in a cursory manner. ...”) and 127 ("But the government has to make the argument to benefit from it.”) (Howard, J., concurring in part and dissenting in part). In any event, even if the result in Dellosantos was wrong, we cannot ñx it now without erring further. See Evans v. Michigan, - U.S. -, 133 S.Ct. 1069, 1081, 185 L.Ed.2d 124 (2013) ("[T]here is no way for antecedent legal errors to be reviewable in the context of judicial acquittals unless those errors are also reviewable when they give rise to jury acquittals....”).