OPINION
GILDEA, Chief Justice.This case presents the issue of whether the United States Supreme Court’s decision in Padilla v. Kentucky, — U.S. —, 130 S.Ct. 1473, 176 L.Ed.2d 284 (2010), holding that the Sixth Amendment right to counsel includes the right to be informed about the deportation consequences of a *483guilty plea, applies retroactively to cases on collateral review. In a motion to withdraw his guilty plea, respondent Rene Reyes Campos argued that Padilla applies retroactively to his conviction. Based on Padilla^ Reyes Campos contended that his attorney’s failure to warn him of the deportation consequences of his guilty plea constituted ineffective assistance of counsel and rendered his plea invalid. The district court determined that Padilla could not be applied to Reyes Campos’ collateral attack on his conviction. The court of appeals reversed. Because we conclude that Padilla announced a new rule of -criminal procedure that does not apply to a collateral review of Reyes Campos’ conviction, we reverse.
On May 26, 2009, the State filed a delinquency petition in juvenile court, charging Reyes Campos with felony simple robbery committed for the benefit of a gang pursuant to Minn.Stat. §§ 609.24 and 609.229, subds. 2, 3(a), 4 (2010).1 Reyes Campos waived his right to a certification hearing, and the district court certified Reyes Campos for prosecution as an adult. Reyes Campos entered a guilty plea in district court on July 10, 2009, to an amended charge of simple robbery in violation of Minn.Stat. § 609.24.
At the plea hearing, defense counsel questioned Reyes Campos on the record regarding his understanding of the plea. Defense counsel asked questions regarding Reyes Campos’ understanding that he was giving up his trial rights, including his presumption of innocence, right to present a defense, and right to be convicted only by a unanimous jury after the State proved its case beyond a reasonable doubt. But Reyes Campos was not questioned or informed about any immigration consequences of his plea.2 Nor was he asked whether he understood the immigration consequences of his guilty plea. See Minn. R.Crim. P. 15.01, subd. l(6)(i) (providing that “[bjefore the judge accepts a guilty plea, the defendant must be sworn and questioned by the judge with the assistance of counsel,” including questioning regarding the defendant’s understanding that “[i]f the defendant is not a citizen of the United States, a guilty plea may result in deportation, exclusion from admission to *484the United States, or denial of naturalization as a United States citizen.”)
The district court accepted Reyes Campos’ guilty plea. The court stayed imposition of sentence, placed Reyes Campos on probation for 3 years, and ordered that, as a condition of probation, he serve 365 days in the Hennepin County workhouse, with credit for 50 days already spent in custody.
When he entered his guilty plea, Reyes Campos had just turned 18 years old, and was a legal permanent resident, but not a citizen, of the United States. Reyes Campos’ conviction for simple robbery, in conjunction with the 365-day workhouse condition, resulted in Reyes Campos having an “aggravated felony” conviction under the Immigration and Nationality Act (INA). 8 U.S.C. § 1101(a)(43)(F), (G) (2006).3 The INA provides that “[a]ny alien who is convicted of an aggravated felony at any time after admission is de-portable.” 8 U.S.C. § 1227(a)(2)(A)(iii) (2006). In January 2010, Immigration and Customs Enforcement detained Reyes Campos, and Reyes Campos has since been deported to Nicaragua, his country of citizenship.
On May 26, 2010, Reyes Campos filed a motion to withdraw his guilty plea pursuant to Minn. R.Crim. P. 15.05, subd. I.4 Reyes Campos relied in his motion on the United States Supreme Court’s March 31, 2010, decision in Padilla, 130 S.Ct. 1473, holding that an attorney provides ineffective assistance of counsel under the Sixth Amendment by failing to inform a client of the deportation consequences of a guilty plea. Reyes Campos argued that Padilla applied retroactively to his case, and that under Padilla’s holding he received ineffective assistance of counsel when his counsel failed to inform him of the immigration consequences of his plea. As a result of his counsel’s ineffectiveness, Reyes Campos contended that his plea was uninformed and therefore invalid. See Hill v. Lockhart, 474 U.S. 52, 56, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985) (holding that a defendant’s guilty plea may be constitutionally invalid if the defendant received ineffective assistance of counsel, rendering his guilty plea not intelligent); State v. Ecker, 524 N.W.2d 712, 718 (Minn.1994) (“When an accused is represented by counsel, the voluntariness of the plea depends on whether counsel’s advice was within the range of competence demanded of attorneys in criminal cases.” (citations omitted) (internal quotation marks omitted)).
The district court denied Reyes Campos’ motion to withdraw his guilty plea. The court ruled that Reyes Campos’ counsel had not provided ineffective assistance when he failed to inform Reyes Campos about the immigration consequences of his *485plea. The court determined that Padilla did not apply retroactively, but had instead announced “a new constitutional rule because it impose[d] a new requirement on counsel under the federal constitution.” The court relied on an earlier decision from our court, Alanis v. State, 583 N.W.2d 573 (Minn.1998), abrogated in part by Padilla v. Kentucky, 130 S.Ct. 1473, holding that counsel need not warn a client about the collateral deportation consequences of a guilty plea. The district court concluded that Reyes Campos had not proved that a manifest injustice occurred warranting withdrawal of his guilty plea, which was, the court found, accurate, voluntary, and intelligent. See Minn. R.Crim. P. 15.05, subd. 1 (noting that a plea may be withdrawn, even after sentencing, if “withdrawal is necessary to correct a manifest injustice”); see also Perkins v. State, 559 N.W.2d 678, 688 (Minn.1997) (“Manifest injustice occurs if a guilty plea is not accurate, voluntary, and intelligent, and thus the plea may be withdrawn.”). Reyes Campos filed a motion to reconsider, which the court also denied.
The court of appeals reversed. Reyes Campos v. State, 798 N.W.2d 565 (Minn.App.2011). The court ruled that Padilla applies retroactively to cases on collateral review, because Padilla did “not announce a new rule of criminal procedure.” Id. at 569. We granted the State’s petition for review.
The primary issue presented in this ease is whether the rule announced in Padilla applies retroactively to a claim of ineffective assistance of counsel raised on collateral review. Reyes Campos argues that under the Court’s holding in Padilla he received ineffective assistance of counsel rendering his plea uninformed and invalid because his attorney failed to inform him about the deportation consequences of his guilty plea. The State contends, however, that Padilla announced a new rule of eon-stitutional criminal procedure and, as a result, Padilla’s holding cannot be applied retroactively to Reyes Campos’ conviction, which became final several months before Padilla was decided. Alternatively, the parties raise the issue of whether Reyes Campos should be allowed to withdraw his guilty plea because the district court failed to provide him with the general immigration advisory required by Minn. R.Crim. P. 15.01, subd. 1(6X0-
We review a district court’s decision to deny a motion to withdraw a guilty plea under Minn. R. Civ. P. 15.05, subd. 1, for an abuse of discretion. See Barragan v. State, 583 N.W.2d 571, 572 (Minn.1998). But the question of whether Padilla applies retroactively to Reyes Campos’ conviction is a legal one that we review de novo. See State v. Houston, 702 N.W.2d 268, 270 (Minn.2005).
I.
Reyes Campos argues that Padilla applies retroactively to his claim of ineffective assistance of counsel raised on collateral review. To provide context for the analysis of the retroactivity question, we begin with a discussion of the law applicable to claims of ineffective assistance of counsel in the guilty plea context raised before Padilla, and then turn to an analysis of Padilla itself. With this background in mind, we then return to the retroactivity question. As set forth below, we conclude that Padilla does not apply retroactively.
A.
Before Padilla, to determine whether a criminal defendant received ineffective assistance of counsel in the context of a guilty plea, we applied the standard from Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). See Scruggs v. State, 484 N.W.2d 21, 25 (Minn.1992); see also Hill v. Lockhart, 474 U.S. *48652, 57-59, 106 S.Ct. 366, 88 L.Ed.2d 203 (1985) (applying the Strickland standard to challenges to guilty pleas based on ineffective assistance of counsel). Under Strickland, the defendant must show that “counsel’s representation fell below an objective standard of reasonableness.” 466 U.S. at 688, 104 S.Ct. 2052; see also King v. State, 562 N.W.2d 791, 795 (Minn.1997). The defendant must also demonstrate “a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Hill, 474 U.S. at 59, 106 S.Ct. 366; see also Premo v. Moore, — U.S. —, 131 S.Ct. 733, 744, 178 L.Ed.2d 649 (2011) (emphasizing that under Strickland’s prejudice prong in the guilty plea context, the prejudice question is whether there was a “reasonable probability that [the defendant] would not have entered his plea but for his counsel’s deficiency”).
Our precedent before Padilla held that the first prong of the Strickland test was not met in cases in which attorneys failed to advise their clients of the deportation consequences of guilty pleas. See Alanis v. State, 583 N.W.2d 573, 575, 578-79 (Minn.1998). In Alanis, we concluded that because deportation was merely a collateral consequence of a guilty plea, the defendant’s “attorney was under no obligation to advise him of the deportation possibility and, therefore, the failure to so inform him could not have fallen below an objective standard of reasonableness as required by Strickland.” Id. at 579. Courts around the country, state and federal, were nearly unanimous on this question, holding as we did in Alanis that the failure to advise a defendant about the collateral deportation consequences of a guilty plea was not ineffective assistance of counsel. See Chaidez v. United States, 655 F.3d 684, 690-91 (7th Cir.2011) (collecting federal cases holding that failure to warn a client of the deportation consequences of a guilty plea did not constitute ineffective assistance of counsel), cert. granted, — U.S. —, 132 S.Ct. 2101, 182 L.Ed.2d 867 (2012); Commonwealth v. Clarke, 460 Mass. 30, 949 N.E.2d 892, 898 (2011) (same); People v. Kabre, 29 Misc.3d 307, 905 N.Y.S.2d 887, 893-94 (N.Y.Crim.Ct.2010) (collecting federal and state cases); Gabriel J. Chin & Richard W. Holmes Jr., Effective Assistance of Counsel and the Consequences of Guilty Pleas, 87 Cornell L.Rev. 697, 706-08 (2002) (collecting state and federal cases holding that the right to effective assistance of counsel did not include the right to be informed about the collateral consequences of a guilty plea).
The Supreme Court, however, reached a different conclusion in Padilla v. Kentucky, — U.S. —, 130 S.Ct. 1473, 1487, 176 L.Ed.2d 284 (2010). Padilla had been a lawful permanent resident of the United States for more than 40 years when he pleaded guilty to the transportation of large quantities of marijuana, and as a result of his plea, he was facing deportation at the time he brought his postconviction claim. Id. at 1477-78. The Kentucky Supreme Court had dismissed Padilla’s request for postconviction relief on the basis of ineffective assistance of counsel because it concluded that deportation was a collateral consequence of Padilla’s guilty plea, and “collateral consequences are outside the scope of representation required by the Sixth Amendment.” Commonwealth v. Padilla, 253 S.W.3d 482, 483 (Ky.2008).
The United States Supreme Court observed that “deportation is a particularly severe ‘penalty.’” Padilla, 130 S.Ct. at 1481 (quoting Fong Yue Ting v. United States, 149 U.S. 698, 740, 13 S.Ct. 1016, 37 L.Ed. 905 (1893)). The Court explained that “recent changes in our immigration law have made removal nearly an automatic result for a broad class of noncitizen offenders.” Id. at 1478. Consequently, under contemporary law, “deportation is an integral part — indeed, sometimes the *487most important part — of the penalty that may be imposed on noncitizen defendants who plead guilty to specified crimes.” Id. at 1480.
After acknowledging that many lower courts had agreed with the analysis of the Kentucky Supreme Court that “collateral consequences are outside the scope of representation required by the Sixth Amendment,” the Court explained that its own jurisprudence had “never applied a distinction between direct and collateral consequences to define the scope of constitutionally ‘reasonable professional assistance’ required under Strickland.” Id. at 1481 (quoting Strickland, 466 U.S. at 689, 104 S.Ct. 2052). Without deciding more generally whether a collateral/direct consequences distinction would ever be appropriate, the Court determined that because of deportation’s “close connection to the criminal process [it is] uniquely difficult to classify [it] as either a direct or a collateral consequence. The collateral versus direct distinction is thus ill-suited to evaluating a Strickland claim concerning the specific risk of deportation.” Id. at 1482. Therefore, the Court held that “advice regarding deportation is not categorically removed from the ambit of the Sixth Amendment right to counsel,” and that Strickland applied to Padilla’s claim. Id.
In analyzing whether representation by Padilla’s counsel “fell below an objective standard of reasonableness,” the Court reiterated that the Strickland inquiry is “linked to the practice and expectations of the legal community: ‘The proper measure of attorney performance remains simply reasonableness under prevailing professional norms.’ ” Padilla, 130 S.Ct. at 1482 (quoting Strickland, 466 U.S. at 688, 104 S.Ct. 2052). The Court then looked to numerous authorities, including the American Bar Association, criminal defense organizations, and state bar associations that “universally require defense attorneys to advise as to the risk of deportation consequences for non-citizen clients,” to conclude that “[t]he weight of prevailing professional norms supports the view that counsel must advise her client regarding the risk of deportation.” Id. Based on “longstanding Sixth Amendment precedents, the seriousness of deportation as a consequence of a criminal plea, and the concomitant impact of deportation on families living lawfully in this country,” the Court held that in order to satisfy the reasonableness standard under Strickland, “counsel must inform her client whether his plea carries a risk of deportation.” Id. at 1486.
There is no dispute in this case that Reyes Campos’ counsel was ineffective under Padilla, because counsel did not advise Reyes Campos of the immigration consequences of his guilty plea. See id.5 But *488under our precedent at the time of Reyes. Campos’ plea, his counsel was not ineffective. See Alanis, 583 N.W.2d at 579. The question we must determine is whether Padilla applies retroactively to collateral review of Reyes Campos’ conviction. We turn to that question now.
B.
Reyes Campos’ conviction was final before Padilla was decided. See O’Meara v. State, 679 N.W.2d 334, 340 (Minn.2004) (discussing finality), overruled on other grounds by Danforth v. Minnesota, 552 U.S. 264, 128 S.Ct. 1029, 169 L.Ed.2d 859 (2008).6 Therefore, Reyes Campos cannot establish an ineffective assistance of counsel claim unless he can avail himself of retroactive application of the Supreme Court’s holding in Padilla. See Beard v. Banks, 542 U.S. 406, 411, 124 S.Ct. 2504, 159 L.Ed.2d 494 (2004) (noting that a ret-roactivity analysis only proceeds when the defendant’s conviction is final before the Supreme Court’s decision at issue is released); Lambrix v. Singletary, 520 U.S. 518, 527, 117 S.Ct. 1517, 137 L.Ed.2d 771 (1997).
We follow the retroactivity principles outlined in Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989), when considering whether a rule of federal constitutional law applies to a criminal conviction that was final when the rule was announced. See Danforth v. State, 761 N.W.2d 493, 499 (Minn.2009). Under Teague, after determining that the defendant’s conviction was final at the time of the Supreme Court decision in question, “we first ask whether the rule of federal constitutional criminal procedure is new, or whether it is merely a predictable extension of a pre-existing doctrine.” State v. Houston, 702 N.W.2d 268, 270 (Minn.2005). Old rules of federal constitutional criminal procedure apply “both on direct and collateral review, but a new rule is generally applicable only to cases that are still on direct review.” Whorton v. Bockting, 549 U.S. 406, 416, 127 S.Ct. 1173, 167 L.Ed.2d 1 (2007); accord Danforth, 761 N.W.2d at 496 (explaining that “if a case is pending on direct review when a new rule of federal constitutional criminal procedure is announced, the defendant is entitled to benefit from that new rule. But if the defendant’s conviction is already final at the time the new rule is announced, then the criminal defendant ordinarily may not avail himself of the new rule.”). A new rule applies retroactively in collateral proceedings under only two narrow exceptions: if the rule announced is substantive or if the rule is a “ ‘watershed rul[e] of criminal procedure’ implicating the fundamental fairness and accuracy of the criminal proceeding.” Saffle v. Parks, 494 U.S. 484, 495, 110 S.Ct. 1257, 108 L.Ed.2d 415 (1990) (quoting Teague, 489 U.S. at 311, 109 S.Ct. 1060).
1.
The threshold question under Teague is whether Padilla announced a new rule. Reyes Campos argues that Padilla did not announce a .new rule, and therefore it ap*489plies retroactively to his guilty plea. He asserts that Padilla is an old rule because it merely applied the “long-standing” principles of Strickland in the context of professional norms that now recognize the importance of immigration consequences in the context of a guilty plea. Moreover, Reyes Campos argues that because Padilla “itself arose in the context of a state post-conviction proceeding” Padilla,’s rule must be retroactive or “Jose Padilla himself could not have benefited from it because he only raised the issue of his trial counsel’s failure to advise him about the immigration consequences of his plea for the first time in his state post-conviction petition.”
The State, on the other hand, contends that Padilla was an application of Strickland in “a novel setting;” and it therefore constitutes a new rule. The State emphasizes that the near uniform agreement among state and federal courts prior to Padilla, holding that “an attorney’s failure to advise a defendant of the immigration consequences of his plea was not ineffective assistance of counsel because such advice related to a collateral matter,” indicates that reasonable jurists clearly would not have felt “compelled” by existing Supreme Court precedent to rule in Reyes Campos’ favor.
A Supreme Court “holding constitutes a new rule within the meaning of Teague if it breaks new ground, imposes a new obligation on the States or the Federal Government, or was not dictated by precedent existing at the time the defendant’s conviction became final.” Graham v. Collins, 506 U.S. 461, 467-68, 113 S.Ct. 892, 122 L.Ed.2d 260 (1993) (citation omitted) (internal quotation marks omitted) (requiring courts to look to the “legal landscape” as it existed at the time of the defendant’s conviction in determining whether a rule is new); see also Houston, 702 N.W.2d at 270. While recognizing that it has “stated variously the formula for determining when a rule is new,” O’Dell v. Netherlands 521 U.S. 151, 156, 117 S.Ct. 1969, 138 L.Ed.2d 351 (1997), the Supreme Court has emphasized that the Teague doctrine serves to “validate[ ] reasonable, good-faith interpretations of existing precedents made by state courts even though they are shown to be contrary to later decisions,” Butler v. McKellar, 494 U.S. 407, 414, 110 S.Ct. 1212, 108 L.Ed.2d 347 (1990). A rule therefore should not be applied to a defendant on collateral review under Teague “unless it can be said that a state court, at the time the conviction or sentence became final, would have acted objectively unreasonably by not extending the relief later sought in federal court.” O’Dell, 521 U.S. at 156, 117 S.Ct. 1969.7
The question of whether Padilla announced a new rule or was mérely an *490application of the Court’s longstanding Strickland analysis has resulted in a split among the United States Courts of Appeals. The Third Circuit in United States v. Orocio, held that Padilla “is retroactively applicable on collateral review.” 645 F.3d 630, 634 (3d Cir.2011). The Fifth, Tenth, and Seventh Circuits have reached the opposite conclusion, holding that Padilla announced a new rule of federal constitutional criminal procedure that is not applicable to defendants whose convictions became final before Padilla was decided. See United States v. Amer, 681 F.3d 211 (5th Cir.2012); United States v. Chang Hong, 671 F.3d 1147 (10th Cir.2011); Chaidez v. United States, 655 F.3d 684 (7th Cir.2011), cert. granted, — U.S. —, 132 S.Ct. 2101, 182 L.Ed.2d 867 (2012). Federal district courts and state courts, using an analysis similar to that of the circuit courts, have also struggled with the retroactivity question and have reached differing conclusions. Compare United States v. Agoro, Nos. CR 90-102 ML, CR 91-074 ML, 2011 WL 6029888, at *5-7 (D.R.I. Nov. 16, 2011) (collecting cases and concluding Padilla is not retroactive); Sarria v. United States, No. 11-20730-CIV, — F.Supp.2d —, —-—, 2011 WL 4949724, at *5-6 (S.D.Fla. Oct. 18, 2011) (same); Doan v. United States, 760 F.Supp.2d 602, 604-05 (E.D.Va.2011) (same), with United States v. Dass, Crim. No. 05-140(3), 2011 WL 2746181, at *2-4 (D.Minn. July 14, 2011) (collecting cases and concluding that Padilla announced an old rule and should apply retroactively to defendants on collateral review); Marroquin v. United States, No. M-10-156, 2011 WL 488985, at *2 (S.D.Tex. Feb. 4, 2011) (same); Commonwealth v. Clarke, 460 Mass. 30, 949 N.E.2d 892, 904 (2011) (same); People v. Garcia, 29 Misc.3d 756, 907 N.Y.S.2d 398, 404-05 (N.Y.Sup.Ct. 2010) (same).
After reviewing our retroactivity analysis under Teague, as well as decisions from other courts considering whether Padilla applies retroactively to cases on collateral review, we conclude that Padilla announced a new rule of federal constitutional criminal procedure. The Teague analysis requires us to determine if the decision in Padilla was compelled by precedent. See Lambrix, 520 U.S. at 538, 117 S.Ct. 1517 (explaining that the question is not whether the decision announced by the Supreme Court is a “reasonable interpretation of prior law,” but rather “whether no other interpretation was reasonable”). It is not enough that the decision was “‘controlled’ or ‘governed’ by prior opinions.” Butler, 494 U.S. at 415, 110 S.Ct. 1212. Even in that situation, where there are “reasonable contrary conclusions reached by other courts,” such that “the outcome ... was susceptible to debate among reasonable minds,” the rule announced is a new one. Id.; see also Houston, 702 N.W.2d at 271 (explaining that “it is not enough that a new constitutional rule of procedure is logically an extension of some precedent, as that is true of virtually all recently announced rules”; rather the result must have been “compelled by existing precedent”).
With this standard in mind, we agree with the Fifth, Seventh, and Tenth Circuits that the state of the law prior to the Padilla decision means that Padilla announced a new rule. Prior to Padilla, our court as well as “the lower federal courts, including at least nine Courts of Appeals, had uniformly held that the Sixth Amendment did not require counsel to provide advice concerning any collateral ... consequences of a guilty plea.” Chaidez, 655 F.3d at 690; see also Padilla, 130 S.Ct. at 1481 & n. 9 (acknowledging that the Kentucky Supreme Court was “far from alone in [its] view” that failure to warn a client of the deportation consequences of a guilty plea did not constitute ineffective assistance of counsel); Chang Hong, 671 F.3d *491at 1154 (“[E]leven federal circuits, more than thirty states, and the District of Columbia have held that lawyers need not explain collateral consequences [under the Sixth Amendment].” (citing Chin & Holmes, supra, at 699)); Clarke, 949 N.E.2d at 898 (explaining that “the Supreme Court in Padilla effectively changed the law in the nine circuit courts of the United States Court of Appeals that had previously addressed the issue”). This overwhelming “unanimity among the lower courts is compelling evidence that reasonable jurists reading the Supreme Court’s precedents [prior to Padilla ] could have disagreed about the outcome of Padilla.” Chaidez, 655 F.3d at 690; see also Caspari v. Bohlen, 510 U.S. 383, 393, 114 S.Ct. 948, 127 L.Ed.2d 236 (1994) (finding it persuasive in determining that a rule advocated by the petitioner was new that two federal courts of appeals and several state courts of last resort had reached different conclusions on the issue, clearly suggesting that “reasonable jurists reviewing our precedents” would not have felt compelled to reach the result urged by the petitioner). Prior to Padilla, it clearly would not have been “illogical or even a grudging application” of Strickland for a court to conclude that effective assistance of counsel did not require an attorney to inform his client of the deportation consequences of a guilty plea. See Butler, 494 U.S. at 415, 110 S.Ct. 1212.8
Our conclusion that Padilla announced a new rule also finds support in the multiple opinions produced by the decision. When the Supreme Court expresses an “array of views” in a case, this can also suggest *492“that the rule announced ... was, in light of [the] Court’s precedent, ‘susceptible to debate among reasonable minds.’ ” O’Dell, 521 U.S. at 159-60, 117 S.Ct. 1969 (quoting Butler, 494 U.S. at 415, 110 S.Ct. 1212); see also Beard, 542 U.S. at 415, 124 S.Ct. 2504 (explaining that “there is no need to guess” whether a rule was not compelled by precedent because in the case in question “four Justices dissented”); Houston, 702 N.W.2d at 272-73 (finding it persuasive that the Blakely decision “was issued by a closely divided five to four court,” in determining that Blakely announced a “new rule of constitutional criminal procedure unavailable for collateral use”).
The majority opinion in Padilla was joined by five justices, with two justices concurring and two justices in dissent. 130 S.Ct. 1473. Justice Alito identified Padilla’s holding as a “dramatic departure from precedent” and a “dramatic expansion of the scope of criminal defense counsel’s duties under the Sixth Amendment,” constituting a “new approach,” that “marks a major upheaval in Sixth Amendment law.” Id. at 1488-92 (Alito, J., concurring). Justice Scalia, in dissent, also described the majority’s holding as well outside the previous scope of the Court’s Sixth Amendment jurisprudence, explaining that “[w]e have never held ... that once counsel is appointed all professional responsibilities of counsel — even those extending beyond defense against the prosecution — become constitutional commands.” Id. at 1495 (Scalia, J., dissenting). Moreover, the dissent noted that in all previous cases regarding the right to effective assistance of counsel “[w]e have limited the Sixth Amendment to legal advice directly related to defense against prosecution of the charged offense,” and have never “required advice of counsel regarding consequences collateral to prosecution.” Id.; see also Chang Hong, 671 F.3d at 1155 (taking “the concurrence and dissent as support for our conclusion that reasonable jurists did not find the rule in Padilla compelled or dictated by the Court’s prior precedent”); Chaidez, 655 F.3d at 689 (“That the members of the Padilla Court expressed such an ‘array of views’ indicates that Padilla was not dictated by precedent.” (quoting O’Dell, 521 U.S. at 159, 117 S.Ct. 1969)). The fact that the Court itself was divided as to the import of its decision further confirms our determination that the result in Padilla was not dictated by Supreme Court precedent.
We acknowledge the Supreme Court’s caution that the mere existence of conflicting law or dissent does not necessarily mean that a rule is new. See Beard, 542 U.S. at 416 n. 5, 124 S.Ct. 2504 (noting that “[b]ecause the focus of the inquiry is whether reasonable jurists could differ as to whether precedent compels the sought-for rule, we do not suggest that the mere existence of a dissent suffices to show that the rule is new”). Courts conducting retroactivity analysis are to engage in their own independent examination of the Supreme Court’s precedent, and not simply rely on the interpretations of other courts. See Williams v. Taylor, 529 U.S. 362, 383, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (stating “[t]he often repeated language that Teague endorses ‘reasonable, good-faith interpretations’ by state courts is an explanation of policy, not a statement of law ... Teague does not direct federal courts to spend less time or effort scrutinizing the existing federal law, on the ground that they can assume the state courts interpreted it properly” (citations omitted) (internal quotation marks omitted)). Even under an independent “objective reading of the relevant cases,” however, we conclude that Padilla was not dictated by precedent. Stringer v. Black, 503 U.S. 222, 237, 112 S.Ct. 1130, 117 L.Ed.2d 367 (1992); see Chang Hong, 671 F.3d at 1156 (“Padilla extended the Sixth *493Amendment right to effective counsel and applied it to an aspect of a plea bargain previously untouched by Strickland,.”)', Chaidez, 655 F.3d at 691.
In urging us to conclude that Padilla announced an old rule, Reyes Campos argues that Padilla could not have announced a new rule because it relied on Strickland, a rule “which of necessity requires a case-by-case examination of the evidence.” Williams, 529 U.S. at 382, 120 S.Ct. 1495 (quoting Wright v. West, 505 U.S. 277, 308-09, 112 S.Ct. 2482 (Kennedy, J., concurring)). The Court has explained that when such a rule is at issue “we can tolerate a number of specific applications without saying that those applications themselves create a new rule.” Id. According to Reyes Campos, because Strickland developed “a rule designed for the specific purpose of evaluating a myriad of factual contexts, it will be the infrequent case that yields a result so novel that it forges a new rule, one not dictated by precedent.” Id.; see also Clarke, 949 N.E.2d at 900 (explaining that Padilla merely applied Strickland, which by its terms requires “that a court reviewing an ineffective assistance claim ‘must judge the reasonableness of counsel’s challenged conduct on the facts of the particular case, viewed as of the time of counsel’s conduct’” (quoting Strickland, 466 U.S. at 690, 104 S.Ct. 2052)).
We disagree with Reyes Campos’ suggestion that the case-by-case examination inherently required under Strickland categorically indicates that a Court decision applying Strickland will never be a new rule. See Chaidez, 655 F.3d at 693. The Court has made clear that even when the principles advanced by precedent “conceived of at a high level of generality ... could be thought to support” the outcome in a particular case, that does not preclude the conclusion that a rule announced is new. See Beard, 542 U.S. at 416, 124 S.Ct. 2504; see also Sawyer v. Smith, 497 U.S. 227, 236, 110 S.Ct. 2822, 111 L.Ed.2d 193 (1990) (explaining that even though the Court “d[id] not doubt that [its] earlier Eighth Amendment cases lent general support to the conclusion” reached in a subsequent case, that fact alone did not compel the conclusion that the rule announced in the subsequent case was old because “the test would be meaningless if applied at this level of generality”).
Moreover, new rules that do not overturn previous cases are often announced in the context of a previous case or line of precedent. That a new rule relied on a previous case, however, is not enough to show that the rule in question is an old one. In Butler v. McKellar, for example, the Court concluded that Arizona v. Roberson, 486 U.S. 675, 108 S.Ct. 2093, 100 L.Ed.2d 704 (1988), holding that the Fifth Amendment barred police-initiated interrogation following a suspect’s request for counsel in a separate investigation, was a new rule not dictated by precedent. 494 U.S. 407, 415, 110 S.Ct. 1212, 108 L.Ed.2d 347 (1990). Though Roberson had cited and relied upon Edwards v. Arizona, 451 U.S. 477, 101 S.Ct. 1880, 68 L.Ed.2d 378 (1981), which required police, during continuous custody, to refrain from all further questioning once an accused invokes the right to counsel on any offense, the Court held that the rule in Roberson was not dictated by Edwards. Butler, 494 U.S. at 415, 110 S.Ct. 1212. In so holding, the Court determined that even though the Roberson opinion “found Edwards controlling,” the “outcome in Roberson was susceptible to debate among reasonable minds,” and therefore constituted a new rule. Butler, 494 U.S. at 415, 110 S.Ct. 1212. Just as Roberson applied well-established Fifth Amendment jurisprudence in a new context, so too did Padilla apply Strickland *494in a new context, not dictated by earlier precedent.
And while we agree with Reyes Campos that Strickland generally sets forth the test relevant to assessing the effectiveness of counsel under the Sixth Amendment, this does not mean that all cases expanding on Strickland’s, jurisprudence will constitute old rules. Padilla is more than just a routine application of Strickland to a unique set of facts. Rather, unlike any previous decision, Padilla “requires a criminal defense attorney to provide advice about matters not directly related to their client’s criminal prosecution.” Chaidez, 655 F.3d at 693; see Chang Hong, 671 F.3d at 1156 (“Padilla is a new rule of constitutional law not because of what it applies — Strickland—but because of where it applies — collateral immigration consequences of a plea bargain.”).
Courts agreeing with Reyes Campos’ contention that Padilla did not constitute a new rule have explained that lower courts misinterpreted the Court’s Strickland precedent when they concluded that advice about the collateral consequences of a guilty plea did not fall within the Sixth Amendment right to the effective assistance of counsel. See Orocio, 645 F.3d at 638. Because “the Padilla Court noted that it had never applied a distinction between direct and collateral consequences to define the scope of constitutionally reasonable professional assistance required under Strickland,” these courts have concluded that “[t]he application of Strickland to the Padilla scenario is not so removed from the broader outlines of precedent as to constitute a new rule.” Orocio, 645 F.3d at 638 (citations omitted) (internal quotation marks omitted).
But the Supreme Court’s precedent pri- or to Padilla did, in fact, provide indications that a direct and collateral consequences distinction was an appropriate one for courts to draw in deciding ineffective assistance of counsel claims. Even though the Court had never expressly endorsed the collateral and direct consequences distinction that we and many other lower courts applied, Padilla was a dramatic departure from previous case law because it announced for the first time that the distinction between collateral and direct consequences of a conviction was inappropriate in the immigration context. Chaidez, 655 F.3d at 691; see also Caspari, 510 U.S. at 393, 114 S.Ct. 948 (concluding that a Double Jeopardy Clause rule was new because though previous cases “may not have foreclosed” the ruling, “neither did any of them apply the Clause in that context”). Earlier Supreme Court cases arguably had implicitly endorsed a distinction between direct and collateral consequences in holding that a defendant’s plea was voluntary when the defendant was “fully aware of the direct consequences” of the plea, making the inclusion of collateral consequences into the Sixth Amendment right to counsel context a departure from previous rulings. Chaidez, 655 F.3d at 691 (emphasis added) (quoting Brady v. United States, 397 U.S. 742, 747, 90 S.Ct. 1468, 25 L.Ed.2d 747 (1970)); see also United States v. Sambro, 454 F.2d 918, 922 (D.C.Cir.1971) (interpreting the statement in Brady to imply that a guilty plea is voluntary if a defendant lacks awareness only of the collateral consequences of a guilty plea, stating “[w]e presume that the Supreme Court meant what it said when it used the word ‘direct ’; by doing so, it excluded collateral consequences”). The language in Brady defining the voluntary nature of a defendant’s plea certainly suggests that Padilla’s holding was not the “sole reasonable interpretation of existing precedent.” Chaidez, 655 F.3d at 692 (emphasis added); see also Lambrix, 520 U.S. at 538, 117 S.Ct. 1517.
Finally, Reyes Campos argues that, based on the procedural posture in *495Padilla, coupled with references in the Padilla opinion to the ruling’s effect on collateral proceedings, we should conclude that Padilla did not announce a new rule. Specifically, Reyes Campos relies on the fact that Padilla challenged his conviction through a collateral postconviction process after his conviction was final for purposes of retroactivity. As Reyes Campos points out, “£j]ust as the Supreme Court found that the Kentucky courts erred in denying Mr. Padilla’s claim, it would necessarily find that the District Court in the instant case erred in deciding Reyes Campos[’] case on the same basis, if the instant case were to be appealed all the way to the Supreme Court” because “Reyes Campos would be in the exact same procedural posture as Mr. Padilla was before the Supreme Court.”9
We acknowledge that Padilla came to the Court on collateral review, and the Court afforded Padilla relief, which could suggest that the Court did not intend Padilla to announce a new rule. But neither Padilla nor the Commonwealth of Kentucky raised the issue of retroactivity in their briefs. See generally Brief for Respondent, Padilla v. Kentucky, — U.S. —, 130 S.Ct. 1473, 176 L.Ed.2d 284 (2010) (No. 08-651), 2009 WL 2473880; Brief for Petitioner, Padilla v. Kentucky, — U.S. —, 130 S.Ct. 1473, 176 L.Ed.2d 284 (2010) (No. 08-651), 2009 WL 1497552. Because the issue of retroactivity was not raised, we decline to speculate as to the Court’s intention on that issue simply based on the procedural posture of the case. See State v. Osborne, 715 N.W.2d 436, 447 n. 8 (Minn.2006) (declining to “speculate” how the United States Supreme Court would decide the question of whether Blakely errors were structural errors); State v. Richter, 270 Minn. 307, 310, 133 N.W.2d 537, 539 (1965) (declining to speculate on the question of whether Mapp v. Ohio should be applied retroactively and deferring resolution of the question to the Supreme Court).
*496We recognize that in Teague, a case that came to the Court on collateral review, the Court concluded that determining “ ‘whether a decision [announcing a new rule should] be given prospective or retroactive effect’ ” is a question that “ ‘should be faced at the time of [that] decision,’ ” and is an issue “properly treated as a threshold question.” 489 U.S. at 300, 109 S.Ct. 1060 (alterations in original) (quoting Paul J. Mishkin, Forward: the High Court, the Great Writ, and the Due Process of Time and Law, 79 Harv. L.Rev. 56, 64 (1965)). In so concluding, the Court expressed dissatisfaction with its earlier retroactivity jurisprudence, which did not always consider retroactivity as a threshold matter, in part because its analysis prior to Teague “led to unfortunate disparity in the treatment of similarly situated defendants on collateral review.” Id. at 305,109 S.Ct. 1060.
But the Court in later cases has backed away from its language in Teague that retroactivity is a “threshold question” to be decided at the time a new rule is announced. In Collins v. Youngblood, the Court held that “[although the Teague rule is grounded in important considerations of federal-state relations, we think it is not ‘jurisdictional’ in the sense that this Court, despite a limited grant of certiorari, must raise and decide the issue sua sponte.” 497 U.S. 37, 41, 110 S.Ct. 2715, 111 L.Ed.2d 30 (1990). Because the State had not relied on Teague in Collins, the Court addressed the petitioner’s collateral claims on the merits. Id.; see also Goeke v. Branch, 514 U.S. 115, 117, 115 S.Ct. 1275, 131 L.Ed.2d 152 (1995) (explaining that “application of Teague is a threshold question in a federal habeas case,” and that while “a court need not entertain the defense if the State has not raised it, a court must apply it if it was raised by the State” (citations omitted)).
The failure of the Commonwealth of Kentucky to raise the issue of retroactivity in Padilla constitutes the type of waiver contemplated by Collins. The Court has consistently stated that courts are free to raise the issue of retroactivity. See Schiro v. Farley, 510 U.S. 222, 229, 114 S.Ct. 783, 127 L.Ed.2d 47 (1994) (explaining that the Court “undoubtedly ha[s] the discretion” to raise and decide the Teague question); Curtis v. Duval, 124 F.3d 1, 5 (1st Cir. 1997) (explaining that the court may “raise the Teague issue on its own initiative if it believes that doing so will further the ends of justice”). But, in the face of a State’s waiver, the Court has never explicitly endorsed the retroactivity-by-silence test that Reyes Campos advocates. Indeed, because the Court has evinced a willingness to raise the issue of retroactivity on its own, or note that it is declining to decide the issue as a result of the State’s waiver, the Court’s silence on retroactivity provides at least as much evidence that the Court did not intend to make a determination regarding the retroactivity of Padilla, as that the Court presumed Padilla would apply retroactively.
The fact that Padilla brought a state postconviction petition, instead of a federal habeas petition, also reinforces our conclusion that we should not rely on the procedural posture of Padilla to divine the Court’s intention on retroactivity. The Teague analysis at its heart is meant to “ensure that state convictions comply with the federal law in existence at the time the conviction became final, and not to provide a mechanism for the continuing reexamination of final judgments based upon later emerging legal doctrine.” Sawyer, 497 U.S. at 234, 110 S.Ct. 2822; see also Houston, 702 N.W.2d at 271 (explaining that “[t]he principle of finality is key” in our retroactivity determinations). In keeping with this general purpose, the Court has clearly determined that states are free to *497“provide broader relief in their own post-conviction proceedings than required by [Teague ].” Danforth v. Minnesota, 552 U.S. 264, 277, 128 S.Ct. 1029, 169 L.Ed.2d 859 (2008). In Danforth, the Court explained that the rule established in Teague “was tailored to the unique context of federal habeas and therefore had no bearing on whether States could provide broader relief in their own postconviction proceedings than required by that opinion.” Id. Because “finality of state convictions is a state interest, not a federal one,” the Court determined that “[i]t is a matter that States should be free to evaluate, and weigh the importance of, when prisoners held in state custody are seeking a remedy for a violation of federal rights by their lower courts.” Id. at 280, 128 S.Ct. 1029.
Given that states are free to extend greater retroactive effect to new rules of constitutional criminal procedure than the Court gives on federal habeas review, the Court’s failure to discuss retroactivity in Padilla, a state postconviction proceeding, arguably has no effect on the Supreme Court’s ability to later determine that Padilla announced a new rule. Kentucky’s waiver of the retroactivity issue in Padilla may suggest instead that the Commonwealth was willing to forego determination of whether Padilla announced a new rule, and apply the rule announced to Padilla’s postconviction petition in any case. Even if the Supreme Court later determines that Padilla announced a new rule that cannot be applied retroactively to federal habeas petitioners, Kentucky will remain free to give greater retroactive effect to Padilla and apply it retroactively to its own defendants on collateral review. Moreover, in the context of state postconviction proceedings, Teague’s policy that similarly-situated defendants should be treated alike has less force. Indeed, states under Dan-forth are allowed to provide more retroactive relief in collateral proceedings, so defendants in different states may receive different access to the benefits of a newly-announced rule of constitutional criminal procedure, at least when state retroactivity rules exceed, the scope of retroactivity provided by Teague.
For all of these reasons we conclude that Padilla announced a new rule of criminal procedure. Because we determine that Padilla announced a new rule of constitutional criminal procedure, it can only be applied to Reyes Campos’ conviction on collateral review if Padilla’s, rule falls within one of the two narrow exceptions to the general retroactivity principles in Teague.
2.
We will only apply new rules retroactively in collateral proceedings if the rule announced is substantive or the rule is a “ ‘watershed rul[e] of criminal procedure’ implicating the fundamental fairness and accuracy of the criminal proceeding.” Saffle v. Parks, 494 U.S. 484, 495, 110 S.Ct. 1257, 108 L.Ed.2d 415 (1990) (quoting Teague, 489 U.S. at 311, 109 S.Ct. 1060). It is clear that the rule announced in Padilla was procedural, not substantive. See Schriro v. Summerlin, 542 U.S. 348, 353, 124 S.Ct. 2519, 159 L.Ed.2d 442 (2004) (explaining that “rules that regulate only the manner of determining the defendant’s culpability are procedural”). Therefore, the only exception that could provide Reyes Campos the relief he seeks is if Padilla announced a watershed rule.
The Court has repeatedly emphasized that the watershed “exception is ‘extremely narrow,’ ” and since its decision in Teague has “rejected every claim that a new rule satisfied the requirements for watershed status.” Whorton, 549 U.S. at 417-18, 127 S.Ct. 1173 (quoting Schriro, 542 U.S. at 352, 124 S.Ct. 2519). Indeed, the Court has indicated that “it is unlikely that any” watershed rules have “yet to emerge.” Schriro, 542 U.S. at 352, 124 *498S.Ct. 2519 (quoting Tyler v. Cain, 533 U.S. 656, 667 n. 7, 121 S.Ct. 2478, 150 L.Ed.2d 632 (2001)).10 A new rule must both be “necessary to prevent an impermissibly large risk of an inaccurate conviction” and “alter our understanding of the bedrock procedural elements essential to the fairness of a proceeding,” in order to qualify as a watershed rule under the Supreme Court’s jurisprudence. Whorton, 549 U.S. at 418,127 S.Ct. 1173. To come within the watershed exception, the rule must institute procedures “implicit in the concept of ordered liberty,” Graham v. Collins, 506 U.S. 461, 478, 113 S.Ct. 892, 122 L.Ed.2d 260 (1993) (citations omitted) (internal quotation marks omitted), and “it is not enough that a new rule ‘is aimed at improving the accuracy of trial,’ or even that it promotes ‘[t]he objectives of fairness and accuracy.’ ” United States v. Mandanici, 205 F.3d 519, 528 (2d Cir.2000) (quoting Sawyer, 497 U.S. at 242, 110 S.Ct. 2822; Saffle, 494 U.S. at 495, 110 S.Ct. 1257). The only case that has ever satisfied this high threshold is Gideon v. Wainwright, in which the Court “held that counsel must be appointed for any indigent defendant charged with a felony.” Whorton, 549 U.S. at 419, 127 S.Ct. 1173 (citing Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963)).
Given the long line of precedent rejecting important new rules as “watershed rulings,” Padilla’s new interpretation of the right to effective assistance of counsel does not qualify as a rule that goes to the heart of a fair proceeding. See, e.g., Schriro, 542 U.S. at 356-57, 124 S.Ct. 2519 (rejecting the contention that Ring v. Arizona, 536 U.S. 584, 122 S.Ct. 2428, 153 L.Ed.2d 556 (2002), which held that a sentencing judge, sitting without a jury, may not find an aggravating circumstance necessary for imposition of the death penalty, announced a watershed rule of criminal procedure); Beard, 542 U.S. at 420, 124 S.Ct. 2504 (declining to find that Mills v. Maryland, 486 U.S. 367, 108 S.Ct. 1860, 100 L.Ed.2d 384 (1988), which announced a new rule invalidating capital sentencing schemes requiring juries to disregard mitigating factors not found unanimously, constituted a watershed ruling); Houston, 702 N.W.2d at 273 (determining that Blakely v. Washington, 542 U.S. 296, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004), had not announced a watershed rule because it did not “impact the accuracy of an underlying determination of guilt or innocence,” but instead only “modifie[d] the manner in which certain factors — those factors justifying upward durational departures ... must be treated”). Requiring counsel to inform his noncitizen client of the immigration consequences of a guilty plea does not decrease the risk of an inaccurate conviction. Padilla is only implicated “in cases where the defendant admits guilt and pleads guilty.” Chang Hong, 671 F.3d at 1158. In such cases, “because the defendant’s guilt is established through his own admission ... Padilla is simply not germane to concerns about risks of inaccurate convictions or *499fundamental procedural fairness.” Id. Moreover, Padilla’s holding, unlike the expansive rule in Gideon establishing a right to counsel in all felony cases, affects only a small subset of defendants, indicating that the rule does not have a fundamental and profound impact on criminal proceedings generally. See Mandanici, 205 F.3d at 528 (explaining that a watershed rule must institute “a ‘sweeping’ change that applies to a large swathe of cases rather than a ‘narrow right’ that applies only to a ‘limited class’ of cases” (quoting O’Dell, 521 U.S. at 167, 117 S.Ct. 1969)); see also Ellis v. United States, 806 F.Supp.2d 538, 549 (E.D.N.Y.2011) (concluding that the rule announced in Padilla was not a watershed rule because “the.rule has nothing to do with the accuracy of a defendant’s conviction,” applied “a relatively narrow holding,” and “only applies to a limited class of defendants — noncitizen defendants who face charges that carry with them immigration consequences”).
Because Padilla is a new rule of constitutional criminal procedure, but is not a watershed rule, we hold that Reyes Campos may not avail himself of the retroactive application of Padilla. In addition, because Minnesota law governing the standard for effective assistance of counsel at the time Reyes Campos’ conviction became final did not require counsel to inform a defendant of the immigration consequences of pleading guilty, we hold that Reyes Campos did not receive ineffective assistance of counsel.
II.
Finally, we turn to Reyes Campos’ contention that even if Padilla is not retroactively applicable to his guilty plea, he is entitled to relief on the additional basis that he did not receive the immigration advisory from the district court at his plea hearing. See Minn. R.Crim. P. 15.01, subd. l(6)(i). Reyes Campos further argues that because Rule 15.01, subd. l(6)(i) did not become effective until January 1, 1999, a year after our ruling in Alanis v. State, 583 N.W.2d 573 (Minn.1998), abrogated in part by Padilla v. Kentucky, 130 S.Ct. 1473, the rationale from Alanis should not apply where the immigration advisory required by the rules was not given to the defendant.11
The court of appeals declined to reach the issue of whether the failure to receive the Rule 15.01 immigration advisory provided a basis for Reyes Campos to withdraw his guilty plea, concluding that the issue was waived in part based on its conclusion that Reyes Campos had not raised the issue on appeal. See Reyes Campos, 798 N.W.2d at 570 n. 2 (explaining that because Reyes Campos’ had only raised the issue in his motion to reconsider, the district court had never addressed the issue, and Reyes Campos had not raised the Rule 15 violation as an issue on appeal, the Rule 15 issue was waived and would not be considered by the court). The court of appeals’ conclusion that Reyes Campos failed to raise the issue on appeal was erroneous. The record reflects that Reyes Campos did raise the Rule 15 violation in his brief to the Court of Appeals, and the State never contended that the issue was waived.
But the court of appeals was correct in concluding that Reyes Campos raised the Rule 15 issue for the first time at the *500district court in his motion to reconsider. The State, however, is not contending that Reyes Campos waived the Rule 15 issue because he raised it for the first time in a motion to reconsider. To the contrary, the State argues to our court that “[t]he Minnesota Court of Appeals should have invalidated Respondent’s guilty plea on [the] basis” that “Respondent did not receive the Rule 15.01, subd. 1(6)(Z) warning prior to pleading guilty.” Based on the unique facts of this case, where the State apparently concedes error and is not arguing that the error was waived, we remand the question of whether Reyes Campos is entitled to withdraw his plea due to lack of compliance with Rule 15.01, subd. l(6)(i).
Reversed and remanded.
Dissenting, PAGE and ANDERSON, PAUL H., JJ.. Under Minn.Stat. § 609.24, "[wjhoever, having knowledge of not being entitled thereto, takes personal property from the person or in the presence of another and uses or threatens the imminent use of force against any person to overcome the person’s resistance or powers of resistance ... is guilty of robbery." Under Minn.Stat. § 609.229, subd. 2, "[a] person who commits a crime for the benefit of, at the direction of, in association with, or motivated by involvement with a criminal gang, with the intent to promote, further, or assist in criminal conduct by gang members is guilty of a crime.” In this section, the legislature also provided longer statutory máxi-mums for crimes committed for the benefit of a gang “than the statutory maximum for the underlying crime." Minn.Stat. § 609.229, subd. 3(a).
. The parties agree that Reyes Campos did not receive any information at the plea hear-mg regarding the effects of the plea on his immigration status. But at the plea hearing, the district court directed defense counsel to complete a plea petition with Reyes Campos. A standard plea petition contains the language: “My attorney has told me and I understand that if I am not a citizen of the United States this plea of guilty may result in deportation, exclusion from admission to the United States of America or denial of citizenship." Minn. R.Crim. P. 15, Appx. A, 27. The plea petition that defense counsel was instructed to prepare, however, does not appear in the district court record, nor does any indication of the alleged petition's contents. There is also no indication in the plea hearing transcript or elsewhere in the record that a plea petition was ever prepared and executed. Counsel for both parties indicated at oral argument before our court that no plea petition was, in fact, prepared prior to Reyes Campos' guilty plea.
. Reyes Campos’ conviction for robbery constitutes an aggravated felony under at least two provisions of the INA. Under the INA an aggravated felony includes “a crime of violence ... for which the term of imprisonment [is] at least one year.” 8 U.S.C. § 1101 (a)(43)(F). A "crime of violence” is further defined as either "an offense that has as an element the use, attempted use, or threatened use of physical force against the person or property of another,” or "any other offense that is a felony and that, by its nature involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.” 18 U.S.C. § 16 (2006). Additionally, "a theft offense (including receipt of stolen property) or burglaty offense for which the term of imprisonment [is] at least one year” constitutes an aggravated felony under the INA. 8 U.S.C. § 1101(a)(43)(G).
. This rule states that "[a]t any time the court must allow a defendant to withdraw a guilty plea upon a timely motion and proof to the satisfaction of the court that withdrawal is necessary to correct a manifest injustice.” Minn. R.Crim. P. 15.05, subd. 1. There is no argument made in this case that Reyes Campos' motion to withdraw his plea was not timely.
. The standard of professional reasonableness announced by Padilla requires that "when the deportation consequence [of a guilty plea] is truly clear,” counsel "give correct advice.” 130 S.Ct. at 1483. But recognizing the complexities of immigration law, the Court clarified that, "[w]hen the law is not succinct and straightforward,” rendering “the deportation consequences of a particular plea ... unclear or uncertain,” “a criminal defense attorney need do no more than advise a noncitizen client that pending criminal charges may carry a risk of adverse immigration consequences.” Id. If Padilla applies retroactively to Reyes Campos’ plea, he has met the first prong of Strickland, because the "deportation consequences” of his guilty plea were likely “sufficiently clear” under even a cursory reading of the INA to invoke counsel’s duty to "give correct advice.” Moreover, even if the immigration consequences of Reyes Campos' plea were unclear or uncertain his counsel did not discharge his duty under Padilla to advise Reyes Campos that his guilty plea could “carry a risk of adverse immigration consequences.” The State does not argue otherwise.
. Because Reyes Campos did not file a direct appeal, his conviction became final for retro-activity purposes when the time to file such an appeal had expired. The 90-day period for Reyes Campos to file a direct appeal began to run on July 10, 2009, when the district court entered a judgment of conviction and sentenced Reyes Campos. See Minn. R.Crim. P. 28.02, subd. 4(3)(a) (providing that an appeal must be filed "within 90 days after final judgment or entry of the order being appealed”); Minn. R.Crim. P. 28.02, subd. 2(1) (“A final judgment ... occurs when die district court enters a judgment of conviction and imposes or stays a sentence.”). Therefore, Reyes Campos' conviction became final for purposes of our retroactivity analysis on October 8, 2009 — several months before the Supreme Court issued its decision in Padilla.
. When the Court explicitly overrules an earlier holding, a new rule is clearly created. See Whorton, 549 U.S. at 416, 127 S.Ct. 1173 (finding that Crawford v. Washington, 541 U.S. 36, 124 S.Ct. 1354, 158 L.Ed.2d 177 (2004), announced a new rule, not applicable on collateral review to defendants whose convictions became final prior to the decision, because Crawford was "flatly inconsistent with the prior governing precedent, Roberts, which Crawford overruled"). The question , becomes "admittedly” more difficult when the Court applies a prior decision "in a novel setting, thereby extending the precedent.” Stringer v. Black, 503 U.S. 222, 228, 112 S.Ct. 1130, 117 L.Ed.2d 367 (1992) (recognizing that "[t]he interests in finality, predictability, and comity underlying our new rule jurisprudence may be undermined to an equal degree by the invocation of a rule that was not dictated by precedent as by the application of an old rule in a manner that was not dictated by precedent”). The mere fact that “a court says that its decision is within the 'logical compass' of an earlier decision, or indeed that it is 'controlled' by a prior decision, is not conclusive for purposes of deciding whether the current decision is a 'new rule' under Teague." Butler, 494 U.S. at 415, 110 S.Ct. 1212.
. We are not persuaded by the reasoning of the Third Circuit in attempting to explain away the plethora of lower court opinions holding contrary to Padilla. The Third Circuit observed that "case law need not exist on all fours to allow for a finding under Teague that the rule at issue was dictated by ... precedent.” Orocio, 645 F.3d at 639 (quoting Lewis v. Johnson, 359 F.3d 646, 655 (3d Cir. 2004)). The court explained that “Strickland did not freeze into place the objective standards of attorney performance prevailing in 1984, never to change again,” and noted that the "[l]ower court decisions not in harmony with Padilla were, with few exceptions, decided before 1995 and pre-date the professional norms that, as the Padilla court recognized, had long demanded that competent counsel provide advice on the removal consequences of a client's plea.” Orocio, 645 F.3d at 640; see also Marroquin, 2011 WL 488985, at *6 (holding that "Padilla abrogated the lower courts' decisions because of their ‘ill-suited’ distinction between direct and collateral consequences” and that “[h]ad the lower courts not dwelled in the direct-versus-collateral distinction, they would have necessarily applied Strickland to the ineffective assistance of counsel claims ... and would have considered the same professional legal standards the Supreme Court examined in reaching its conclusion”). But the "few exceptions” identified by the Third Circuit include the opinions of five circuit courts, reaffirming that counsel need not .provide advice about collateral immigration consequences, after passage of major immigration reforms in 1996. See Chaidez, 655 F.3d at 690-91 (citing cases); see, e.g., Santos-Sanchez v. United States, 548 F.3d 327, 337 (5th Cir.2008) (determining that enactment of the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 "has not so altered the nature of deportation as to render it a direct consequence of a guilty plea”), abrogated by Padilla, 130 S.Ct. 1473. Additionally, the Court has held that a decision announced a new rule for retroactivity purposes even when only a few lower courts had not reached the same result as that reached in the case announcing the new rule. See Caspari, 510 U.S. at 393-94, 114 S.Ct. 948 (finding it persuasive that a rule was new when "one Federal Court of Appeals and two state courts of last resort” had reached a holding contrary to the new rule); Butler, 494 U.S. at 415, 110 S.Ct. 1212 (noting that a rule was "susceptible to debate among reasonable minds” when the Fourth and Seventh Circuits had taken "differing positions” on the issue). The almost unanimous approach of the lower courts before Padilla clearly confirms that a reasonable reading of Strickland did not dictate Padilla's result.
. Moreover, Reyes Campos argues that language from the Padilla opinion dispelling Kentucky's concern that the Court's holding would open the floodgates to collateral challenges to guilty pleas, also supports the classification of Padilla as an old rule. The Court concluded that "[i]t seems unlikely that our decision today will have a significant effect on those convictions already obtained as a result of plea bargains.” Padilla, 130 S.Ct. at 1485 (emphasis added). The Court’s conclusion was based in part on the Court’s assumption that because the majority of professional norms over the past 15 years have "imposed an obligation on counsel to provide advice on the deportation consequences of a client's plea,” there would be few defendants in the past 15 years who had pled guilty without receiving counsel's advice regarding the potential immigration consequences of the plea. Id. Courts have relied on this language to suggest that the Court contemplated Padilla-based challenges by defendants whose convictions were already final on March 31, 2010. See Orocio, 645 F.3d at 641 (explaining that the Court’s discussion of opening the floodgates to claims from petitioners whose convictions were already final at the time of the decision indicated “that the Padilla Court anticipated the retroactive application of its holding on collateral review”); United States v. Hubenig, No. 6:03-mj-040, 2010 WL 2650625, at *7 (E.D.Cal. July 1, 2010) ("If the Court intended Padilla to be a new rule which would apply only prospectively, the entire 'floodgates' discussion would have been unnecessary.”).
In our view, the "floodgates” discussion in Padilla is an insufficient basis to conclude that Padilla applies retroactively. We agree with the Tenth Circuit that it is "unwise to imply retroactivity based on dicta — and abandon the Teague analysis entirely.” Chang Hong, 671 F.3d at 1159. Because Teague retroactivity analysis "exists to promote the finality of convictions by shielding them from collateral attacks mounted on new procedural rules of constitutional law,” it would "completely ignore this goal” to "imply retroactivity from an isolated phrase in a Supreme Court opinion.” Id.
. Though we are not bound by the Supreme Court’s determination of whether a rule constitutes a watershed rule and have expressly reserved the right to conduct our own fairness analysis, see Danforth, 761 N.W.2d at 500, we have never had the opportunity to define the extent to which our definition of fundamental fairness differs from that of the United States Supreme Court. The parties do not suggest, and we do not find, that the facts of this case require us to depart from persuasive Supreme Court precedent regarding whether a rule so affects the fundamental fairness of a criminal proceeding such that it qualifies as a watershed rule. Because we determine that the rule announced in Padilla clearly is not so essential to fundamental fairness that it warrants application to all defendants, even those on collateral review, we need not determine the full scope of our fairness exception.
. In Alanis, we held not only that the failure to advise a defendant of the deportation consequences of a guilty plea did not constitute ineffective assistance of counsel, but also that no warning of any kind about deportation consequences was required for a guilty plea to be intelligent. 583 N.W.2d at 578. We clarified that "[w]hile we have said that for a guilty plea to be intelligent the defendant must be aware of the consequences of pleading guilty, it is the direct consequences of the guilty plea to which we refer.” Id.