filed a dissenting opinion in which PRICE, WOMACK and JOHNSON, JJ., joined.
I respectfully dissent. The majority denies habeas relief because applicant did not show that he was prejudiced under the second prong of Strickland.1 I would hold that applicant met the first prong of Strickland (that trial counsel was deficient) and also its second prong, i.e., “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” 466 U.S. at 694, 104 S.Ct. 2052.
I. Deficient Performance
Because the majority does not hold whether' counsel was deficient, i.e., whether “counsel’s representation fell below an objective standard of reasonableness,” id. at 688-89,104 S.Ct. 2052,1 find it unnecessary to discuss it at great length, other than to say that we have held, in a variety of contexts, that failing to follow clear procedural rules will usually be considered objectively unreasonable representation. See e.g., Ex parte Coy, 909 S.W.2d 927, 928 (Tex.Crim.App.1995); Ex parte Dietzman, 790 S.W.2d 305 (Tex.Crim.App.1990); cf. Hernandez v. State, 726 S.W.2d 53, 58 (Tex.Crim.App.1986) (rejecting applicant’s claim that his lawyer was unfamiliar with the rules of evidence).
I would add, however, that in my view, the motion for community supervision was adequately verified. The “defect” in the motion, according to the majority is that the signature line where applicant’s name should appear is blank on the affidavit attached to the motion; however, applicant’s signature appears on the face of the motion. But more importantly, the affidavit is subscribed' by the trial court clerk, and he or she avers that applicant appeared and swore to the facts contained in the motion.2 This is and should be sufficient; otherwise, we are elevating form *820over content.3
II. Prejudice
When evaluating whether an applicant was prejudiced by the unprofessional errors of his attorney, a habeas court may not apply an outcome determinative test. Strickland, 466 U.S. at 693-94, 104 S.Ct. 2052. That is, we must not place a burden on the applicant to show that he was more likely to have prevailed than not if his lawyer had not committed the unprofessional errors. See id. This standard is inappropriate because the defendant suffers from “the absence of one of the crucial assurances that the result of the proceeding is reliable.” Id.
After a careful review of the appellate and habeas records, it appears to me that the majority placed a burden of proof upon the applicant to show that the evidence preponderates in his favor; namely, that applicant did not demonstrate that he would have received probation or a lesser sentence had the jury had the option of granting community supervision. Specifically, the majority concludes that “the central issue of prejudice under Strickland, is whether there is a reasonable probability that applicant’s sentencing jury would have recommended probation had the issue been submitted to it.”4 (citing Woodford v. Visciotti 537 U.S. 19, 123 S.Ct. 357, 154 L.Ed.2d 279 (2002)). Setting aside for the moment that Visciotti is plainly distinguishable, it is also plain that, according to the majority’s holding today, applicant would have to demonstrate that there was a reasonable probability that the jury would have recommended probation or a lesser sentence in order to be entitled to relief. This is, in the plain words of Strickland, inappropriate — the holding here is nothing more than an application of the outcome determinative standard prohibited by Strickland.
Now, back to Woodford v. Visciotti. Without more than a parenthetical summary of Visciotti, the majority denies relief because applicant must show (but did not show) that the jury would have recommended a more favorable sentence. However, Visciotti does not set forth, as a substantive rule, that second-prong Strickland error during sentencing must be such that the jury would have given the defendant a lesser sentence. Rather, the Supreme Court in Visciotti simply concluded, as it has many times, that federal habeas review of state-court habeas proceedings must focus on whether the state court’s judgment was an unreasonable — not just incorrect — application of federal law. Deciding that the Ninth Circuit had substituted its judgment for the state habeas court, the Supreme Court reversed, noting it may have agreed with either the state or the federal court’s interpretation of whether Visciotti was prejudiced by his lawyer’s failure to introduce mitigating evidence, but further held that the federal court’s *821duty is not to consider whether the state-court decision is incorrect, but rather, if it involved an unreasonable application of Strickland. As such, Visciotti is not precedent as substantive law to be applied by the state habeas court; instead, it is a decision which limits, pursuant to statute, the federal court’s right to reverse a state court’s ruling on habeas corpus relief.
After explaining that relief in this case would be based on “pure conjecture and speculation,”5 the majority cites this Court’s “highly questionable” pre-Strickland holding in Mercado v. State, 615 S.W.2d at 228, to deny relief. See GeoRGE E. Dix & RobeRT 0. Dawson, 43A Texas Practice and PROCEDURE § 89.22 (2001) [hereinafter Dix & Dawson]. Mercado was charged with possession of heroin, a jury found him guilty, and assessed punishment at 17 years imprisonment. 615 S.W.2d at 226. In a motion for new trial, Mercado complained that he received ineffective assistance of counsel because his lawyer never told him that he was eligible for an instruction on probation and his lawyer did not file a motion for probation. At the hearing on the motion for new trial, trial counsel testified that he indeed failed to file the motion or inform Mercado that he was entitled to the instruction on probation. Importantly, however, trial counsel testified (unlike trial counsel here; see Strickland, 466 U.S. at 690-91, 104 S.Ct. 2052) that his decision not to submit the issue of probation to the jury was one of trial strategy; i.e., that such a request could be viewed by the jury as a partial admission of guilt. 615 S.W.2d at 227.
As a pre-Strickland case, the Mercado panel devoted most of the opinion to hashing out the proper standard of review, and concluded in two sentences (the latter of which is an alternative holding) that Mercado was not entitled to relief because (1) there was no evidence admitted at the motion for new trial or in the record to show that Mercado was eligible for probation, and (2) “[fjurther, under the penalty assessed by the jury under the facts of the case, the jury would not have reached the consideration of any motion for probation, even if one had been filed.” Id. at 228.
Thus, the “rule” in Mercado is most certainly dicta, as it was clearly an alternative holding to the conclusion that Mercado had not shown he was entitled to probation in the first place. But the dicta in Mercado has, sadly, morphed into some sort of “rule” which has become to be known as such — when a jury assesses punishment at more than ten years, the improper failure to submit probation as a punishment option cannot be reversible error because the jury would not have reached the consideration of a motion for probation. See Resendez v. State, 2005 WL 1025375 *1, 2005 TexApp. Lexis 3340 *3-1 (Tex.App.-San Antonio 2005, no pet.) (not designated for publication); Dix & Dawson, 43 Texas PRACTICE and PROCEDURE § 36.58. Indeed, this “rule” is no less speculative or based on conjecture than a conclusion that the lack of a probation instruction did not prejudice applicant. Vasquez v. State, 830 S.W.2d 948, 951 (Tex.Crim.App.1992) (per curiam).
Simply put, there is not a sufficiently strong logical connection between the majority’s conclusion (that the jury would not have given a more lenient sentence had it had the probation option before it) and the *822premise (the jury assessed a 40-year sentence).6 See Dix & Dawson, 43A Texas PRACTICE AND PROCEDURE § 39.22 (2001) (if defendant was entitled to have issue of probation submitted to the jury, but it was not, that would surely affect the jury’s punishment deliberations because it would not be considering the full range of punishment applicable to the case).
Alas, my belief that Mercado’s dicta is unsound because it is based on pure conjecture and speculation, has not swayed a majority of the judges of this Court to overrule this mischievous precedent, but my voice is not without strong allies. Namely, Professors Dix and Dawson have decried the “rule” in Mercado, and more eloquently than I, explain its fallacy as follows:
[The analysis in Mercado ] seems incomplete because failure to charge on probation may well affect the length of the sentence the jury ultimately gives because failure to charge on probation gives a false picture of the seriousness of the offense. Further, the jury is instructed that it must assess a punishment of 10 years or less for the defendant to be eligible for probation. [Tex.Code CRIm. Proc. Ann. art. 42.12 § 4(d)(1) ]. It would not be unusual for a jury to assess 10 years and give probation, when that is an option, but to assess more than 10 years in the absence of the option of probation. , In other words, the fact that without a probation option the jury assessed more than 10 years does not mean that it would have rejected probation had that option been made available to it.
Dix & Dawson, 43 Texas Practice and Procedure § 36.58 (citing Snow v. State, 697 S.W.2d 663, 665 (Tex.App.-Houston [1st Dist.] 1985)).
In Snow, the court of appeals explained the “gravitational influence” that the probation option may have had on a jury’s ultimate verdict:
[TJhere is a reasonable probability that a jury instruction concerning probation would have altered the outcome of the punishment by influencing the jury to seriously consider lower sentence that 10 years imprisonment. The right to be considered for probation is valuable, even if probation is not given, because the jury instruction concerning probation forcefully directs the jury’s attention to the lowest punishment allowed by law.
I would “overrule” Mercado and follow Snow. See 697 S.W.2d at 665.
The majority cannot say with any assurance that the result of the proceeding was reliable when admitting that the jury did not have before it the proper range of punishment. See Strickland, 466 U.S. at 668, 104 S.Ct. 2052. Nor does the majority dispute that the proper range of punishment in this case would have included the *823option of community supervision. See Maj. Op. at 817 & n. 4. Applicant was fifteen years old at the time of the present offense, and the record reflects that he had one prior offense for unauthorized use of a vehicle, committed when he was fourteen years old. This prior offense would not prevent the trial court from granting a jury charge on community supervision. See Tex. Fam.Code Ann. § 51.13(d). Therefore, there was no legal bar to submitting the question of probation to the jury, and the trial court should have done so.7
We have repeatedly held that any punishment selected by the jury must fit within the statutory range of punishment. Cartwright v. State, 833 S.W.2d 134, 135 (Tex.Crim.App.1992); Stein v. State, 515 S.W.2d 104, 108 (Tex.Crim.App.1974); see also Gonzales v. State, 672 S.W.2d 618, 620 (Tex.App.-Amarillo 1984); Dix & Dawson, 43 Texas PRACTICE and PROCeduee § 36.53. As Professors Dix and Dawson opined, “it seems abundantly clear that the mere fact that the jury selected a punishment within the correct range does not mean automatically that the defendant was not harmed by the mistaken instruction.” Drx & Dawson, 43 Texas Practioe and Prooedure § 36.53. A jury is likely to attempt to assess the culpability of the defendant’s offense within a continuum of severity as reflected in the range of punishment it was erroneously given by the trial court. Id. Therefore, if the range of punishment is overstated, that would tend to cause the jury to select a more severe sentence than it might have selected had it been correctly instructed. Id. (citing Uribe v. State, 688 S.W.2d 534, 538 (Tex.Crim.App.1985), overruled on other grounds by Cartwright, 833 S.W.2d at 135).
We discussed an analogous problem in Uribe v. State, saying that when the jury is given a punishment greater than what the law allows, “[t]he court’s misdirection thus harms the defendant because the jury is instructed to consider his offense as more serious than the law grades it.” 688 S.W.2d at 538. Similarly, since applicant’s jury did not receive a charge with the proper range of punishment, and instead received one that erroneously failed to show that the law would have allowed for probation, applicant was surely harmed because it likely caused the jury to consider his crime far more serious than the law graded it. See id. In sum, the fact that applicant suffered from the absence of the minimum punishment allowed by law, applicant sufficiently showed that the outcome of the case was undermined. See Strickland, 466 U.S. at 696, 104 S.Ct. 2052.8
III. Conclusion
We must always be mindful that when reviewing claims of ineffective assistance of counsel, our ultimate duty is to insure the “fundamental fairness of the proceed*824ing.” See id. The process here has not been fair as contemplated by Strickland. See id. Moreover, the record reflects an unreliable result due to a breakdown in the adversarial process, which had more than a mere conceivable affect on the outcome of the proceeding. See id. at 693, 104 S.Ct. 2052. Because I do not believe that the proceedings produced a “just result” in this case, I respectfully dissent. See id. at 696, 104 S.Ct. 2052; Vasquez v. State, 830 S.W.2d at 951.
. Strickland v. Washington, 466 U.S. 668, 687-88, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).
. The State acted inconsistently by complaining on direct appeal that the motion was not properly verified, and then shifted positions on habeas review (where such an argument would be grounds for relief) and argued that the motion was indeed properly verified. Obviously then, the State took the mutually exclusive positions when it would be favorable to itself. This created a fundamentally unfair situation for both applicant and his lawyer. See e.g., New Hampshire v. Maine, 532 U.S. 742, 749, 121 S.Ct. 1808, 149 L.Ed.2d 968 (2001) ("Where a party assumes a certain position in a legal proceeding, and succeeds in maintaining that position, he may not thereafter, simply because his interests have changed, assume a contrary position....”); Tex.Code Crim. Proc. Ann. art. 2.01 (It shall be the primary duty of all prosecuting attorneys not to convict but to see that justice is done).
. But even though I believe that trial counsel should not be found deficient for failing to insure that the motion for community supervision was verified properly, applicant’s argument that trial counsel was deficient for failing to put on any evidence to show that applicant was indeed entitled to probation is a more compelling argument. See Tex.Code Crim. Proc. Ann. art 42.12 § 4(e); see e.g., Mercado v. State, 615 S.W.2d 225, 228 (Tex.Crim.App. [Panel Op.] 1981).
. While it is hard to imagine how applicant could ever meet such a burden as set forth by the majority today, see Tex.R. Evid. 606(b), Strickland clearly states that the applicant need not do so to be entitled to habeas relief. Strickland, 466 U.S. at 694, 104 S.Ct. 2052 ("The result of a proceeding can be rendered unreliable, and hence the proceeding itself unfair, even if the errors of counsel cannot be shown by a preponderance of the evidence to have determined the outcome.”).
. The majority concludes that the basis for granting relief in this case would amount to “conjecture and speculation.” Webster defines "conjecture” as "a conclusion deduced by surmise or guesswork." Webster's Ninth New Collegiate Dictionary 277 (1983). "Speculate” means "to take to be true on the basis of insufficient evidence.” Id. at 1133. Would it not though be equally true that the majority’s holding today, and the reasoning in Mercado, are also based on pure conjecture and speculation?
. The "Modus Ponens” reasoning of Mercado goes something like this:
If "P” then "Q”.
"P"; Therefore "Q”
Assign "the jury gave a forty-year sentence in this case without the probation option” to P; Assign "the jury would not have considered a lighter sentence even with probation option” to Q. The equation we end up with is this:
If the applicant received a forty-year sentence, then the jury would not have considered probation or a lighter sentence. Applicant received a forty-year sentence; therefore, the jury would not have considered probation or a lighter sentence.
This reasoning merely affirms the consequence, which is a non-sequitur (“it does not follow”) and faulty logic. See A.R. Lacey. A Dictionary of Philosophy (3d ed.1996). By plugging other facts into the model, the faulty logic is easier to see:
If the streets are wet, then it’s raining. The streets are wet; therefore, it's raining.
I would solidly renounce Mercado.
. "[A] a court should presume ... that the judge or jury acted according to law.” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. I find footnote 6 of the majority opinion worthy of comment. Strickland instructs that we must presume the judge and the jury acted in accordance with the law. As the majority has nearly conceded, applicant was entitled to the probation instruction. Accordingly, whether the trial court would have committed error anyway is not a proper consideration and should not be a factor in this Court's reasoning upon applicant’s ineffective assistance of counsel claim. This improper consideration further indicates a fundamentally unfair result in this case. See id. at 696, 104 S.Ct. 2052.
. Errors that undermine confidence in the fundamental fairness of the proceeding justify habeas relief. See, e.g., Teague v. Lane, 489 U.S. 288, 311-314, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989).