This is an appeal from the denial of Defendant’s (Appellant’s) petition for post-conviction relief. The sole issue is the retrospective application of Lawrence v. State (1972), 259 Ind. 306, 286 N.E.2d 830, wherein we adopted a prophylactic rule requiring a bifurcated trial in habitual criminal charges.
In Enlow v. State (1973), 261 Ind. 348, 303 N.E.2d 658 we held that the Lawrence rule was applicable to cases decided on direct appeal after September 11, 1972, the date that Lawrence was handed down. Defendant now urges its application to his conviction in 1962 which was affirmed by this Court on direct appeal in 1966. MePhearson v. State (1966), 247 Ind. 579, 219 N.E. 2d 907.
We recently handed down Prophet v. State (1974), 262 Ind. 312, 315 N.E.2d 699 holding that Prophet was not entitled to have his convictions in a Pre-Lawrence unified trial set aside. Our opinion, however, left a suggestion that we might give consideration to such challenges upon a case-to-case basis—a door that we now believe had best be closed.
The considerations pertinent to the determination of whether or not to give retrospective application to a duly announced rule affecting criminal trials were related in Enlow v. State, supra.
“ ‘The criteria guiding resolution of the question implicate (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.’ Stovall v. Denno (1966), 388 U.S. 293. 297. 87 S. Ct. 1967. 18 L. Ed. 2d 1199.” 303 N.E.2d at 659
*470■ We further acknowledged in Enlow, mpra, that the first of those factors was the most important.
“It is clear that the most important and compelling of these three factors is the purpose to be served by the newly announced rule. Desist v. U.S. (1968), 394 U.S. 244, 89 S. Ct. 1030, 22 L. Ed. 2d. 248; Witherspoon v. Illinois (1967), 391 U.S. 510, 88 S. Ct. 1770, 20 L. Ed. 2d 776. If that purpose is to correct serious flaws in the fact finding process at trial then the rule is applied retrospectively. Roberts v. Russell (1967), 392 U. S. 293, 88 S. Ct. 1921, 20 L. Ed. 2d 1100.
“ ‘We have given complete retroactive effect to the new rule, regardless of goodfaith reliance by law enforcement authorities or the degree of impact on the administration of justice where the major purpose of new constitutional doctrine is to overcome an aspect of the criminal trial that substantially impairs its truthfinding function and so raises serious questions about the accuracy of guilty verdicts in past trials.’ Adams v. Illinois (1972), 405 U.S. 278, 280, 92 S. Ct. 916, 31 L. Ed. 2d 202.
“When the new principle being applied goes to the fairness of the trial itself and is designed to eliminate a previously existing danger of convicting the innocent, rather than merely constructing a more effective remedy by which to enforce other constitutional rights not necessarily connected with the fact finding function, then the principle is deemed to be retrospective. Linkletter v. Walker (1965), 381 U.S. 618, 85 S. Ct. 1731, 14 L. Ed. 2d 601; Jackson v. Denno (1963), 378 U.S. 368, 84 S. Ct. 1774, 12 L. Ed. 2d 908; Gideon v. Wainwright (1962), 372 U.S. 335, 83 S. Ct. 792, 9 L. Ed. 2d 799; Griffin v. Illinois (1956), 351 U.S. 12, 76 S. Ct. 585, 100 L. Ed. 891; Monserrate v. State (1971), 256 Ind. 623, 271 N.E.2d 420.
“With these principles in mind an examination of our holding in Lawrence indicates that it militates a retrospective application.” 303 N.E. at 659, 660.
To say that the purpose to be served by the new standard is the most important of the three considerations, however, is not to say that it totally eclipses the others, and to say that the relative importance of the vital considerations militate a retrospective application of our holding in Lawrence is not to say that such application is mandated. The con*471siderations are relative, and in the final analysis we must determine each issue of retroactivity in context.
Acknowledging that the purpose of the Lawrence rule is to correct serious flaws in the “fact finding process” in the broad sense, we nevertheless, should not be governed by this factor alone. Regardless of how we hold, some good and some evil will flow from our decision today. We are constrained, then, to elect that course of action which will, in our opinion, result in more good than harm.
As we stated in Enlow, supra, the Lawrence type situation goes to the “fact finding process.” Almost everything admitted into evidence weighs upon the minds of the fact-finder. In the broad sense, then, all evidence affects the “fact finding process.” As rational men, however, we know that some evidence weighs more heavily than other and if erroneously admitted, the seriousness of the error is relative. However, our analysis of the United States Supreme Court cases leads us to believe that their use of the term “fact finding process” is in a much narrower sense, limited to include only those matters that appear to bear directly and persuasively upon the determination of the guilt or innocence of the defendant, and that that Court would apply a rule retroactively if the flaw corrected bore directly and persuasively upon the determination of guilt. However, if the flaw to be removed has only an indirect effect upon such determination or, even though direct, it does not appear to be of substantial impact, retroactivity has not been mandated.
The flaw we desired to eliminate in Lawrence was the “splash-over” from the presentation of evidence of prior convictions into the determination of guilt or innocence upon the primary criminal charge. The harm from such flaw arises when the probative evidence upon , the primary charge leaves a reasonable doubt as to the defendant’s guilt. In those cases where no reasonable doubt remains, the premature entry of evidence of the prior convictions is harmless.
*472Notwithstanding the seriousness of the flaw corrected by Lawrence, supra, it is our opinion that the considerations weighing in favor of its total retrospective application do not preponderate in the entire picture. Additionally, although the “splash-over” effect of the prior convictions evidence is likely to be prejudicial, such prejudice does not bear directly upon the determination of guilt or innocence upon the primary charge. Rather, the detriment to the defendant is one of impeachment. Logically, the propensity of the flaw for imperiling the accused is not so great as that eliminated by Bruton v. United States (1968), 391 U.S. 123, 88 S. Ct. 1620, 20 L. Ed. 2d 476, wherein the extra-judicial confession of a co-defendant implicating the defendant was held to be inadmissible. Bruton was given retroactive application in Roberts v. Russell (1968), 392 U.S. 293, 88 S. Ct. 1921, 20 L. Ed. 2d 1100 in light of the same considerations reviewed in the Enlow case, supra. The situation in the case at hand is more akin to that in Ashton v. Anderson (1972), 258 Ind. 51, 279 N.E.2d 210, with respect to which we have said there is no retroactivity. Dexter v. State (1973), 260 Ind. 608, 297 N.E.2d 817, 818.
The judgment of the trial court is affirmed.
Arterburn, C.J., Givan and Hunter, JJ., concur; DeBruler, J., dissents with opinion.