concurring in the judgment.
While I could not disagree with the majority that sufficient evidence supported West's conviction, see, e.g., ante, at 295-297, I do not think the Court should reach that issue. We have often said that when the principles first developed in Teague v. Lane, 489 U. S. 288 (1989), pose a threshold question on federal habeas review, it is only after an answer favorable to the prisoner that a court should address the merits. See, e. g., Collins v. Youngblood, 497 U. S. 37, 40-41 (1990); Penry v. Lynaugh, 492 U. S. 302, 313, 329 (1989); Teague, supra, at 300 (plurality opinion). This habeas case begins with a Teague question, and its answer does not favor West. I would go no further.1
I
Under cases in the line of Teague v. Lane, supra, with two narrow exceptions not here relevant, federal courts conducting collateral review may not announce or apply a “new” rule for a state prisoner’s benefit, Butler v. McKellar, 494 U. S. 407, 412 (1990); Teague, supra, at 310 (plurality opinion), a new rule being one that was “not ‘dictated by precedent existing at the time the defendant’s conviction became final,’ ” Sawyer v. Smith, 497 U. S. 227, 234 (1990) (quoting Teague, supra, at 301 (plurality opinion)) (emphasis in original). Put differently, the new-rule enquiry asks “whether a state court considering [the prisoner’s] claim at the time his conviction became final would have felt compelled by existing precedent to conclude that the rule [the prisoner] seeks was required by the Constitution.” Saffle v. Parks, 494 U. S. 484, 488 *311(1990). Or, put differently yet again, if “reasonable jurists [might have] disagree^]” about the steps the law would take next, its later development will not be grounds for relief. Sawyer v. Smith, supra, at 234; see also Butler, supra, at 415 (“susceptible to debate among reasonable minds”).
The Teague line of cases reflects recognition of important “interests of comity and finality.” Teague, supra, at 308 (plurality opinion). One purpose of federal collateral review of judgments rendered by state courts in criminal eases is to create an incentive for state courts to “‘“conduct their proceedings in a manner consistent with established constitutional standards,” ’ ” Butler, supra, at 413 (quoting Teague, supra, at 306 (plurality opinion)), and “[t]he ‘new rule’ principle” recognizes that purpose by “validating] reasonable, good-faith interpretations of existing precedents made by state courts even though they are shown to be contrary to later decisions.” Butler, supra, at 414 (citing United States v. Leon, 468 U. S. 897, 918-919 (1984)).
The crux of the analysis when Teague is invoked, then, is identification of the rule on which the claim for habeas relief depends. To survive Teague, it must be “old” enough to have predated the finality of the prisoner’s conviction, and specific enough to dictate the rule on which the conviction may be held to be unlawful. A rule old enough for Teague may of course be too general, and while identifying the required age of the rule of relief is a simple matter of comparing dates, passing on its requisite specificity calls for analytical care.
The proper response to a prisoner’s invocation of a rule at too high a level of generality is well illustrated by our cases. In Butler, supra, for example, the prisoner relied on the rule of Arizona v. Roberson, 486 U. S. 675 (1988), which we announced after Butler’s conviction had become final. We held in Roberson that the Fifth Amendment forbids police interrogation about a crime after the suspect requests counsel, even if his request occurs in the course of investigating a *312different, unrelated crime. Id., at 682. Butler argued that he could invoke Roberson’s rule because it was “merely an application of Edwards [v. Arizona, 451 U. S. 477 (1981)],” in which we held that, if a person is in custody on suspicion of a crime, the police must stop questioning him about that crime once he invokes his right to counsel, id., at 484-485, “to a slightly different set of facts.” 494 U. S., at 414. We rejected this argument, saying that it “would not have been an illogical or even a grudging application of Edwards to decide that it did not extend to the facts of Roberson.” Id., at 415.
Likewise, in Sawyer, supra, the petitioner sought the benefit of Caldwell v. Mississippi, 472 U. S. 320 (1985), which had been announced after Sawyer’s conviction was final. We held in Caldwell that the Eighth Amendment prohibits resting “a death sentence on a determination made by a sen-tencer who has been led to believe that the responsibility for determining the appropriateness of the defendant’s death rests elsewhere.” Id., at 328-329., Sawyer argued that he was entitled to the benefit of Caldwell’s rule as having been “dictated by the principle of reliability in capital sentencing,” Sawyer, supra, at 236, which, he said, had been established by cases announced before his conviction became final, Eddings v. Oklahoma, 455 U. S. 104 (1982), and Lockett v. Ohio, 438 U. S. 586 (1978), among them. We rejected the argument, saying that
“the [Teague] test would be meaningless if applied at this level of generality. Cf. Anderson v. Creighton, 483 U. S. 635, 639 (1987) (‘[I]f the test of “clearly established law” were to be applied at this level of generality, . . . [pjlaintiffs would be able to convert the rule of qualified immunity that our cases plainly establish into a rule of virtually unqualified liability simply by alleging violation of extremely abstract rights’).” 497 U. S., at 236 (internal quotation brackets in original).
*313Although the principle that Sawyer invoked certainly “lent general support to the conclusion reached in Caldwell,” ibid., we said that “ ‘it does not follow that [Eddings and Lockett] compel the rule that [petitioner] seeks,”’ ibid, (second set of brackets in original) (quoting Saffle, supra, at 491).
In sum, our eases have recognized 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.” Stringer v. Black, 503 U. S. 222, 228 (1992). This does not mean, of course, that a habeas petitioner must be able to point to an old case decided on facts identical to the facts of his own. But it does mean that, in light of authority extant when his conviction became final, its unlawfulness must be apparent. Cf. Anderson v. Creighton, 483 U. S. 635, 640 (1987).
J — J
In this ease, the Court of Appeals overruled the Commonwealth’s Teague objection by saying that West merely claimed that the evidence had been insufficient to support his conviction, so that the result he sought was dictated by Jackson v. Virginia, 443 U. S. 307 (1979), a case announced before petitioner’s conviction became final for Teague purposes in 1980. 931 F. 2d 262, 265-267 (CA4 1991). Having thus surmounted Teague’s time hurdle, the court went on to say that “the evidence here consisted entirely of... the ... facts . . . that about one-third in value of goods stolen between December 13 and December 26, 1978, were found on January 10, 1979, in the exclusive possession of . . . West, coupled with [West’s] own testimony explaining his possession as having come about by purchases in the interval.” 931 F. 2d, at 268. Applied in this context, the court held, the unadorned Jackson norm translated into the more specific rule announced in Cosby v. Jones, 682 F. 2d 1373 (CA11 *3141982), which held that the evidence of unexplained or unconvincingly explained possession of recently stolen goods was not, without more, sufficient to prove theft, but must be weighed more exactly after asking five questions: (1) Was “the possession ... recent, relative to the crime”? (2) Was, a large majority of the stolen items found in the defendant’s possession? (8) Did the defendant attempt to conceal the stolen items? (4) Was the defendant’s explanation, “even if discredited by the jury,... ‘so implausible or demonstrably false as to give rise to positive evidence in favor of the government’”? and (5) Was there corroborating evidence supporting the conviction? 931 F. 2d, at 268 (quoting Cosby, supra, at 1383, n. 19).
Applying Cosby to the facts of this case, the Court of Appeals found that all five factors were either neutral or advantageous to West: (1) Two to four weeks elapsed between the theft and the possession described in testimony,2 a time period consistent with West’s explanation that he had bought the goods in the interval; (2) measured by value, a mere third of Cardova’s belongings surfaced in West’s possession; (3) the stolen items were found in plain view in West’s home; (4) while “there was no third person testimony corroborating [West’s] explanation and on cross-examination West exhibited confusion about the exact circumstances of some of the purehases[,]... he maintained his general explanation that he had purchased all the items at flea markets, and there was nothing inherently implausible about this explanation”; and, finally, (5) there was no evidence corroborating theft by West. 931 F. 2d, at 269-270. The Court of Appeals concluded that “the evidence here, assessed in its entirety and in the light most favorable to the prosecution, was not sufficient to persuade any rational trier of fact of [West’s] guilt....” Id., at 270.
*315It is clear that the Court of Appeals misapplied the commands of Teague by defining the rule from which West sought to benefit at an unduly elevated level of generality. There can, of course, be no doubt that, in reviewing West's conviction, the Supreme Court of Virginia was not entitled to disregard Jackson, which antedated the finality of West’s conviction. But from Jacksoris rule, that sufficiency depends on whether a rational trier, viewing the evidence most favorably to the prosecution, could find all elements beyond a reasonable doubt, it does not follow that the insufficiency of the evidence to support West's conviction was apparent. Virginia courts have long recognized a rule that evidence of unexplained or falsely explained possession of recently stolen goods is sufficient to sustain a finding that the possessor took the goods. See, e. g., Montgomery v. Commonwealth, 221 Va. 188, 190, 269 S. E. 2d 352, 353 (1980); Henderson v. Commonwealth, 215 Va. 811, 812-813, 213 S. E. 2d 782, 783-784 (1975); Bazemore v. Commonwealth, 210 Va. 351, 352, 170 S. E. 2d 774, 776 (1969); Bright v. Commonwealth, 4 Va. App. 248, 251, 356 S. E. 2d 443, 444 (1987). In this ease, we are concerned only with the Virginia rule’s second prong. West took the stand and gave .an explanation that the jury rejected, thereby implying a finding that the explanation was false.3 Thus, the portion of the state rule under attack here is that falsely explained recent possession suffices to identify the possessor as the thief. The rule has the virtue of much common sense. It is utterly reasonable to conclude that a possessor of recently stolen goods who lies about where he got them is the thief who took them, and it should come as no surprise that the rule had been accepted as good law against the backdrop of a general state sufficiency standard no less stringent than that of Jackson. See, e. g., Bishop v. Commonwealth, 227 Va. 164, 169, 313 S. E. 2d 390, 393 (1984); *316Inge v. Commonwealth, 217 Va. 360, 366, 228 S. E. 2d 563, 568 (1976). It is simply insupportable, then, to say that reasonable jurists could not have considered this rule compatible with the Jackson standard. There can be no doubt, therefore, that in the federal courts West sought the benefit of a “new rule,” and that his claim was barred by Teague.
On this ground, I respectfully concur in the judgment of the Court.
Because my analysis ends the case for me without reaching historical questions, I do not take a position in the disagreement between Justice Thomas and Justice O’Connor.
The Court of Appeals overlooked that West testified that he came into possession of Cardova's goods around January 1. See App. 25-27. Thus, a more accurate estimate of the time lapse would be one to three weeks.
The jury’s finding must of course be accepted under the Jackson v. Virginia, 443 U. S. 307 (1979), requirement to judge sufficiency by viewing the evidence “in the light most favorable to the prosecution.” Id., at 319.