*748 A jury convicted Melvin Hiram Thomas II in 2003 for receiving a stolen vehicle ( Pen. Code, § 496d, subd. (a), unlabeled statutory citations refer to this code) and active participation in a criminal street gang (§ 186.22, subd. (a) ). 1 To support the gang conviction, the People offered a gang expert whose testimony included testimonial, out-of-court statements about the specific facts of Thomas's case.

On direct appeal, Thomas challenged the admissibility of the gang expert's testimony as testimonial hearsay which violated his Sixth Amendment confrontation rights under Crawford v. Washington (2004) 541 U.S. 36 , 124 S.Ct. 1354 , 158 L.Ed.2d 177 ( Crawford ). Another panel of this court concluded Crawford did not undermine the established rule that experts may testify about the bases of their opinions without running afoul of hearsay and confrontation clause problems because such evidence is not submitted for the truth of the matter, as our Supreme Court held in People v. Gardeley (1996) 14 Cal.4th 605 , 59 Cal.Rptr.2d 356 , 927 P.2d 713 ( Gardeley ). ( Thomas , supra , 130 Cal.App.4th at p. 1210 , 30 Cal.Rptr.3d 582 .) Subsequently, in People v. Sanchez (2016) 63 Cal.4th 665 , 686, 204 Cal.Rptr.3d 102 , 374 P.3d 320 ( Sanchez ), the California Supreme Court rejected the Gardeley rule and held introducing out-of-court testimonial statements about case-specific facts through an expert witness violates the confrontation clause (as interpreted in Crawford ) unless the person who made the statement is unavailable and the defendant had a prior opportunity for cross-examination.

*749 Thomas now petitions for a writ of habeas corpus, claiming his conviction for active participation in a criminal street gang is invalid after Sanchez , which established a new rule we should apply retrospectively to his case. We issued an order to show cause, and requested briefing on the issues whether Sanchez was retroactive and whether any error was harmless beyond a reasonable doubt. We conclude the Teague v. Lane (1989) 489 U.S. 288 , 310, 109 S.Ct. 1060 , 103 L.Ed.2d 334 ( Teague ) retroactivity *868 standard governing federal habeas petitions doesn't govern California habeas petitions, but conclude Sanchez isn't retroactive under the three-factor analysis set out by the California Supreme Court in In re Johnson (1970) 3 Cal.3d 404 , 90 Cal.Rptr. 569 , 475 P.2d 841 ( Johnson ) and other decisions. We therefore deny the petition.

I

FACTS

On the evening of May 15, 2003, Judith Barrera heard a man outside her house yell, "Fuck you, guys. E.Y.C.," an acronym for the Elsinore Young Classics gang. She went to the front of her house, where she saw Thomas and a codefendant in a parked car. The codefendant got out of the car, entered a truck belonging to a member of a rival gang, and drove off. 2

Barrera and her aunt pursued them and, about 10 or 15 minutes later, found the codefendant pushing the truck toward the gas pumps at a convenience store. They saw Thomas get out of the truck and enter the convenience store. Barrera yelled, "That's our truck" and called 911. The two men fled into a field behind the store, where sheriff's deputies later found them-concealed in the weeds-and arrested them.

At the jail, Thomas asked a sheriff's deputy why they had been arrested. The deputy said they'd been arrested for stealing a truck and said they also faced a gang enhancement because someone had yelled out, "Fuck you, E.Y.C." Thomas's codefendant said Thomas hadn't been present when he had yelled the acronym.

At trial, Riverside Sheriff's Officer Robert Kwan testified as a gang expert. He described the Elsinore Young Classics gang generally, focusing on its subgroups and primary activities. "They have the P.W.L.'s, which is the Pee-Wee Locos, the kids in the elementary school levels. They have the Tiny Winos, which is between 12 and 18 years old, which their acronym is T.W.S., *750 and then they have the Nite Owls, which are the guys that are 18 and older." Kwan said E.Y.C. was primarily a Latino gang, although some members were white. Its primary activities ranged "from graffiti to robbery, to burglary, to attempt murder, up to murder," and included stealing cars.

Kwan told the jury he believed Thomas and his codefendant were members of E.Y.C. gang and they committed the crime for its benefit. Kwan based his opinion Thomas was a gang member on his own "training and experience, reports written where [Thomas was] a suspect, times [Kwan had] contacted [Thomas] being in the presence of other gang members, when he was caught with [his codefendant]; also with-that day being caught with another gang member." Kwan said Thomas had admitted "to commit[ting] other crimes with other gang members." He mentioned specifically a 1992 robbery where Thomas and other gang members had stolen "some bikes and hats off some kids" and an incident in February 2002 when Kwan found Thomas in the concealed basement of a house where Kwan was searching for another E.Y.C. member suspected of an attempted murder. Kwan also mentioned seeing an incident report which said Thomas had been present at a knife fight or stabbing in 1995 involving another E.Y.C. member, though Thomas was not charged in connection with that incident. In addition, Kwan said he had talked with other E.Y.C. members about Thomas, and they had told him Thomas was a member of *869 E.Y.C. whose gang moniker was "Little Casper" or "Villain."

Kwan also said Thomas had numerous gang-related tattoos. "He's got 'Elsinore' on his neck, on his eyebrow; 'Y.C.' on his eyebrow; 'P.W.L.' on his head underneath his hair; 'Y.C.' on the back of his head. 'P.W.L.' on his arms; 'E.Y.C.' across his whole midsection and chest. Numerous other tattoos depicting 'South Side' or 'I.E.'; 'SUR,' S-U-R, 'Y.C.' on his hands." Thomas had the number "13" tattooed on his arm, which Kwan explained represents "the 13th letter of the alphabet is M, which is the EME. [T]he EME runs the southern faction of the prison system, in the state prison system ... If you are from the southern, you will tattoo 13, showing your affiliation to what they call 'South Side Surenos.' " Kwan said Thomas's head had been shaved when he was arrested so the tattoos on the back and side of his head were visible. In Kwan's opinion, that meant Thomas was still active in the gang; otherwise, he would have grown his hair out to conceal the tattoos.

Kwan said, based on his training and experience, Thomas and his codefendant had committed the crime for the benefit of E.Y.C. because it caused fear and intimidation for rival gang members. He mentioned that a week before the pickup was taken, someone had painted the codefendant's gang moniker on the fence next door to Barrera's house.

*751 On cross-examination, Kwan conceded Thomas did not appear in the photographs of E.Y.C. gang members shown to the jury. He also admitted he was not aware of any crimes Thomas had committed recently. He knew only that Thomas had committed crimes 11 and nine years earlier and had served two years in prison. Kwan said Thomas was about 30 years old, and E.Y.C. was a gang "predominantly geared towards younger age groups." Kwan said much of his expertise concerning Elsinore gangs had been provided by other Elsinore officers and deputies as well as by speaking with E.Y.C. gang members.

Kwan described having several conversations with gang members concerning Thomas. "[J]ust a lot of consensual conversation, you know. We just started talking and names get thrown out on who's who and monikers and-." He said he had not documented those conversations. The only record he had in his file concerning Thomas concerned the incident in February 2002 when Kwan found Thomas hiding in a concealed basement. Kwan did not have any field identification cards for Thomas. There was no record Thomas had bragged about committing any crimes. Thomas had not been charged with any gang enhancement in the much earlier robbery of bikes and hats.

The jury found Thomas guilty of receiving a stolen vehicle ( § 496d, subd. (a) ) and actively participating in a criminal street gang (§ 186.22, subd. (a) ), but not guilty of vehicle theft. In bifurcated proceedings, the jury found true the allegations Thomas had suffered a prior prison term (former § 667.5, subd. (b)(5) ), two prior serious convictions (§ 667, subd. (a) ), and three strike priors (§§ 667, subds. (c), (e), 1170.12, subd. (c)(2)(A) ). The trial court dismissed two of Thomas's strike priors on count 2 and sentenced him on count 2 to the aggravated term of six years as a second strike. However, the trial court refused to strike any of the priors on count 3 and sentenced Thomas to a concurrent term of 25 years to life plus an additional and consecutive five-year term for each of the two serious felony priors for a total sentence of 35 years to life.

On direct appeal, Thomas argued the gang expert's opinion that he was an active member of a criminal street gang violated the confrontation clause under Crawford , because it was based on testimonial hearsay. A panel of this court concluded it was *870 bound to follow the California Supreme Court's holding in Gardeley , supra , 14 Cal.4th at pp. 618-619, 59 Cal.Rptr.2d 356 , 927 P.2d 713 , that experts may testify as to their opinions on relevant matters and, if questioned, may relate the information and sources on which they relied in forming those opinions, even if those sources include testimonial hearsay. The panel also concluded Crawford does not undermine that established rule, "because an expert is subject to cross-examination about his or her opinions and additionally, the materials on which the expert bases his or her opinion are not elicited for the *752 truth of their contents ; they are examined to assess the weight of the expert's opinion." ( Thomas , supra , 130 Cal.App.4th at p. 1210 , 30 Cal.Rptr.3d 582 , italics added.) The panel therefore affirmed Thomas's conviction.

More than a decade later, the California Supreme Court revisited the issue in Sanchez and held the Crawford rule limits expert witnesses from relating case-specific testimonial hearsay in explaining the basis for their opinions. Thomas filed a petition for writ of habeas corpus and, after soliciting briefs and appointing counsel for Thomas, we issued an order to show cause why relief should not be granted.

II

DISCUSSION

A. Change in Law Regarding Expert Basis Testimony

In Sanchez , the California Supreme Court changed how courts in California are permitted to treat expert testimony about the basis of their opinions. "When any expert relates to the jury case-specific out-of-court statements, and treats the content of those statements as true and accurate to support the expert's opinion, the statements are hearsay. It cannot logically be maintained that the statements are not being admitted for their truth. If the case is one in which a prosecution expert seeks to relate testimonial hearsay, there is a confrontation clause violation unless (1) there is a showing of unavailability and (2) the defendant had a prior opportunity for cross-examination, or forfeited that right by wrongdoing." ( Sanchez , supra , 63 Cal.4th at p. 686 , 204 Cal.Rptr.3d 102 , 374 P.3d 320 .)

By so holding, the Supreme Court parted ways with a well-established line of authority. It rejected the view it had previously endorsed in Gardeley that an expert could rely on and report to the jury out-of-court statements without violating the rule against hearsay, because such a statement would be introduced to support the expert's opinion, rather than establish the truth of the out-of-court statement itself. ( Gardeley , supra , 14 Cal.4th at p. 618 , 59 Cal.Rptr.2d 356 , 927 P.2d 713 ["an expert witness whose opinion is based on such inadmissible matter can, when testifying, describe the material that forms the basis of the opinion"]; see also People v. Vy (2004) 122 Cal.App.4th 1209 , 1223, fn. 9, 19 Cal.Rptr.3d 402 ["experts-may give opinion testimony that is based upon hearsay, including conversations with gang members as well as with the defendant"].) Now, after Sanchez , an expert may introduce such testimony only "through an applicable hearsay exception ... [or] the evidence can be admitted through an appropriate witness and the expert may *753 assume its truth in a properly worded hypothetical question in the traditional manner." ( Sanchez , supra , 63 Cal.4th at p. 684 , 204 Cal.Rptr.3d 102 , 374 P.3d 320 .)

The Supreme Court's decision had additional fallout. When it recognized expert testimony reporting out-of-court statements not subject to an exception are hearsay, it introduced potential problems under the confrontation clause of the Sixth Amendment. "Under Crawford , ... if [such] hearsay was testimonial and" unless the declarant is unavailable and the defendant *871 had a prior opportunity to cross-examine the declarant or forfeited the right "defendant [w]ould have [to have] been given the opportunity to cross-examine the declarant or the evidence [w]ould have [to be] excluded. Improper admission of such prosecution evidence would also be an error of federal constitutional magnitude." ( Sanchez , supra , 63 Cal.4th at p. 685 , 204 Cal.Rptr.3d 102 , 374 P.3d 320 ; see also id. at pp. 679-680, 204 Cal.Rptr.3d 102 , 374 P.3d 320 .)

In this case, Thomas contends the gang expert related testimonial, case-specific, out-of-court statements to explain to the jury how he concluded Thomas was an active gang member. The People do not contest the claim. Admitting such testimony today would violate Thomas's confrontation rights. ( Sanchez, supra , 63 Cal.4th at pp. 694-697, 204 Cal.Rptr.3d 102 , 374 P.3d 320 [holding police reports, STEP notices, and field identification cards may be testimonial].) However, the trial court permitted the testimony many years before the Sanchez Court announced the change in the law, and under the prior rule endorsed in Gardeley there was no error. (See Thomas , supra , 130 Cal.App.4th at p. 1210 , 30 Cal.Rptr.3d 582 ["Thus, because the statements were not offered to establish the truth of the matter asserted, but merely as on one of the bases for an expert witness's opinion, the confrontation clause, as interpreted in Crawford , does not apply"].) Thomas asks us to decide the confrontation clause rule announced in Sanchez applies retroactively and find he would have escaped conviction on the gang charge had the trial court applied the Sanchez rule to his case.

B. Retroactivity

1. California, not federal, retroactivity law governs state habeas petitions

As our sister court recognized in a recent case which raised the same issues, deciding whether a change in criminal decisional law should apply retroactively on collateral review is a complex task that raises important issues "of freedom, justice, and pragmatism." ( In re Ruedas (2018) 23 Cal.App.5th 777 , 782, 233 Cal.Rptr.3d 555 ( Ruedas ).) We have studied the Ruedas opinion carefully, and commend its careful discussion of the convoluted history of retroactivity jurisprudence, including the divergent development of the law in the federal and California judicial systems. ( *754 Id. at pp. 792-799, 233 Cal.Rptr.3d 555 .) Nevertheless, we part ways with our colleagues in refusing to apply the retroactivity doctrine that governs federal habeas practice in favor of California's own retroactivity rules under the state habeas statute.

Both the federal and state retroactivity doctrines have their roots in Linkletter v. Walker (1965) 381 U.S. 618 , 85 S.Ct. 1731 , 14 L.Ed.2d 601 ( Linkletter ), where the U.S. Supreme Court refused to apply the new rule established in Mapp v. Ohio (1961) 367 U.S. 643 , 81 S.Ct. 1684 , 6 L.Ed.2d 1081 ( Mapp ) to cases already subject to final judgment. In Mapp , the Supreme Court had held courts must exclude evidence acquired in searches that violated the search and seizure clause of the Fourth Amendment. Before Linkletter , the Court had always "applied new constitutional rules to cases finalized before the promulgation of the [new] rule." ( Linkletter , at p. 628, 85 S.Ct. 1731 .) " Linkletter [citation] was the first instance in which the Court declined to apply a new doctrine respecting one of the provisions of the Bill of Rights retroactively for the benefit of a previously convicted defendant." ( Brown v. Louisiana (1980) 447 U.S. 323 , 327, 100 S.Ct. 2214 , 65 L.Ed.2d 159 .)

So, in *872 Eskridge v. Washington State Prison Board of Prison Terms & Paroles (1958) 357 U.S. 214 , 78 S.Ct. 1061 , 2 L.Ed.2d 1269 , the U.S. Supreme Court reached back to alter a 1935 conviction by applying a new rule requiring states to provide trial transcripts to indigents on appeal. 3 In Gideon v. Wainwright (1963) 372 U.S. 335 , 83 S.Ct. 792 , 9 L.Ed.2d 799 ( Gideon ), the Court applied the new rule that the state must appoint counsel to represent indigent felony defendants even though Gideon's conviction was final, and he had attacked the judgment under the federal habeas statute. Similarly, in Jackson v. Denno (1964) 378 U.S. 368 , 84 S.Ct. 1774 , 12 L.Ed.2d 908 , the Court applied a new rule against allowing juries to determine whether an allegedly coerced confession was voluntary, though the petitioner's conviction was final and he attacked the judgment under the federal habeas statute. (See also McNerlin v. Denno (1964) 378 U.S. 575 , 84 S.Ct. 1933 , 12 L.Ed.2d 1041 [directing appeals court to apply Jackson v. Denno retroactively].)

In Linkletter , however, the U.S. Supreme Court concluded its past approach was not required. "[T]he Constitution neither prohibits nor requires retrospective effect." ( Linkletter , supra , 381 U.S. at p. 629 , 85 S.Ct. 1731 .) As a result, the court reasoned, "we must ... weigh the merits and demerits in each case [of applying a new rule retroactively] by looking to the prior history of the rule in question, its purpose and effect, and whether retrospective operation will further or retard its operation." ( Ibid. ) The court emphasized that analysis *755 should include a consideration of the importance of the new rule as well as the importance of respecting the finality of past judgments.

To accomplish these goals, the court introduced a now familiar three-factor analysis for deciding whether the courts should give a new rule retroactive application. "[W]e must look to the purpose of the [new] rule; the reliance placed upon the [old] doctrine; and the effect on the administration of justice of a retrospective application of [the new rule]." ( Linkletter , supra , 381 U.S. at p. 636 , 85 S.Ct. 1731 ; see also Stovall v. Denno (1967) 388 U.S. 293 , 297, 87 S.Ct. 1967 , 18 L.Ed.2d 1199 ( Stovall ) [courts must consider "(a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards"].) In effect, the court directed lower courts to weigh the importance of the new rule and the scope of its impact against the disruption of enforcing it to past cases.

In the case of the rule requiring the exclusion of evidence gathered in illegal searches and seizures, the Supreme Court held these factors did not support retroactive application. First, the purpose of the rule was to deter future illegal searches by the police, not protect against wrongful convictions. "[A]ll of the cases ... requiring the exclusion of illegal evidence have been based on the necessity for an effective deterrent to illegal police action. [Citation.] We cannot say that this purpose would be advanced by making the rule retrospective. The misconduct of the police prior to [the new rule] has already occurred and will not be corrected by releasing the prisoners involved." ( Linkletter , supra , 381 U.S. at pp. 636-637, 85 S.Ct. 1731 .) Second, the court emphasized "interests in the administration of justice and the integrity of the judicial process" weighed against applying the exclusionary rule retroactively. "Hearings would have to be held on the excludability of evidence long since destroyed, misplaced or deteriorated. If it is excluded, the witnesses available at the time of the original trial will not be available or if located their memory *873 will be dimmed." ( Id. at p. 637, 85 S.Ct. 1731 .) The court concluded these interests outweighed the purpose of the rule, which "has no bearing on guilt." ( Id. at p. 638, 85 S.Ct. 1731 .)

Subsequent federal decisions applying the Linkletter three-factor analysis established the purpose of a new rule can be the overriding consideration. "[O]ur decisions establish that '[foremost] among these factors is the purpose to be served by the new constitutional rule,' [citation], and that we will give controlling significance to the measure of reliance and the impact on the administration of justice 'only when the purpose of the rule in question [does] not clearly favor either retroactivity or prospectivity.' " ( Brown v. Louisiana , supra , 447 U.S. at p. 328 , 100 S.Ct. 2214 .)

*756 Ensuring convictions are legitimate and factual is the purpose of the highest order. "Where the major purpose of new constitutional doctrine is to overcome an aspect of the criminal trial that substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts in past trials, the new rule has been given complete retroactive effect. Neither good-faith reliance by state or federal authorities on prior constitutional law or accepted practice, nor severe impact on the administration of justice has sufficed to require prospective application in these circumstances." ( Williams v. United States (1971) 401 U.S. 646 , 653, 91 S.Ct. 1148 , 28 L.Ed.2d 388 .) Still, "[t]he extent to which a condemned practice infects the integrity of the truth-determining process at trial is a 'question of probabilities.' " ( Stovall , supra , 388 U.S. at p. 298 , 87 S.Ct. 1967 .) As a result, the courts applied new procedural rules "only when an assessment of those probabilities indicates that the condemned practice casts doubt upon the reliability of the determinations of guilt in past criminal cases." ( Brown v. Louisiana , supra , 447 U.S. at p. 329 , 100 S.Ct. 2214 .)

In 1970, the California Supreme Court adopted the Linkletter approach to deciding whether a new rule applies retroactively in California state courts. ( Johnson , supra , 3 Cal.3d 404 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) The year before, the U.S. Supreme Court had decided convictions for failing to pay the federal marijuana transfer tax ( 26 U.S.C. § 4744 , subd. (a)(1) ) were unconstitutional because requiring someone to file the necessary paperwork violated the Fifth Amendment right against self-incrimination. ( Leary v. United States (1969) 395 U.S. 6 , 89 S.Ct. 1532 , 23 L.Ed.2d 57 ( Leary ).) Johnson filed a petition for a writ of habeas corpus challenging his two concurrent 10-year minimum sentences for two counts of selling marijuana with two priors, because one of the priors was a 1950 federal conviction for failing to pay the marijuana transfer tax disapproved in Leary . ( Johnson , at p. 407, 90 Cal.Rptr. 569 , 475 P.2d 841 .) The court held "[i]t is clear that petitioner could not now be convicted of the same offense for which he was convicted in 1950 [citation], and that petitioner may now challenge his sentence to the extent that it is augmented by an unconstitutional prior [citations]. The question[ ] remaining [is] whether Leary applies retroactively to render the 1950 prior unconstitutional." ( Id. at p. 410, 90 Cal.Rptr. 569 , 475 P.2d 841 .)

The California Supreme Court turned to federal case law and the Linkletter three-factor analysis to decide whether to give Johnson the benefit of the Leary rule. "The retrospective effect of a law-making opinion is to be determined by '(a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards.' [Citation.] It is also clear that the factors of reliance and burden on the administration of justice are of *874 significant relevance only when the question of retroactivity is a close one after the purpose of the new rule is considered." ( Johnson , supra , 3 Cal.3d at p. 410 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) *757 After surveying federal retroactivity case law, the court concluded "the more directly the new rule in question serves to preclude the conviction of innocent persons, the more likely it is that the rule will be afforded retrospective application. Further, if the rule relates to characteristics of the judicial system which are essential to minimizing convictions of the innocent, it will apply retroactively regardless of the reliance of prosecutors on former law, and regardless of the burden which retroactivity will place upon the judicial system." ( Johnson , supra , 3 Cal.3d at p. 413 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) In short, though the court recognized "the states are free to give greater retroactive impact to a decision than the federal courts choose to give," it nevertheless imported the federal approach whole, including its emphasis on guarding against false convictions over concerns with the finality of convictions. "The overwhelming concern ... with the relation of the rule in question to the reliability of the truth-determining process at trial is but a corollary to the ultimate test of the integrity of the judicial process: its capacity to ensure the acquittal of the innocent." ( Id. at pp. 415-416, 90 Cal.Rptr. 569 , 475 P.2d 841 .) The Court applied Leary retroactively because it held defendants, like Johnson, who asserted the Fifth Amendment right against self-incrimination against charges of failing to pay the federal marijuana tax are innocent as a matter of law. ( Johnson , at pp. 415-416, 90 Cal.Rptr. 569 , 475 P.2d 841 .)

Though it was widely adopted, the Linkletter approach never found unanimous support in the U.S. Supreme Court. Justices Black and Douglas dissented from the Linkletter opinion because they believed the court should have applied even the exclusionary rule retroactively. They objected they could not "understand why those who suffer from the use of evidence secured by a search and seizure in violation of the Fourth Amendment would be treated differently from those who have been denied other guarantees of the Bill of Rights." ( Linkletter , supra , 381 U.S. at p. 646 , 85 S.Ct. 1731 (dis. opn. of Black, J.).) They also objected to giving concerns about the finality of judgments any importance in deciding who should benefit from a new rule. "Linkletter, convicted in the state court by use of 'unconstitutional evidence,' is today denied relief by the judgment of this Court because his conviction became 'final' before Mapp was decided. Linkletter must stay in jail; Miss Mapp, whose offense was committed before Linkletter's, is free. This different treatment of Miss Mapp and Linkletter points up at once the arbitrary and discriminatory nature of the judicial contrivance utilized here to break the promise of Mapp by keeping all people in jail who are unfortunate enough to have had their unconstitutional convictions affirmed before June 19, 1961." ( Id. at p. 641, 85 S.Ct. 1731 .) In short, some justices adhered to the view that refusing to apply new constitutional rules retroactively worked injustice and could not be counterbalanced by the interest of maintaining the finality of convictions.

Justice Harlan later articulated, in a series of dissenting and concurring opinions, a thorough-going and ultimately influential critique of the Linkletter *758 approach. He noted the three-factor analysis had created a welter of inconsistent opinions. ( Desist v. United States (1969) 394 U.S. 244 , 258, 89 S.Ct. 1030 , 22 L.Ed.2d 248 (dis. opn. of Harlan, J.) ; see also Mackey v. United States (1971) 401 U.S. 667 , 692-693, 91 S.Ct. 1160 , 28 L.Ed.2d 404 (conc. opn. of Harlan, J.).) He then distinguished cases *875 where defendants seek the benefit of new rules introduced while their cases are still pending from cases where defendants seek to apply new rules in collateral attacks on final state convictions through the federal habeas statute. Justice Harlan urged new rules be applied in all cases still pending and subject to direct review when the rule is announced. ( Desist , at p. 258, 89 S.Ct. 1030 (dis. opn. of Harlan, J.).) However, he argued collateral attacks under the federal habeas statutes require a different approach. On federal habeas review, he argued federal courts should not apply new rules retroactively except when the rules (1) place certain kinds of primary, private individual conduct beyond the power of the state to proscribe, or (2) require the observance of procedures "implicit in the concept of ordered liberty." ( Mackey , at pp. 692-693, 91 S.Ct. 1160 (conc. opn. of Harlan, J.).)

The U.S. Supreme Court later largely adopted Justice Harlan's proposals. In United States v. Johnson (1982) 457 U.S. 537 , 102 S.Ct. 2579 , 73 L.Ed.2d 202 and Griffith v. Kentucky (1987) 479 U.S. 314 , 328, 107 S.Ct. 708 , 93 L.Ed.2d 649 , the Court adopted his view on the treatment of appellants seeking to benefit from new rules on direct review. In Griffith , the court held "a new rule for the conduct of criminal prosecutions is to be applied retroactively to all cases, state or federal, pending on direct review or not yet final, with no exception." ( Griffith , at p. 328, 107 S.Ct. 708 .) A few years later, a majority of the court endorsed Justice Harlan's view that new rules should only rarely be applied retroactively on federal habeas review. ( Teague , supra , 489 U.S. at p. 310 , 109 S.Ct. 1060 (plur. opn. of O'Connor, J.); id . at p. 317, 109 S.Ct. 1060 (conc. opn. of White, J.); id . at pp. 319-320, 109 S.Ct. 1060 (conc. opn. of Stevens, J.).) Though this portion of Teague had the support of only a plurality of the justices, the full Court subsequently endorsed the test. "A new rule applies retroactively in a collateral proceeding only if (1) the rule is substantive or (2) the rule is a ' "watershed rul[e] of criminal procedure" implicating the fundamental fairness and accuracy of the criminal proceeding.' " ( Whorton v. Bockting (2007) 549 U.S. 406 , 416, 127 S.Ct. 1173 , 167 L.Ed.2d 1 ( Whorton ).)

Teague emphasized the need to respect the finality of state criminal judgments as well as the importance of federal comity for state judgments. "We agree with Justice Harlan's description of the function of habeas corpus. '[T]he Court never has defined the scope of the writ simply by reference to a perceived need to assure that an individual accused of crime is afforded a trial free of constitutional error.' [Citation.] Rather, we have recognized that interests of comity and finality must also be considered in determining the *759 proper scope of habeas review." ( Teague , supra , 489 U.S. at p. 308 , 109 S.Ct. 1060 .) The Court then overruled the Linkletter line of cases, which gave primary importance to protecting against false convictions, and held "new constitutional rules of criminal procedure will not be applicable to those cases which have become final before the new rules are announced," unless they fall within one of two very narrow exceptions. ( Teague , at p. 310, 109 S.Ct. 1060 .) Those exceptions are cases where a new rule (i) places " 'certain kinds of primary, private individual conduct beyond the power of the criminal law-making authority to proscribe,' " or (ii) "requires the observance of 'those procedures that ... are "implicit in the concept of ordered liberty," ' " in the sense that the rule is a "watershed rule[ ] of criminal procedure." ( Id. at p. 311, 109 S.Ct. 1060 , quoting Mackey , supra , 401 U.S. 667 at pp. 692-693, 91 S.Ct. 1160 .)

The exception for watershed rules of criminal procedure is extraordinarily narrow. The Court has observed it is unlikely *876 "any such rules ' "ha[ve] yet to emerge," ' " and noted it had "rejected every claim that a new rule satisfied the requirements for watershed status." ( Whorton , supra , 549 U.S. at pp. 417-418, 127 S.Ct. 1173 .) "In order to qualify as watershed, a new rule must meet two requirements. First, the rule must be necessary to prevent 'an " 'impermissibly large risk' " ' of an inaccurate conviction. [Citations.] Second, the rule must 'alter our understanding of the bedrock procedural elements essential to the fairness of a proceeding.' " ( Id. at p. 418, 127 S.Ct. 1173 .) "[T]he only case that [the Supreme Court] ha[s] identified as qualifying under this exception" is Gideon , where "the Court held that counsel must be appointed for any indigent defendant charged with a felony. When a defendant who wishes to be represented by counsel is denied representation. Gideon held, the risk of an unreliable verdict is intolerably high." ( Whorton , at p. 419, 127 S.Ct. 1173 .) In Whorton , the U.S. Supreme Court held the new rule in Crawford , which plays a role in this case, does not qualify as a watershed rule of criminal procedure. ( Whorton , at pp. 419-420, 127 S.Ct. 1173 .)

Our colleagues in the Fourth District, Division Three concluded the federal Teague standard is nearly insurmountable and one the rule in Sanchez does not satisfy. ( Ruedas , supra , 23 Cal.App.5th at p. 798 .) We agree. If the Teague test determined whether Sanchez applies retroactively to Thomas's case, we would be constrained to conclude it does not. ( Ruedas , at p. 798, 233 Cal.Rptr.3d 555 .) "After all, Sanchez is but an extension of Crawford ; it merely considered the degree to which Crawford applies in the context of expert testimony," ( ibid. ), and in Whorton , the U.S. Supreme Court held the new Crawford rule is not a watershed rule of criminal procedure. ( Whorton , supra , 549 U.S. at pp. 419-420, 127 S.Ct. 1173 .) Unlike our sister court, however, we conclude the Teague standard does not govern because Thomas is here on a state habeas petition, not a federal habeas petition.

*760 The distinction is critical. It means the Teague decision is not binding on us. The U.S. Supreme Court reached the same conclusion in Danforth v. Minnesota (2008) 552 U.S. 264 , 128 S.Ct. 1029 , 169 L.Ed.2d 859 ( Danforth ). "Like Linkletter , Teague arose on federal habeas. Unlike in Linkletter , however, this procedural posture was not merely a background fact in Teague . A close reading of the Teague opinion makes clear that the rule it established 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. at p. 277, 128 S.Ct. 1029 , italics added.) The court concluded Teague "clearly indicates that [its] general rule of nonretroactivity was an exercise of this Court's power to interpret the federal habeas statute," so "it cannot be read as imposing a binding obligation on state courts." ( Danforth , at p. 278, 128 S.Ct. 1029 .) The California Supreme Court agrees we are " 'free to give greater retroactive impact to a decision than the federal courts choose to give.' " ( In re Gomez (2009) 45 Cal.4th 650 , 655, fn. 3, 88 Cal.Rptr.3d 177 , 199 P.3d 574 , quoting Johnson , supra , 3 Cal.3d at p. 415 , 90 Cal.Rptr. 569 , 475 P.2d 841 .)

The Danforth Court emphasized the Teague general rule against retroactivity in federal habeas proceedings bottoms out on comity. The Teague decision "justified the general rule of nonretroactivity in part by reference to comity and respect for the finality of state convictions. Federalism and comity considerations are unique to federal habeas review of state convictions. [Citation.] If anything, considerations of comity militate in favor of allowing state courts to grant habeas relief to a broader class of individual than is required by *877 Teague . And while finality is, of course, implicated in the contest of state as well as federal habeas, finality of state convictions is a state interest, not a federal one. It 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." ( Danforth , supra , 552 U.S. at pp. 279-280, 128 S.Ct. 1029 .)

We agree with the Danforth Court's analysis and conclude we must look for guidance to the California Supreme Court in evaluating whether to apply Sanchez retroactively. The Supreme Court of Missouri reached the same conclusion about its own courts in State v. Whitfield (2003) 107 S.W.3d 253 , 267. "It is up to each state to determine whether to apply the rule set out in Teague , to continue to apply the rule set out in Linkletter - Stovall , or to apply yet some other rule appropriate for determining retroactivity of a new constitutional rule to cases on collateral review. So long as the state's test is not narrower than that set forth in Teague , it will pass constitutional muster." ( Ibid. )

We conclude the three-factor balancing test articulated in Johnson still governs whether we should apply Sanchez retroactively when a petitioner *761 seeks state habeas review. Johnson was itself a state habeas case, and the California Supreme Court weighed the three Linkletter factors to determine whether a new rule would apply to a state habeas petition attacking a final state judgment. ( Johnson , supra , 3 Cal.3d at p. 410 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) The Supreme Court has never disavowed Johnson and California courts continue to apply its retroactivity analysis in the context of state collateral review. (See, e.g., In re Lucero (2011) 200 Cal.App.4th 38 , 45, 132 Cal.Rptr.3d 499 [applying the Johnson factors and giving retroactive effect to the new rule that the crime of shooting at an occupied vehicle merges with the crime of homicide]; In re Hansen (2014) 227 Cal.App.4th 906 , 918-919, 174 Cal.Rptr.3d 146 [same]; compare Lozano v. Diaz (9th Cir. 2014) 586 Fed.Appx. 413 [refusing to apply the same new rule on federal habeas review].)

For all these reasons, we reject the approach taken by the Ruedas court, applying the Teague standard to decide whether Sanchez applies retroactively in state habeas proceedings. 4 ( Ruedas , supra , 23 Cal.App.5th at pp. 797-798, 233 Cal.Rptr.3d 555 .) We are bound by our own Supreme Court's case law concerning retroactivity in state habeas proceedings. Even if we believed a different standard should apply-whether the Teague standard or a standard specially designed to be applied in the context of state habeas petitions-we do not have the power to adopt it. ( Auto Equity Sales, Inc. v. Superior Court (1962) 57 Cal.2d 450 , 455, 20 Cal.Rptr. 321 , 369 P.2d 937 .)

2. Under California law Sanchez created a new rule

Under Johnson and its progeny, "[t]o determine whether a decision should be given retroactive effect, the California courts first undertake a threshold inquiry: does the decision establish a new rule of law? If it does, the new rule may or may not be retroactive ... but if it does not, 'no question of retroactivity arises' because there is no material change in the law." ( People v. Guerra (1984) 37 Cal.3d 385 , 399, 208 Cal.Rptr. 162 , 690 P.2d 635 ( Guerra ).)

Decisions establish "new rules" when they depart from clear contrary *878 rules established in prior judicial decisions. In practice, that means decisions establish new rules when they (1) explicitly overrule a precedent of the California Supreme Court, or (2) disapprove a practice implicitly sanctioned by prior decisions of the Supreme Court, or (3) disapprove a longstanding and widespread practice expressly approved by a near-unanimous body of lower court authorities. ( In re Lucero, supra, 200 Cal.App.4th at p. 45 , 132 Cal.Rptr.3d 499 , citing Guerra , supra, 37 Cal.3d at p. 401 , 208 Cal.Rptr. 162 , 690 P.2d 635 .) *762 By contrast, decisions do not establish new rules if they merely "explain or refine the holding of a prior case, ... apply an existing precedent to a different fact situation, ... draw a conclusion that was clearly implied in or anticipated by previous opinions," or give effect to a statutory rule or existing binding decision of the United States Supreme Court. ( Guerra , supra , 37 Cal.3d at p. 399 & fn. 13, 208 Cal.Rptr. 162 , 690 P.2d 635 .) In such cases, the rule in question already exists, and the case law merely applies it or fills out its boundaries.

Though a judicial decision may articulate a rule for the first time, it is not a new rule for retroactivity purposes if there was no prior contrary rule. So, there is no question of the retroactive application of a judicial holding where a decision "resolve[s] a conflict between lower court decisions, or address[es] an issue not previously presented to the courts," including cases where "the issue was presented but not squarely decided in the prior opinions, or in which prior Court of Appeal decisions resolving it were vacated by grants of hearing in [the Supreme Court]." ( Guerra , supra , 37 Cal.3d at p. 400 , 208 Cal.Rptr. 162 , 690 P.2d 635 .) Similarly, a decision does not establish a new rule of law if the court is only giving effect " 'to a statutory rule that the courts had theretofore misconstrued [citation.]' 'Whenever a decision undertakes to vindicate the original meaning of an enactment, putting into effect the policy intended from its inception, retroactive application is essential to accomplish that aim.' " ( Woosley v. State of California (1992) 3 Cal.4th 758 , 794, 13 Cal.Rptr.2d 30 , 838 P.2d 758 .)

The rule articulated in Sanchez plainly constitutes a new rule under the standard articulated in these cases. In Gardeley , the California Supreme Court held an expert witness may inform the jury of the basis of her opinion even if that basis would in other circumstances be inadmissible as testimonial hearsay. The rationale was such expert basis testimony is not presented for the truth of the matter, but only to support the expert's conclusion. In Sanchez , the Supreme Court rejected its prior holding. It rejected the premise that such evidence could be admitted to support the expert's opinion without implicating its truth because the basis evidence wouldn't support the expert's conclusion if it weren't true. As a result, the Supreme Court reversed itself and held case-specific testimonial hearsay must satisfy the normal limitations on the introduction of hearsay under the rules of evidence and, where the hearsay is testimonial, under the U.S. Supreme Court's decision in Crawford . ( Sanchez , supra , 63 Cal.4th at pp. 683-684, 686, 204 Cal.Rptr.3d 102 , 374 P.3d 320 .)

In our view, that change constitutes an express overruling of formerly binding precedent. Indeed, a prior panel of this court regarded Gardeley as such when it upheld Thomas's conviction and sentence on direct appeal. ( Thomas , supra , 130 Cal.App.4th at p. 1210 , 30 Cal.Rptr.3d 582 .) But even if Sanchez did not *763 overrule Gardeley expressly, "it did 'disapprove' of that decision 'to the extent it suggested an expert may properly testify regarding case-specific out-of-court statements without *879 satisfying hearsay rules.' " ( Ruedas , supra , 23 Cal.App.5th at p. 799 , 233 Cal.Rptr.3d 555 , quoting Sanchez , supra , 63 Cal.4th at p. 686, fn. 13 , 204 Cal.Rptr.3d 102 , 374 P.3d 320 .) The upshot of the change has constitutional dimension because it implicates the U.S. Supreme Court's confrontation clause decision in Crawford . However you designate the change, like our colleagues in Division Three, we conclude "[b]y renouncing this practice, Sanchez created a new rule for purposes of state retroactivity analysis." ( Ruedas , at p. 799, 233 Cal.Rptr.3d 555 .)

3. Sanchez does not apply retroactively in state habeas cases

If the change in the law constitutes a new rule, we next determine whether it should apply retroactively in collateral proceedings by considering (a) the purpose to be served by the new rule, (b) the extent of reliance by law enforcement authorities on the old rule, and (c) the effect on the administration of justice of applying the new rule retroactively. ( Johnson , supra , 3 Cal.3d at p. 410 , 90 Cal.Rptr. 569 , 475 P.2d 841 ) That is, we weigh the new rule's importance and impact against the disruption that would be caused by applying the new rule to final cases where law enforcement, including prosecutors, relied on the old rule in investigating and prosecuting those cases originally.

If the new rule aims to deter future misconduct or to define procedural rights merely incidental to a fair determination of guilt or innocence, it will generally not be given retroactive effect. ( People v. Garcia (1984) 36 Cal.3d 539 , 549, 205 Cal.Rptr. 265 , 684 P.2d 826 , citing Donaldson v. Superior Court (1983) 35 Cal.3d 24 , 38, 39, 196 Cal.Rptr. 704 , 672 P.2d 110 .) On the other hand, if a decision goes to the integrity of the fact-finding process ( People v. Kaanehe (1977) 19 Cal.3d 1 , 10, 136 Cal.Rptr. 409 , 559 P.2d 1028 ), or "implicates questions of guilt and innocence" ( Pryor v. Municipal Court (1979) 25 Cal.3d 238 , 258, 158 Cal.Rptr. 330 , 599 P.2d 636 ), retroactivity is the norm.

The Ruedas court concluded the new Sanchez rule should not apply retroactively because it "simply established an evidentiary framework for analyzing the admissibility of expert basis evidence." ( Ruedas, supra, 23 Cal.App.5th at p. 800 , 233 Cal.Rptr.3d 555 .) We disagree with that characterization. Applying the criteria of Johnson, we conclude Sanchez should not apply retroactively despite the fact it relates to the integrity of the fact-finding process. We reach that conclusion because the new rule does not target deep problems with the factuality of convictions, but law enforcement relied heavily on the old rule and going back to apply the new rule to those old cases would be exceedingly disruptive.

*764 a. The purpose of the Sanchez rule

Of the three Johnson factors, the first factor is of primary importance and will generally control if the purpose of the new rule plainly favors retroactive or prospective application. ( Johnson, supra, 3 Cal.3d at p. 410 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) We determine the purpose of the new rule by examining its relationship to the reliability and integrity of the fact-finding process at trial. ( Id . at pp. 411, 416, 90 Cal.Rptr. 569 , 475 P.2d 841 .)

Sanchez established certain kinds of evidence commonly used in proving gang charges and enhancements not only are inadmissible, but their admission violates the accused's Sixth Amendment confrontation rights. Because the old rule permitted such evidence, some people convicted of gang charges may be in prison based only or largely on such testimonial hearsay. For example, some convictions, like Thomas's, *880 likely turned on fact-specific hearsay statements in police reports recounted to juries as true by gang experts.

As the U.S. Supreme Court explained in Crawford , "the principal evil at which the Confrontation Clause was directed was the civil-law mode of criminal procedure, and particularly its use of ex parte examinations as evidence against the accused." ( Crawford , supra , 541 U.S. at p. 50 , 124 S.Ct. 1354 .) In the civil law and early British common law systems, judicial officers could-at least in some cases-examine witnesses and report the testimony in later trial proceedings as evidence without requiring the actual witness to testify in front of the accused or ever submit to cross-examination. ( Id. at pp. 43-44, 124 S.Ct. 1354 .) One example "was the 1603 trial of Sir Walter Raleigh for treason. Lord Cobham, Raleigh's alleged accomplice, had implicated him in an examination before the Privy Council and in a letter. At Raleigh's trial, these were read to the jury." ( Id. at p. 44, 124 S.Ct. 1354 .) Though Raleigh objected and asked to be allowed to confront his accuser, confident he would recant, the court refused, the jury convicted, and Raleigh was sentenced to death. ( Ibid. )

The example is dramatic, but before Sanchez , experts could act in much the same way. Take the facts of Thomas's case. Officer Kwan testified at trial that he believed Thomas was a member of the Elsinore Young Classics gang based in part on reports saying Thomas had been present at prior crimes involving other E.Y.C. members. ( Thomas , supra , 130 Cal.App.4th at pp. 1206-1207, 30 Cal.Rptr.3d 582 .) This means Officer Kwan was permitted to recount to the jury statements given to other law enforcement officers during the investigation of those criminal activities. Effectively, these law enforcement officers filled the role of the judicial officers, which the U.S. Supreme Court identified as inspiring the confrontation principle. Experts play the role of delivering the statements recorded during law enforcement investigations, reporting them as factual without the actual witnesses ever being tested in the *765 crucible of cross-examination. To the extent Sanchez puts an end to that process, we conclude it has articulated a new rule related to the integrity of the fact-finding process which implicates questions of guilt and innocence. 5 ( Pryor v. Municipal Court , supra , 25 Cal.3d at p. 258, 158 Cal.Rptr. 330 , 599 P.2d 636 ; People v. Kaanehe , supra , 19 Cal.3d at p. 10 , 136 Cal.Rptr. 409 , 559 P.2d 1028 .)

That doesn't end our analysis, however, because Johnson and its progeny require us to balance the benefit of employing the new rule to fix past wrongful convictions against the harm to the administration of justice of undertaking that effort, especially in the face of law enforcement and prosecutorial reliance on the old rule. We start by noting that, though the Sanchez rule does implicate the factuality of convictions, "[t]he extent to which a condemned practice infects the integrity of the truth-determining process at trial is a 'question of probabilities.' " ( Stovall , supra , 388 U.S. at p. 298 , 87 S.Ct. 1967 .) Our Supreme Court has admonished, "the more *881 directly the new rule in question serves to preclude the conviction of innocent persons, the more likely it is that the rule will be afforded retrospective application." ( Johnson , supra , 3 Cal.3d at p. 413 , 90 Cal.Rptr. 569 , 475 P.2d 841 .) A corollary is the less directly the new rule precludes innocent convictions, the less likely we should apply it retroactively.

We conclude the connection between the Sanchez rule and avoiding wrongful convictions is significant, but not strong. This is not a case, like Gideon , where the old practice raises doubts about the fundamental fairness of past trials. Nor is it a case where following the old practice threatened to lead to the introduction of corrupt evidence supporting conviction, like compelled convictions. ( Jackson v. Denno , supra , 378 U.S. at pp. 386-387, 84 S.Ct. 1774 .) Rather, it's a case where the evidence supporting the expert's opinion lacks force. So, hearsay statements in a police report are not strong support for the expert's opinion because they aren't tested by confronting the declarants. However, as the Sanchez court emphasized, the remedy is not to exclude expert opinions that the accused were gang members; it's to allow experts to present their opinions without spelling out the hearsay basis. Alternatively, the prosecution may present a witness to provide the foundation for the expert's opinion, and ask the expert his opinion based on hypothetical questions that assume the other witness's testimony is true. ( Sanchez , supra , 63 Cal.4th at pp. 685-686, 204 Cal.Rptr.3d 102 , 374 P.3d 320 .) Thus, even giving Sanchez its most far reaching *766 effect, prosecutors would have been able to put on weakened expert testimony of gang membership. "Not every rule that 'tends incidentally' to avoid unfairness at trial must be accorded retroactive effect." ( Brown v. Louisiana , supra , 447 U.S. at p. 329 , 100 S.Ct. 2214 .) In our judgment, the connection with the factuality of convictions is attenuated and does not raise serious questions about the accuracy of guilty verdicts in past trials. (See Williams , supra , 401 U.S. at p. 653, 91 S.Ct. 1148 .)

b. Reliance and the administration of justice

That brings us to the questions of the degree to which law enforcement relied on the old rule and the burden on the administration of justice of applying the new rule retroactively to cases already final.

Before Sanchez , prosecutors reasonably relied on Gardeley and its predecessors in deciding how to present their cases to juries. The settled rule gave them no reason to expend scarce resources and expend scarce trial time developing and presenting additional witnesses who could testify about the contents of police reports, STEP notifications, or field identification cards. The Supreme Court of California had said they could present that evidence through gang experts, so they did. (E.g., People v. Sisneros (2009) 174 Cal.App.4th 142 , 153, 94 Cal.Rptr.3d 98 ; People v. Williams (2009) 170 Cal.App.4th 587 , 622, 88 Cal.Rptr.3d 401 ; In re I.M. (2005) 125 Cal.App.4th 1195 , 1207, 23 Cal.Rptr.3d 375 .) As the Ruedas court noted, " Gardeley alone was cited in over 2,000 appellate decisions between the time it was decided in 1996 and the time Sanchez was decided in 2016." ( Ruedas , supra , 23 Cal.App.5th at p. 801, fn. 9 , 233 Cal.Rptr.3d 555 .)

Approaching prosecutions in that way was a shortcut, but it wasn't always necessary. Had the Sanchez rule been in place at the time of these prosecutions, in many cases the prosecutors could have put on independent evidence to show what appeared in the reports, notifications, and cards. Having done so, they could have asked their gang experts to assume what that evidence indicated and respond to hypothetical questions to elicit their opinion *882 as to gang membership and activity. The California Supreme Court explicitly endorsed this practice in Sanchez . ( Sanchez , supra , 63 Cal.4th at p. 684 , 204 Cal.Rptr.3d 102 , 374 P.3d 320 ["Alternatively, the evidence can be admitted through an appropriate witness and the expert may assume its truth in a properly worded hypothetical question in the traditional manner"].) As a result, we conclude applying Sanchez retroactively to final cases would result in reopening thousands of cases in which the prosecution used a shortcut even though it could have obtained a conviction using other evidence.

We also believe applying the new rule retroactively to final cases would be too disruptive and costly. The alternative evidence prosecutors could have *767 used in cases tried years and even decades ago likely has gone stale. That means many of the people who would receive relief would be offenders who could have been convicted using the Sanchez rule originally, but could not be convicted using the same rule today, though not because of innocence. We must also consider the raw cost to the court system of reopening these thousands of final cases for retrial and subsequent appeals. We conclude the benefits of applying the new rule retroactively do not justify the disruption to the administration of justice.

The bottom line is the purpose of the new rule in Sanchez is to improve the integrity of criminal trials involving gang experts, but its effect is neither so fundamental nor so far-reaching as to justify applying it to cases already final, especially in view of prosecutors' reasonable and routine reliance on the old rule. We concede our decision will leave some number of people in prison who wouldn't have been convicted of gang-related crimes had Sanchez been the law when they were tried. We have to bite that bullet. Our conclusion results from the fact that the retroactivity rule gives importance to finality as well as factuality. For that purpose, Johnson provides a good, workable, though imperfect approach to deciding whether to apply new rules to final cases in state habeas challenges. Here, the analysis counsels against retroactive application. It is for our Supreme Court to decide whether it is preferable to adopt a rule that gives more weight to factuality than finality.

For all these reasons, we conclude we should not give the Sanchez rule retroactive effect.

III

DISPOSITION

We deny Thomas's petition for habeas corpus.

We concur:

McKINSTER, Acting P. J.

MILLER, J.

The opinion in the direct appeal incorrectly said the jury convicted him of receiving stolen property (§ 496, subd. (a) ). ( People v. Thomas (2005) 130 Cal.App.4th 1202 , 1207, 30 Cal.Rptr.3d 582 ( Thomas ).)

We take the facts about the incident and trial proceedings from our opinion in Thomas's direct appeal, Thomas , supra , 130 Cal.App.4th 1202 , 30 Cal.Rptr.3d 582 .

Griffin v. Illinois (1956) 351 U.S. 12 , 76 S.Ct. 585 , 100 L.Ed. 891 .

The Ruedas court also held, in the alternative, the new rule announced in Sanchez is not retroactive under Johnson 's three-factor analysis. ( Ruedas , supra , 23 Cal.App.5th at pp. 797-798, 233 Cal.Rptr.3d 555 .)

The Ruedas court concludes the Sanchez rule is not central to the integrity of the truth-determining process, because it "only pertains to a particular subset of testimony-that derived from expert witnesses," making "its impact on the accuracy of criminal trials ... less far reaching than Crawford ." ( Ruedas , supra , 23 Cal.App.5th at p. 801 , 233 Cal.Rptr.3d 555 .) That's the wrong way of looking at it. A rule is not central to the factuality of convictions because it affects a large universe of cases; it's central to the factuality of convictions if it protects against the conviction of innocent persons in those cases to which it applies .