¶1 This case presents the question of whether our holding in State v. Winterstein, 167 Wn.2d 620, 220 P.3d 1226 (2009), that the inevitable discovery doctrine is inconsistent with article I, section 7 of the Washington State Constitution, applies retroactively to judgments final when the case was decided. Because we hold that Winter-stein does not apply retroactively, we must also address Nadder Baron Haghighi’s claim that his appellate counsel provided ineffective assistance, which requires first determining whether this claim is timely. We affirm the Court of Appeals and hold that Winterstein announced a new rule and that Haghighi’s ineffective assistance of appellate counsel claim is time barred.
Facts and Procedural History
¶2 A jury convicted Haghighi of one count of theft in the first degree and seven counts of unlawful issuance of checks or drafts. The unlawful issuance of checks or drafts counts relate to checks drawn on an account he opened with Allstate Bank, which provides only Internet banking. The theft count is tied to Venture Bank, a local operation. The underlying issue in this personal restraint petition (PRP) involves the admissibility of evidence obtained by way of a warrant issued in Washington but faxed to Allstate’s offices in Illinois. The State does not dispute that this method of securing evidence from Illinois was improper.
¶3 Several victims initially identified Haghighi in a photo montage and provided copies of fraudulent checks he had written to them. Based on this information, police obtained a search warrant in Washington for Haghighi’s bank records with Allstate. However, Allstate’s office is in Illinois, and the detective faxed the warrant to Allstate’s offices there. A representative at Allstate then provided the records requested by the detective, which did not comply with Illinois law on the domestication of out-of-state warrants.
*439¶4 Before trial, Haghighi moved to suppress the records on the grounds that the issuing court lacked jurisdiction to issue a warrant that would be served out of state. He did not, however, challenge the finding of probable cause upon which the warrant was issued. Based on the improper search, he argued that the records should be excluded. Although the trial court found the warrant unenforceable in Illinois, it ruled Haghighi’s constitutional rights had been properly considered and that under the inevitable discovery doctrine, the Allstate records should not be excluded. The court also denied Haghighi’s request for an inevitable discovery evidentiary hearing.
¶5 After being convicted, Haghighi appealed. As relevant here, he argued that the State failed to prove it would have inevitably discovered the Allstate records and that the trial court erred in not allowing a hearing on the issue. In an unpublished decision, the Court of Appeals affirmed the convictions, concluding that the State would have inevitably discovered the Allstate records. State v. Haghighi, noted at 151 Wn. App. 1047 (2009).
¶6 When the Court of Appeals issued its decision, Haghighi’s appointed counsel, Casey Grannis, wrote to Haghighi telling him of the decision. He stated that he did not think a motion for reconsideration would be successful but that Haghighi could file one himself if he so chose. Grannis also discussed filing a petition for review. He noted that he had been assigned to represent Haghighi only in the Court of Appeals but that occasionally and at its discretion his office extended the representation to petitions for review. He promised to follow up with Haghighi about whether his office would file a petition for review but informed Haghighi that a petition for review was required if Haghighi planned to file a federal habeas corpus action.
¶7 On August 20, 2009, Grannis again wrote Haghighi, telling him about a case pending before this court on a competency issue similar to an issue in Haghighi’s case. He offered to file a petition for review on that one issue, unless *440this court issued an unfavorable ruling before the deadline for filing. He did not believe the other issues warranted a petition for review. Grannis also informed Haghighi that he could file only one petition for review, so all issues needed to be raised. If he wanted to raise additional issues, he should hire an attorney or file pro se.
¶8 On September 14, Grannis wrote to Haghighi informing him that this court had issued an unfavorable ruling on the competency issue, so his office would no longer be filing a petition for review. He stated, “My substantive involvement with your direct appeal is now over, although I formally remain your attorney for the direct appeal in the Court of Appeals until the mandate terminating review is issued.” Br. of Pét’r, App. E at 2. The mandate issued on September 25, 2009.
¶9 On December 22, Grannis wrote to Haghighi informing him of our decision in Winterstein, where we struck down the inevitable discovery doctrine as an exception to the exclusionary rule. He called this a “major change in the law” and told Haghighi that although he was no longer Haghighi’s attorney, Haghighi could file a PRP on the Winterstein issue, but generally only one PRP could be filed and it must be filed within one year. Pet’r’s Suppl. Br., App. D at 1.
¶10 On March 6, 2010, Haghighi filed a timely PRP arguing several theories, including the Winterstein issue, but did not include an ineffective assistance of appellate counsel theory. Roughly 10 months later, on December 22, the chief judge dismissed all of the issues, except for the Winterstein issue. The judge appointed Grannis’s firm to represent Haghighi, but on January 6, 2011, the firm notified the court it could not represent Haghighi due to a conflict, since ineffective assistance of counsel was an issue. The firm did not copy Haghighi on this letter or inform him of this issue.
¶11 On January 27, Nancy Collins from the Washington Appellate Project was appointed counsel. She filed an *441amended PRP and argued that the rule announced in Winterstein should apply to Haghighi and that Grannis had been ineffective in not presenting or preserving the inevitable discovery doctrine issue, especially with Winterstein before this court at the time. In a published opinion, the Court of Appeals rejected both arguments.1 It held that Winterstein announced a new rule that did not apply retroactively to Haghighi’s case, which was final at the time. Further, it held that the ineffective assistance claim was added beyond the one-year time limit for PRPs and that equitable tolling should not be applied. Therefore, the claim was time barred. We granted review of this decision.
Analysis
a. Retroactivity
¶12 Since Teague v. Lane, 489 U.S. 288, 109 S. Ct. 1060, 103 L. Ed. 2d 334 (1989), this court has consistently and repeatedly followed and applied the federal retroactivity analysis as established in Teague. Oddly, the concurrence/ dissent claims that Teague’s applicability is an “open question,” concurrence in part/dissent in part at 461, but, even before Teague, we “attempted from the outset to stay in step with the federal retroactivity analysis.” In re Pers. Restraint of St. Pierre, 118 Wn.2d 321, 324, 823 P.2d 492 (1992) (citing In re Pers. Restraint of Sauve, 103 Wn.2d 322, 326-28, 692 P.2d 818 (1985)).
¶13 In St. Pierre, we followed this precedent and adopted Teague, which we consistently applied in later cases. E.g., In re Pers. Restraint of Scott, 173 Wn.2d 911, 271 P.3d 218 (2012); In re Pers. Restraint of Eastmond, 173 Wn.2d 632, 272 P.3d 188 (2012); In re Pers. Restraint of Rhome, 172 Wn.2d 654, 260 P.3d 874 (2011); State v. Robinson, 171 Wn.2d 292, 253 P.3d 84 (2011); State v. Williams-Walker, 167 Wn.2d 889, 225 P.3d 913 (2010); State v. Kilgore, 167 *442Wn.2d 28, 216 P.3d 393 (2009); State v. Abrams, 163 Wn.2d 277, 178 P.3d 1021 (2008); In re Pers. Restraint of Domingo, 155 Wn.2d 356, 119 P.3d 816 (2005); State v. Evans, 154 Wn.2d 438, 114 P.3d 627 (2005); In re Pers. Restraint of Markel, 154 Wn.2d 262, 111 P.3d 249 (2005); In re Pers. Restraint of Hinton, 152 Wn.2d 853, 100 P.3d 801 (2004); State v. Hanson, 151 Wn.2d 783, 91 P.3d 888 (2004); State v. Summers, 120 Wn.2d 801, 846 P.2d 490 (1993); In re Pers. Restraint of Rupe, 115 Wn.2d 379, 798 P.2d 780 (1990). Similarly, the Court of Appeals has not questioned the applicability of the Teague framework. E.g., In re Pers. Restraint of Jagana, 170 Wn. App. 32, 282 P.3d 1153 (2012); In re Pers. Restraint of Haghighi, 167 Wn. App. 712, 276 P.3d 311, review granted, 175 Wn.2d 1021, 287 P.3d 595 (2012); In re Pers. Restraint of Hacheney, 169 Wn. App. 1, 288 P.3d 619 (2012); State v. Chetty, 167 Wn. App. 432, 272 P.3d 918 (2012).
¶14 Here, although Haghighi argues for abandonment of Teague, he — in the concurrence/dissent’s own words — “provide [s] no specific alternative.” Concurrence in part/dissent in part at 459 n.21. In an attempt to justify abandonment of our long-settled precedent, the concurrence/dissent points to Danforth v. Minnesota, 552 U.S. 264, 128 S. Ct. 1029, 169 L. Ed. 2d 859 (2008), where the United States Supreme Court held that Teague did not bind state courts. But nothing in Danforth is new. We recognized this proposition in 2005 before the Supreme Court ever considered Danforth. Evans, 154 Wn.2d at 448-49. Moreover, as the concurrence/dissent acknowledges, other states have continued to apply their own analyses despite Teague. Concurrence in part/dissent in part at 460 (“ ‘New York . . . has continued to rely on the [Linkletter v. Walker, 381 U.S. 618, 85 S. Ct. 1731, 14 L. Ed. 2d 601 (1965)] criteria, even post-Teague.’’ ” (alterations in original) (quoting People v. Marshall, No. 3018/1995, 2013 WL 1729392, at *2, 2013 N.Y. Misc. LEXIS 1602, *3-4 (N.Y. Sup. Ct. Apr. 19, 2013) (unpublished))). Finally, in Danforth itself the Court *443described the issue as whether “Teague constrains the authority of state courts to give broader effect to new rules,” and stated, “We have never suggested that it does, and now hold that it does not.” Danforth, 552 U.S. at 266. Thus, Danforth is not the groundbreaking case described by the concurrence/dissent but, rather, the United States Supreme Court’s acknowledgment of a principle long established in our country’s courts. The Teague framework is supported by roughly 25 years of precedent, and neither Haghighi nor the concurrence/dissent provide adequate basis for jettisoning such a firmly established principle of law.
¶15 Under the Teague analysis, a new rule of criminal procedure applies retroactively to all cases pending on direct review or not yet final. A new rule, however, will not apply retroactively to final judgments unless the rule places certain kinds of private conduct beyond the State’s power to proscribe or requires the observance of procedures implicit in the concept of ordered justice. A new rule is considered one that “ ‘breaks new ground or . . . was not dictated by precedent existing at the time the defendant’s conviction became final.’ ” Eastmond, 173 Wn.2d at 639 (quoting Teague, 489 U.S. at 301). Moreover, if “ ‘reasonable jurists could disagree on the rule of law, the rule is new.’ ” Eastmond, 173 Wn.2d at 639-40 (quoting Evans, 154 Wn.2d at 444).
¶16 Haghighi argues that he should gain the benefit of our decision in Winterstein, where we held that the inevitable discovery doctrine is inconsistent with article I, section 7 of the Washington State Constitution. Before Winter-stein, the Court of Appeals had upheld trial court decisions allowing evidence to be admitted under the inevitable discovery doctrine. E.g., State v. Avila-Avina, 99 Wn. App. 9, 17, 991 P.2d 720 (2000), abrogated by Winterstein, 167 Wn.2d 620; State v. Reyes, 98 Wn. App. 923, 930, 993 P.2d 921 (2000), abrogated by Winterstein, 167 Wn.2d 620. Consistent with this Court of Appeals precedent, the Court of Appeals in Haghighi’s case upheld the trial court’s denial of the motion to suppress based on the inevitable discovery *444doctrine. Thus, if Haghighi is entitled to benefit from Winterstein’s rule, he arguably may have been entitled to suppression of the Allstate records. However, Winterstein was published shortly after Haghighi’s conviction became final. Thus, he must show that Winterstein was dictated by existing precedent and therefore not a new rule.2
¶17 The decision and analysis in State v. O’Neill, 148 Wn.2d 564, 62 P.3d 489 (2003), guides our resolution of this issue. There, the suspect had been lawfully detained but not arrested when police searched his car. Although there were numerous arguments regarding the propriety of this search, one of the State’s arguments was that the evidence would have been inevitably discovered when the suspect was arrested. Contrary to Haghighi’s argument here, in O’Neill we declined to address the broader issue of whether the inevitable discovery doctrine was ever valid under Washington law. We stated that the rule “cannot be applied in these circumstances.” O’Neill, 148 Wn.2d at 592 (emphasis added). Further, we opted to “leave for another case the question whether the rule might apply in another context,” which was “a question we ha[d] not decided.” O’Neill, 148 Wn.2d at 592 n.ll. Thus, O’Neill recognized the issue of inevitable discovery was unresolved and left the question unanswered. Not until Winterstein did we again consider and resolve the issue. Because we expressly left the issue undecided in O’Neill and based on the Court of Appeals’ approach existing prior to Winterstein, Haghighi cannot establish that the “Winterstein rule” was an existing rule dictated by precedent. Since Haghighi’s conviction was final when we decided Winterstein, he is not entitled to benefit from the rule announced in that case.
¶18 Haghighi also argues that we should not apply Teague because this is solely a state law claim. But as noted above, we have been consistent in our application of the *445Teague framework. Moreover, no explanation is offered as to why our state would favor finality of judgments to a lesser extent than the federal system. Thus, we see no reason to depart from our established retroactivity analysis.
¶19 Finally, Haghighi attempts to justify retroactive application of Winterstein based on the Rules of Appellate Procedure and the Revised Code of Washington, specifically RAP 16.4(c)(4) and RCW 10.73.100. As applicable here, both rules apply the same standard and permit relief in a personal restraint petition if “[t]here has been a significant change in the law” that is “material to the conviction” and “sufficient reasons exist to require retroactive application” of the new rule. RAP 16.4(c)(4); RCW 10.73.100(6). By its terms, RAP 16.4 does not provide a separate avenue for relief. Rather, the rule is restricted to situations where the petitioner is entitled to relief under RCW 10.73.100. RAP 16.4(d). We have interpreted RCW 10.73.100 as a procedural rule that is entirely consistent with the federal retroactivity analysis. Abrams, 163 Wn.2d at 291. For these reasons, Haghighi is not entitled to relief based on a retroactive application of our decision in Winterstein under RAP 16.4 or RCW 10.73.100.
b. Timeliness of the Ineffective Assistance of Appellate Counsel Claim
¶20 In the alternative, Haghighi argues that his appellate counsel was ineffective for not appealing the inevitable discovery issue or for not preserving the issue until Winterstein was decided. The Court of Appeals rejected this claim as untimely because although Haghighi timely filed his initial PRP, his amended PRP adding the ineffective assistance of appellate counsel claim was not filed until more than one year after his appeal became final. Generally, a PRP is time barred when filed more than one year after the mandate issues, unless subject to an exception. RCW 10.73.090. No statutory exception exists, so Haghighi offers two distinct theories why we should reach the merits of his ineffective assistance of appellate counsel claim.
*446 ¶21 First, he argues that the ineffective assistance claim is not really a new claim because he is simply amending his original PRP, suggesting that this amendment should relate back to the original filing as might be allowed under CR 15(c). However, PRPs are governed by the Rules of Appellate Procedure, not the Civil Rules. Although the concurrence/dissent accuses us of adopting a new rule, we have consistently noted that the RAPs neither authorize nor prohibit amendments to PRPs and have allowed amendment only if made within the one-year statutory time limit. In re Pers. Restraint of Bonds, 165 Wn.2d 135, 140, 196 P.3d 672 (2008). The concurrence/dissent argues that we approved of untimely “amendment” in Bonds, but that case involved equitable tolling, an entirely different doctrine than the “relating back” procedure contemplated by CR 15(c). Nothing in Bonds suggests that we approved of allowing untimely amendments to relate back in a PRP. CR 15 is simply inapplicable to the current case.
¶22 Moreover, we already rejected this very argument in In re Personal Restraint of Benn, 134 Wn.2d 868, 952 P.2d 116 (1998), where we recognized that the RAPs neither provide for a “relating back” procedure analogous to CR 15(c) nor allow the petitioner to add a later untimely claim. Thus, under both Bonds and Benn, an “amended” PRP does not relate back to the original filing and any “amendment” or new claim must be timely raised.3
¶23 Haghighi relies on In re Personal Restraint of Wilson, 169 Wn. App. 379, 279 P.3d 990 (2012), as a case where a new legal theory was added after the time bar. However, the court in Wilson did not allow the claim to *447relate back but instead found it to be part of the original claim. There, defense counsel had proposed an erroneous jury instruction, and generally, review of such error is precluded under the invited error doctrine. But where the error is the result of ineffective assistance of counsel, review is not precluded. Thus, in order for the petitioner to prevail on his timely claim, he would have had to prove ineffective assistance of counsel, which made the claim “ ‘part and parcel’ ” of the original claim and not a new claim. Wilson, 169 Wn. App. at 387.
¶24 Further, the ineffective assistance claim is, in fact, a new and separate claim. While the timely Winterstein issue is intimately tied to the ineffective assistance claim, the latter is a much broader issue requiring examination of when the representation started and ended, a lawyer’s obligations to his clients, and whether Haghighi would have prevailed on appeal. Resolution of this issue rests on facts and legal theory entirely distinct from whether Winterstein applies retroactively. Thus, even if we were to allow the amended PRP to “relate back” to the original filing, the Winterstein issue is only part of the ineffective assistance of appellate counsel claim and, therefore, the claim did not arise out of the same “conduct, transaction, or occurrence” as required by CR 15(c).
¶25 Haghighi’s second theory is that the time bar should be equitably tolled so that the untimely ineffective assistance of appellate counsel claim can be added. We have recognized that the time bar in RCW 10.73.090 is not jurisdictional in nature and may, under limited circumstances, be subject to equitable tolling. Although this court has not previously settled what standard should be applied in this context, traditionally we have allowed equitable tolling when justice requires its application and when the predicates of bad faith, deception, or false assurances are met, and where the petitioner has exercised diligence in pursuing his or her rights. Bonds, 165 Wn.2d at 140-41.
¶26 In any context, the doctrine of equitable tolling is a narrow doctrine to be used only sparingly and not *448applicable more generally to “garden variety” claims of neglect. In determining whether the doctrine applies, its application must be consistent with the general purpose of the statute setting out the statute of limitations. Bonds, 165 Wn.2d at 140-41.
¶27 Consistent with the general rules and policies governing PRPs, we find it both unwise and unnecessary to expand the doctrine beyond the traditional standard. RCW 10.73.090’s time bar promotes finality of judgments, a principle especially important in this context because a petitioner cannot obtain federal habeas corpus relief until his or her judgment is final. Any lower standard would require the courts to constantly define the doctrine’s boundaries and call into question the statutorily established finality.
¶28 Moreover, the general framework governing PRPs shows that equitable tolling has a more limited role than exists in other contexts, which makes it necessary to adhere to a stricter standard. In a “normal” situation, equitable tolling might be the only way in which a party is not deprived of his or her remedy. In a collateral attack, however, the petitioner not only had the right to make numerous timely challenges in the form of appeals or other motions, but he or she can also take advantage of multiple other grounds for tolling the statute of limitations. These other grounds include those expressed in RCW 10.73.100, including newly discovered evidence, facial invalidity of the judgment and sentence, and double jeopardy violations. We have also recognized the possibility of the actual innocence doctrine, which provides perhaps the broadest equitable protections where justice requires.4 Thus, an expansive equitable tolling doctrine provides limited benefit to petitioners and adds unnecessary ambiguity in the law. Consistent with the narrowness of the doctrine’s applicability, principles of finality, and the multiple avenues available for *449postconviction relief, we apply the civil standard and require the predicates of bad faith, deception, or false assurances, which Haghighi does not establish.
¶29 Haghighi knew all the facts relevant to his ineffective assistance of counsel claim when he filed his initial appeal. He knew that Grannis had not raised the inevitable discovery doctrine in the Court of Appeals and that this court had subsequently held the doctrine unconstitutional under Washington law. Nothing prevented Haghighi from raising the ineffective assistance of appellate counsel claim when he initially filed his timely PRP. We agree with the Court of Appeals’ conclusion that Haghighi’s claims are time barred.5
Conclusion
¶30 We affirm the Court of Appeals and hold that Winterstein announced a new rule from which Haghighi does not benefit because his case was final when Winterstein was decided. Further, we hold that Haghighi’s ineffective assistance of appellate counsel is time barred.
Madsen, C.J., and Owens, Fairhurst, J.M. Johnson, and González, JJ., concur.In re Pers. Restraint of Haghighi, 167 Wn. App. 712, 276 P.3d 311, review granted, 175 Wn.2d 1021, 287 P.3d 595 (2012).
Haghighi does not argue that the Winterstein rule placed certain kinds of private conduct beyond the State’s power to proscribe or required the observance of procedures implicit in the concept of ordered justice.
The concurrence/dissent also cites In re Personal Restraint ofPirtle, 136 Wn.2d 467, 471 n.1, 965 P.2d 593 (1998), as a case where an untimely amendment was allowed to relate back to the initial filing. Though we did allow untimely issues to be brought, it is not clear that the court allowed the untimely PRP to “relate back” or if we equitably tolled the filing deadline. Regardless, we emphasized the extraordinary circumstances underlying the filing of the petition and that the death penalty was involved. The case does not cite CR 15(c), and there is certainly no indication that the court sought to overrule Benn on this point of law.
In re Pers. Restraint of Weber, 175 Wn.2d 247, 284 P.3d 734 (2012); In re Pers. Restraint of Carter, 172 Wn.2d 917, 263 P.3d 1241 (2011).
Even if the ineffectiveness of appellate counsel claim were considered, we find nothing ineffective in an attorney not raising a claim on appeal that had been rejected in the cases considering the issue and where no current case supported the theory.