Alexander Durrive is serving a sentence of 120 months’ imprisonment for conspiring to distribute cocaine and using the telephone to facilitate that crime. We affirmed the conviction and sentence on direct appeal. 902 F.2d 1221 (7th Cir.1990). Durrive then began this proceeding under 28 U.S.C. § 2255, arguing that the district court had not complied with Fed.R.Crim.P. 32 in imposing sentence. Anticipating that he might have trouble presenting the Rule 32 argument on collateral review, Durrive launched an attack on counsel, contending that his lawyer furnished ineffective assistance in not ensuring compliance with that rule.

Durrive was correct to perceive that noncompliance with Rule 32 may not be raised by collateral attack. Hill v. United States, 368 U.S. 424, 82 S.Ct. 468, 7 L.Ed.2d 417 (1962); Johnson v. United States, 805 F.2d 1284 (7th Cir.1986). See also United States v. Timmreck, 441 U.S. 780, 99 S.Ct. 2085, 60 L.Ed.2d 634 (1979); Scott v. United States, 997 F.2d 340, 341-42 (7th Cir.1993). A prisoner may of course contend that the procedures used in his case violated the Constitution, but the contentions Durrive advances — that he did not read the presentence report personally; that the judge did not ask him whether he had read the report; that the judge did not state why he chose 120 months’ imprisonment from the range of 108-35 months determined under the Guidelines; that the judge misinterpreted the import of some evidence in the record — do not demonstrate constitutional shortcomings. The Constitution permits courts substantial leeway in the procedures used to arrive at a[*550] sentence. E.g., United States v. Dunnigan, — U.S. -, ---, 113 S.Ct. 1111, 1116-18, 122 L.Ed.2d 445 (1993); United States v. Grayson, 438 U.S. 41, 98 S.Ct. 2610, 57 L.Ed.2d 582 (1978); Williams v. Oklahoma, 358 U.S. 576, 79 S.Ct. 421, 3 L.Ed.2d 516 (1959); Williams v. New York, 337 U.S. 241, 69 S.Ct. 1079, 93 L.Ed. 1337 (1949).

Durrive has never denied that he discussed the presentence report with his counsel and that he had an opportunity to read that document. Durrive had an opportunity to contest the conclusions of the presentence report, and his lawyer did so at some length. After his lawyer had finished, the judge asked Durrive if he had anything to add; Durrive spoke briefly but did not make substantive objections. The Constitution does not require a district judge to agree with the defendant’s assessment of the evidence; it requires only a procedure that is likely to lead to accurate decisions. Hill, 368 U.S. at 428, 82 S.Ct. at 471; Johnson, 805 F.2d at 1288. See also United States ex rel. Villa v. Fairman, 810 F.2d 715, 718-19 (7th Cir.1987). The district court’s procedure satisfied that standard.

Thus everything comes down to the contention that counsel furnished ineffective assistance. The district judge rejected this contention, concluding that, even with the benefit of hindsight, he would have imposed the same sentence. Thus, the judge concluded, Durrive cannot satisfy the “prejudice” component of ineffective assistance under Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). The prosecutor adds on appeal that counsel’s performance was not constitutionally deficient. Although counsel did not object to the judge’s casual treatment of the defendant’s access to the presentence report, and did not request the district judge to state reasons for imposing sentence at a particular point in the guideline range, counsel was otherwise diligent. He objected strenuously to portions of the presentenee report, argued for a lower sentence, and took an appeal presenting substantial questions for this court’s consideration.

Strickland defines “prejudice” as a probable effect on the outcome. 466 U.S. at 694, 104 S.Ct. at 2068. An “outcome” of guilt rather than innocence, or of death rather than a term of imprisonment, presents the effect starkly. But what of other sentences, where small differences in the quantity of drugs affect the range computed under the Sentencing Guidelines? A superior presentation by the lawyer may induce the judge to subtract a few offense levels or give a sentence lower in the range. Because the judge both determines a range and selects from within that range, almost any of counsel’s actions has a potential effect on the sentence. Small failings by counsel are not enough to turn a probable acquittal into a probable conviction; equally slight failings could turn a 115 month sentence into a 120 month sentence. Do effects of this kind satisfy the prejudice component of Strickland? If so, the sixth amendment becomes the means of vindicating on collateral attack all manner of arguments under rules and statutes. Almost any error at sentencing may be recast as a challenge to counsel. Had the lawyer ensured compliance with Rule 32, the argument goes, the district court would have selected a lower sentencing range, and this difference, defendant would have us believe, yields a probable effect on the outcome, and therefore “prejudice,” under Strickland. A Rule 32 claim thus is open for argument just as if this were a direct appeal. Yet this contradicts the holdings of Hill, Johnson, and many other cases that violations of the rules of criminal procedure are not grounds for collateral relief. Does Strickland admit through the back door considerations barred at the front?

The answer must be yes if probable effect on the outcome means but-for causation (or probable but-for causation). Cases since Strickland show, however, that causation is not enough, because not all effects are of equal weight. “The essence of an ineffective-assistance claim is that counsel’s unprofessional errors so upset the adversarial balance between defense and prosecution that the trial was rendered unfair and the verdict rendered suspect.” Kimmelman v. Morrison, 477 U.S. 365, 374, 106 S.Ct. 2574, 2582, 91 L.Ed.2d 305 (1986). That understanding led the Court in Lockhart v. Fretwell, —[*551] U.S. -, 113 S.Ct. 838, 122 L.Ed.2d 180 (1993), to reject the equation between causation and prejudice. “[A]n analysis focussing solely on mere outcome determination, without attention to whether the result of the proceeding was fundamentally unfair or unreliable, is defective. To set aside a conviction or sentence solely because the outcome would have been different but for counsel’s error may grant the defendant a windfall to which the law does not entitle him.” Id. at -, 113 S.Ct. at 842-43. If an effect on the outcome is not enough, what is? According to Lockhart, the defendant must establish that counsel’s shortcomings “render[ed] the result ... unreliable or the proceeding fundamentally unfair.” Id. at -, 113 S.Ct. at 844. This takes us back to the question whether a violation of Rule 32 makes a sentence “unreliable or ... fundamentally unfair.” Hill and Johnson answer that question “no.” Similarly, Scott holds that a misapplication of the Sentencing Guidelines does not make the sentence “unreliable or ... fundamentally unfair.” Lockhart thus avoids the two-step process by which an invocation of the sixth amendment opens to collateral review errors that do not themselves call for a writ of habeas corpus.

Grave errors by judge and counsel might make a sentence under the Guidelines “unreliable or ... fundamentally unfair.” For instance, an error that produces a large effect on the sentence could be condemned on this basis. “In order to avoid turning Strickland into an automatic rule of reversal in the non-capital sentencing context, ... in deciding such an ineffectiveness claim, a court must determine whether there is a reasonable probability that but for trial counsel’s errors the defendant’s non-capital sentence would have been significantly less harsh.” Spriggs v. Collins, 993 F.2d 85, 88-89 (5th Cir.1993) (emphasis in original). The court added: “‘Significance’ is a relative term here. Of course, it is arguable that any amount of liberty of which a person is unnecessarily deprived is ‘significant.’ However, our reading of the Supreme Court’s capital ineffectiveness jurisprudence leads us to believe that ‘prejudice’ must be rather appreciable before a new [proceeding] is warranted in view of counsel’s error. If an appreciable showing of prejudice is required in the capital context, a requirement for a showing of significant prejudice applies a fortiori in the non-capital context.” Id. at 89 n. 5 (emphasis in original; citations omitted). We need not decide whether there are other ways to establish that the outcome of a sentencing proceeding is “unreliable or ... fundamentally unfair”; Durrive’s claim fails by almost any standard.

The district court determined that Dur-rive’s conspiracy planned to acquire three kilograms of cocaine. Durrive says that meticulous adherence to Rule 32’s procedures would have led the court to understand that the quantity under consideration was less than two kilograms, reducing the offense level from 28 to 26. This difference would have dropped the sentencing range to 87-108 months, leading to at least one year’s reduction in the sentence. As the fifth circuit recognized, an extra year in custody is significant from the defendant’s perspective. But the difference between 120 months and 108 or even 98 (the middle of the lower range) does not demonstrate that the actual sentence is “unreliable ... or fundamentally unfair” — is not, in the language of Spriggs, a “significant” difference. That opinion held that a lawyer’s failure to contest the presen-tence report’s description of the defendant’s prior offenses did not lead to a “significant” increase in the sentence even though the judge cited those offenses when imposing a term of 35 years’ imprisonment. Durrive protests a lesser injury (if there was any error at all). Adjusting the offense level by two or three steps is exactly the routine decision that is supposed to be handled at sentencing and on direct appeal. Before the advent of the Sentencing Guidelines, no one would have dreamed that choices influencing the term of imprisonment within such a narrow range could be relitigated on collateral attack. Nothing in the Guidelines expands the scope of such review, see Scott, and we therefore conclude that Durrive has not established “prejudice” within the meaning of Strickland.

Our conclusion that Durrive may not obtain an indirect review of his Rule 32 claims through the vehicle of disputing his counsel’s performance does not imply that the district[*552] judge acted properly. The judge failed to comply with 18 U.S.C. § 3553(c)(1), which requires the court to give reasons for selecting the point within a sentencing range that exceeds 24 months. The judge also failed to comply with Fed.R.Crim.P. 32. Rule 32(e)(3)(A) provides that “[a]t least 10 days before imposing sentence, unless this minimum period is waived by the defendant, the court shall provide the defendant and the defendant’s counsel with a copy of the report of the presentence investigation”. Rule 32(a)(1)(A) adds that before imposing sentence the court shall “determine that the defendant and defendant’s counsel have had the opportunity to read and discuss the pre-sentence investigation report made available pursuant to subdivision (c)(3)(A)”. Durrive did not receive the presentence report 10 days before sentencing, and the district court neither elicited an express waiver nor determined that Durrive had the requisite opportunity to read the report. Instead the judge relied on counsel to ensure that Durrive had such access as counsel thought appropriate. The day before sentencing Durrive’s parents hired him a new lawyer, who advised the judge that neither he nor his client had read the report. Although the judge volunteered to defer the sentencing (counsel declined the offer), the judge did not make the inquiry that Rule 32(a) requires. Counsel represented Durrive vigorously on appeal. His current lawyer’s attack on that lawyer’s appellate performance does not require discussion. It may be that counsel disdained these potential issues because of a perception that a remand would do no good. (So the judge said when denying the petition under § 2255.) Effective lawyers confine themselves to the issues with the greatest potential for assisting their clients. Jones v. Barnes, 463 U.S. 745, 751-53, 103 S.Ct. 3308, 3312-13, 77 L.Ed.2d 987 (1983). Still, complying with Rule 32 at sentencing would have been far preferable — first because Rule 32 is the law, and second because it would have saved everyone the time and aggravation this collateral proceeding has entailed.

Affirmed.