In Re Anthony Roberson v. the State of Texas

TEXAS COURT OF APPEALS, THIRD DISTRICT, AT AUSTIN

NO. 03-24-00325-CR

In re Anthony Roberson

FROM THE 299TH DISTRICT COURT OF TRAVIS COUNTY NO. D-1-DC-95-953274, THE HONORABLE KAREN SAGE, JUDGE PRESIDING

OPINION

This is an appeal from an order finding no reasonable probability of non-conviction following the results of post-conviction forensic DNA testing. Because we agree with the convicting court that the reported results do not support a finding otherwise, we overrule Roberson’s issues challenging that finding. We also overrule Roberson’s issues claiming due process violations in the Chapter 64 proceedings. Finally, we dismiss his issues challenging his underlying conviction—all of which are beyond the scope of a Chapter 64 appeal.

BACKGROUND

1. The Assault and The Trial In Case No. D1-DC-95-953274

“In the early morning hours of November 4, 1990, R.P. was sexually assaulted in her Northwest Hills home in Austin. She was unable to identify her assailant.” Roberson v. State, 16 S.W.3d 156, 159 (Tex. App.—Austin 2000, pet. ref’d). At the 1998 trial, DPS DNA analyst Michelle Lockhoof testified that she performed restriction fragment length polymorphism (RFLP) testing on semen stains from R.P.’s nightgown, panty liner, and vaginal swab, and Roberson’s *2 known blood. She testified Roberson could not be excluded as a donor of the semen stains. Id. at 161-62. Lockhoof stated that Roberson’s profile occurred in approximately 1 in 420 billion persons, but because the DPS crime lab does not report ratios higher than the world’s known population she stated that Roberson’s profile occurred in 1 in 5.5 billion persons. Id. at 162.

During the sentencing hearing of that case, the trial court heard about four other home-invasion aggravated sexual assaults in the early 1990’s, all in the Northwest Hills neighborhood of Austin, and all tied to Roberson by DNA. Id. at 163. First, M.S. testified Roberson had been convicted (in Case No. D-1-DC-94-942556) of aggravated sexual assault in 1995 based on an August 16, 1991 offense against her in her home. See Roberson v. State, No. 03-96-00067-CR, 1997 WL 420979 (Tex. App.—Austin July 24, 1997, no pet.) (not designated for publication). Also, “S.T., G.B., and H.B. each testified to an extraneous unadjudicated aggravated sexual assault that occurred in their homes in the early morning hours of June 30, July 13, and July 15, 1991, respectively.” Roberson, 16 S.W.3d at 163. None could identify Roberson as the perpetrator, but Roberson “could not be excluded as the donor of the sperm in each case.” Id. “Lockhoof again testified that there was a 1 in 5.5 billion chance in each offense that [Roberson’s] DNA profile would match another individual.” This court affirmed Roberson’s conviction for the aggravated sexual assault over his sufficiency complaints. Id. at 159, 171-72. 2. The Motion for DNA Re-Testing

Roberson moved for re-testing of samples associated with his 1998 conviction— including the vaginal swab from R.P.’s sexual assault kit. Roberson noted that comparisons between the DNA from the vaginal swab and Roberson were made in 1995, but the evidence could now be “subjected to testing with newer testing techniques that provide a reasonable likelihood of *3 results that are more accurate and probative than the results of the previous test.” See Tex. Code Crim. Proc. art. 64.01(b)(2)(A). The State did not oppose the motion on the conditions that (1) DNA analysis be conducted in all five of the aggravated sexual assault cases that were attributed to Roberson via DNA, and (2) all analysis be conducted in the same lab. The convicting court ordered re-testing of the samples associated with the convictions; the State sent the samples associated with the unadjudicated offenses. Bode Technology conducted all the testing. 3. The Reported Results

Regarding the sample associated with the challenged conviction, Bode Technology reported as follows:

• • CCX2320-0521-E-01 (victim R.P./vaginal smear swab): “The Y-STR profile obtained from sample CCX2320-0521-E-01 is consistent with a mixture of at least two individuals including a major contributor (Male 1). Anthony Roberson cannot be excluded as a possible contributor of the major component Y-STR profile deduced from sample CCX2320-0521-E-O1. The major component Y-STR profile deduced is at least 1,923 times more likely to be observed if it originated from Anthony Roberson than if from an unknown, unrelated male individual from the US population. Due to the possibility of allelic dropout, no conclusions can be made on the minor alleles.”

Regarding the four samples associated with the prior conviction and three unadjudicated offenses, Bode Technology reported as follows:

• • CCX2320-0520-E-01 (victim M.S./vaginal smear swab): “The DNA profile obtained from the sperm fraction (SF) of sample CCX2320-0520-E-01 is consistent with a male contributor. This DNA profile matches the DNA profile obtained from Anthony Roberson. 8to the interpretable portion of the mixture DNA profile. Assuming a mixture of two individuals, this mixture DNA profile obtained is at least 2.4 octillion times more likely to be observed if it originated from Anthony Roberson and one unknown, unrelated individual than if from two unknown, unrelated individuals.”
• • CCX2320-0524-E-01 (victim’s name redacted/semen positive stain): “The DNA profile obtained from the sperm fraction (SF) of sample CCX2320-0524-E-01 is consistent with a

male contributor. This DNA profile matches the DNA profile obtained from Anthony Roberson. The DNA profile obtained is at least 4.4 octillion times more likely to be observed if it originated from Anthony Roberson than if from an unknown, unrelated individual.”

• • CCX2320-0522-E-01 (victim’s name redacted/semen positive stain): “The DNA profile obtained from the sperm fraction (SF) of sample CCX2320-0522-E-01 is consistent with a male contributor. This DNA profile matches the DNA profile obtained from Anthony Roberson. The DNA profile obtained is at least 4.5 octillion times more likely to be observed if it originated from Anthony Roberson than if from an unknown, unrelated individual.”

The convicting court concluded that the results of the DNA testing were “not exculpatory,” and that, if they had been available during the 1998 trial, “it is reasonably probable that [Roberson] would have been convicted.” 4. The Issues Raised in This Appeal

Roberson now appeals the convicting court’s ruling in twelve issues, eight in his initial brief, and two each in two supplemental briefs. We address the issues in three groups: those challenging the convicting court’s finding that the results were not favorable; those claiming due process violations in the Chapter 64 proceedings, and those challenging his underlying conviction.

ANALYSIS

1. Issues Challenging the Convicting Court’s Finding that Results Were Not Favorable

Roberson’s first five issues challenge the convicting court’s conclusion that the results were not favorable. Chapter 64 provides that after examining the results of testing under Article 64.03, “the convicting court shall hold a hearing and make a finding as to whether, had the results been available during the trial of the offense, it is reasonably probable that the person would not have been convicted.” Tex. Code Crim. Proc. art. 64.04. This is not an actual innocence standard; the question is “whether the results ‘would cast affirmative doubt upon the validity of *5 the inmate’s conviction.’” Flores v. State, 491 S.W.3d 6, 9 (Tex. App.—Houston [14th Dist.] 2016, pet. ref’d) (quoting Raby v. State, No. AP–76,970, 2015 WL 1874540, at *5–6 (Tex. Crim. App. April 22, 2015) (not designated for publication)).

“When reviewing an Article 64.04 favorability finding, we apply a bifurcated standard of review, affording almost total deference to a trial court’s resolution of historical facts and mixed questions that turn on credibility and demeanor, but we review de novo mixed questions that do not turn on credibility and demeanor and questions of law.” Dunning v. State, 572 S.W.3d 685, 692 (Tex. Crim. App. 2019). Ultimate questions—such as whether a reasonable probability exists that, given the results of the testing, the person would not have been convicted—are mixed questions reviewed de novo. See Rivera v. State, 89 S.W.3d 55, 59 (Tex. Crim. App. 2002).

Roberson argues that the convicting court inverted the language in the statute, requiring him to show actual innocence; invoked an erroneous assessment of law; failed to focus on how a reasonable juror would evaluate the evidence; erroneously assessed the facts and erroneously applied the law; and ignored that any number of individuals could have committed the crime. All these arguments rely on the fact that the new results do not tie him as strongly to the crime. Evidence that the DNA profile obtained from the semen sample in the challenged conviction is 1,923 times more likely to be observed if it originated from Roberson than if from an unknown, unrelated individual is better for Roberson than is evidence that the profile is 5.5 or 470 billion times more likely to be observed in the same. The newly reported result may be favorable in the colloquial sense, but favorable in this context is a term of art.

Under Article 64.04, the convicting court “must consider the results after testing is performed to determine the exculpatory value, if any, of the test results and whether the results meet the standard set out in Article 64.04.” Dunning, 572 S.W.3d at 697. Thus, the meaning of *6 “favorable” is tied to whether the DNA evidence, if available at the time of trial, would have created a reasonable probability that the person would not have been convicted.1 Tex. Code Crim. Proc. art. 64.04.

There is no doubt that the correctly calculated DNA statistic undermines the reliability of the most damning DNA evidence presented at the guilt-innocence phase of trial, but the discrete question before the convicting court, and now before us on de novo review, is whether the weakened inculpatory inference from the DNA evidence created a reasonable likelihood of a different outcome. It does not; the revised estimate still inculpates Roberson, and the four results in the related cases were unaffected. These cases could have been available for use at guilt-innocence—admissible under the “mark of Zorro” or “modus operandi” rationales to show identity. See Segundo v. State, 270 S.W.3d 79, 88-89 (Tex. Crim. App. 2008).

In addition, at trial, Steve Robertson, a supervisor at the DPS Crime Lab in Austin, testified that he “tested the semen on the vaginal swabs for blood group typing and secretor status” and “determined that [R.P.’s] assailant was a blood group O secretor or else was a nonsecretor of any blood type with a PGM of 1 minus or 1 minus, 1 plus.” Roberson, 16 S.W.3d at 160. Robertson later received Roberson’s blood sample “and the blood typing revealed [Roberson] to be a blood group O secretor with a PGM type of 1 minus, 1 plus.” Id. at 160-61. Roberson, therefore,

1 In its first iteration, the statute provided that, “After examining the results of testing under Article 64.03, the convicting court shall hold a hearing and make a finding as to whether the results are favorable to the convicted person. For the purposes of this article, results are favorable if, had the results been available before or during the trial of the offense, it is reasonably probable that the person would not have been prosecuted or convicted.” Act of April 5, 2001, 77th Leg., R.S., ch. 2, § 2, 2001 Tex. Gen. Laws 4. The reference to “favorable” does not appear in later versions of the statute. See Act of May 9, 2003, 78th Leg., R.S., ch. 13, § 4, 2003 Tex. Gen. Laws 16; Act of June 17, 2011, 82nd Leg., R.S., ch. 278, § 7, 2011 Tex. Gen. Laws 885; Act of June 17, 2011, 82nd Leg., R.S., ch. 366, § 3, 2011 Tex. Gen. Laws 1016. Nevertheless, as is evident in Dunning, use of the word persists.

“could not be eliminated as being the donor of the semen in question, including the semen stain on [R.P.’s] nightgown.” Id. Robertson stated that “blood group O with a PGM type 1 minus, 1 plus was found in approximately nine to ten percent of the male population.” Id. at 161.

Also, Dr. David Bing, the scientific director of Genomice Collaborative, testified that he performed polymerase chain reaction (PCR) testing on R.P.’s blood, one of the vaginal swabs from the rape kit, and Roberson’s blood. Id. at 162. Bing concluded that Roberson could not be excluded as a donor of the sperm found on the vaginal swab and that the frequency of Roberson’s DNA profile in the population would be 1 in 1600 (under 1992-93 guidelines) and 1 in 1800 (under 1996 guidelines). Id.

Given the complete post-conviction results and trial evidence, the weakened inculpatory inference from the DNA evidence did not create a reasonable likelihood of a different outcome. Dunning, 572 S.W.3d at 697. We overrule Roberson’s five issues challenging the court’s non-favorability finding. 2. Issues Claiming Due Process Violations in the Chapter 64 Proceedings

In his eighth issue, Roberson argues the convicting court violated his constitutional rights by failing to rule on his motions for counsel and for a trial record on appeal. In his second issue in each of his supplemental briefs, Roberson argues the convicting court violated his constitutional rights by refusing to provide him access to the trial record.

Before Roberson filed his briefs, we abated this appeal and remanded it to the convicting court. We ordered the convicting court to conduct a hearing to (1) determine whether Roberson was presently indigent, and if he was so, to order the preparation of the reporter’s record at no cost to him and (2) consider Roberson’s pending request, if any, for the appointment of *8 appellate counsel. See In re Roberson, No. 03-24-00325-CR, 2024 WL 3187641, *1 (Tex. App.— Austin June 27, 2024, order) (per curiam) (not designated for publication). The convicting court held that hearing and (1) found Roberson indigent, noted that it had held the hearing contemplated by article 64.04, and explained that although Roberson would normally be entitled to a free record of that hearing, that hearing had not been transcribed; and (2) allowed the Innocence Project, which had represented Roberson in the Chapter 64 proceedings to that point, to withdraw. When the convicting court asked why the Project wanted to withdraw, counsel replied, “Our representation was tied to the Chapter 64 testing, and when that concluded, so did our representation of Mr. Roberson. We notified him by writing. I believe he’s accepted that as he’s appearing pro se, [he’s] not reached out to us to help him with this appeal, Your Honor.” Counsel acknowledged that if the Project were to file a brief, “it would be an Anders brief.” Roberson stated that he was not requesting counsel “at this time. And I don’t want him representing me at this time.” Roberson requested a trial record. The convicting court told Roberson the issue of entitlement to a trial record was beyond the scope of the issues before it on remand.

The convicting court did not violate Roberson’s limited rights under Chapter 64. First, Roberson does have a limited right to counsel for Chapter 64 proceedings. “The convicting court shall appoint counsel for the convicted person if the person informs the court that the person wishes to submit a motion under this chapter, the court finds reasonable grounds for a motion to be filed, and the court determines that the person is indigent.” Tex. Code Crim. Proc. art. 64.01(c). Roberson got that counsel, and, after the results came back and the Project moved to withdraw from representation, Roberson informed the convicting court he was not requesting replacement counsel. Under these circumstances, we cannot find that the convicting court violated Roberson’s right to counsel as provided for in Chapter 64.

Second, although indigent defendants have the right to a free trial record on direct appeal, they do not have that right for collateral attacks on their convictions. In re Bonilla, 424 S.W.3d 528, 532 (Tex. Crim. App. 2014). It follows that, without a statutory right to such a record, a convicting court would not abuse its discretion in denying a free trial record in a proceeding subsidiary to a collateral attack, which, as discussed below, a Chapter 64 proceeding is. No such statutory right appears in Chapter 64; we therefore overrule Roberson’s issues tied to appellate counsel and a free trial record. 3. Issues Challenging The Underlying Conviction

In issues six and seven, Roberson asserts that Lockhoof and R.P. testified falsely at the 1998 trial. And in the first issue in each of Roberson’s supplemental briefs, Roberson asserts that the evidence is insufficient to sustain his conviction. These trial-based claims are beyond the scope of a Chapter 64 appeal. The remedy in a Chapter 64 proceeding is limited to obtaining DNA evidence and findings based on any test results. Wood v. State, 693 S.W.3d 308, 327 (Tex. Crim. App. 2024). That evidence might then be used in a state or federal habeas proceeding. Id. The Chapter does not confer jurisdiction on an appellate court to consider collateral attacks on the trial court’s judgment or to review anything beyond the scope of the articles in it. In re Garcia, 363 S.W.3d 819, 822 (Tex. App.—Austin 2012, no pet.). We therefore lack subject matter jurisdiction over Roberson’s complaints about the 1998 trial and dismiss those complaints.

CONCLUSION

Having overruled all cognizable issues raised in Roberson’s appeal of the convicting court’s non-favorability finding, we affirm the order making that finding. We dismiss the trial-based complaints.2

__________________________________________ Chari L. Kelly, Justice

Affirmed in Part; Dismissed in Part Before Chief Justice Byrne, Justices Kelly and Ellis Filed: October 2, 2025 Publish

2 We also deny as moot Roberson’s Motion Requesting Court to Take Judicial Notice of Adjudicative Fact (evidence he alleges supports his false evidence claim).