NEWELL, J., delivered the opinion of the unanimous Court. Andrea Gooden was a laboratory technician who, as everyone seems to agree, properly analyzed Appellant’s blood for alcohol content in this case. After Appellant’s trial, it was revealed that she had—before the trial—mistakenly certified a blood alcohol analysis report in an unrelated case where a police officer had mislabeled the submission form accompanying a blood sample. Due to her self-report of the erroneous certification to her supervisor, Gooden had been temporarily removed from casework at the time of Appellant’s trial so she could research and document this incident. The prosecutors in this case, unaware of the problem in the unrelated case, failed to disclose this information to Appellant prior to Gooden’s testimony in Appellant’s trial. The question before us is whether this evidence is material. The post-conviction habeas court concluded it was not and denied Article 11.072 relief. Based upon the record before us, we agree. We reverse the court of appeals’ holding to the contrary and uphold the habeas court’s ruling. BACKGROUND Appellant was convicted for misdemeanor driving while intoxicated (DWI).[1] She did not appeal her conviction. But after a jury convicted Appellant, Andrea Gooden, the lab technician who had analyzed Appellant’s blood, self-reported to the Texas Forensic Science Commission (TFSC) that the crime lab had violated quality control and documentation protocols in another, unrelated case. An investigation followed, and reports by the Houston Office of the Inspector General (OIG) and the TFSC were given to Appellant after they became available. In response, Appellant filed an application for a writ of habeas corpus under Texas Code of Criminal Procedure Article 11.072,[2] alleging that the State had suppressed favorable impeachment evidence in violation of her right to due process.[3] Specifically, she claimed that the Stated failed to disclose that, before Gooden testified in her trial: (1) Gooden certified a mislabeled lab report in an unrelated case; and (2) Gooden’s supervisor, William Arnold, had temporarily removed Gooden from her casework. Appellant argued that the undisclosed evidence would have enabled her to impeach Gooden and either exclude her testimony entirely or discredit it, resulting in an acquittal or a deadlocked jury. The habeas court, presided over by the same judge who presided over the trial, conducted an evidentiary hearing on Appellant’s claims. Both Arnold and Gooden testified at the habeas hearing. The habeas court ultimately denied Appellant’s writ application, finding that the undisclosed evidence was neither favorable nor material. The court of appeals disagreed and reversed. We granted the State’s petition for discretionary review to address whether the court of appeals failed to apply the standard of review correctly in conducting its materiality analysis. Facts Developed at Trial While conducting a traffic stop on another vehicle, Deputy Bounds saw Appellant speed past him in the lane closest to his patrol car. Bounds got into his patrol car, turned on his lights, and pursued the vehicle for a long time before Appellant finally stopped. During the pursuit, Appellant made several unsafe lane changes without signaling, which caused other drivers to slam on their brakes. When Appellant stepped out of her vehicle, she staggered and could not keep her balance. She appeared disoriented. She said that she was coming from a golf course at a country club but was unable to identify the name or location of the club, despite being asked multiple times. She admitted she had consumed three Bud Light beers that day. There was one open can of beer and two cold, unopened cans of beer in her vehicle. She and her car smelled strongly of alcohol. She was visibly intoxicated: she had red, glassy eyes and slurred speech. She appeared confused, being unable to identify the medication she was taking. Bounds requested another deputy to assist him with the traffic stop. Deputy Francis arrived and administered two standard field sobriety tests: the Walk and Turn (WAT) and the One-leg Stand (OLS). Although Deputy Francis made a few mistakes while administering the tests, Appellant failed both tests. During the WAT test, Appellant exhibited five out of eight clues of intoxication.[4] And during the OLS test, Appellant exhibited four out of four clues of intoxication.[5] Deputy Bounds determined that Appellant was “intoxicated” because she had lost the normal use of her mental and physical faculties.[6] Deputy Bounds placed Appellant under arrest for DWI. After Appellant refused to give a sample of her breath or blood for alcohol analysis, Deputy Bounds secured a warrant to obtain a sample of her blood. A registered nurse drew Appellant’s blood. Appellant’s blood vials were labeled with Deputy Bound’s initials, Appellant’s name, and the case number. The vials were placed into an envelope, labeled with Appellant’s name and the case number, and sealed with tape. Deputy Bounds delivered the blood vials to a secure lockbox at the Houston Police Department. Andrea Gooden, an analyst from the Houston Police Department Crime Lab, retrieved the sealed envelope containing Appellant’s blood samples. There did not appear to be any tampering with the envelope. Prior to testing Appellant’s blood sample, Gooden observed Appellant’s name on the envelope and Appellant’s name on the blood vials inside that envelope. The vials contained a manufacturing label that, in addition to Appellant’s name, had the same case number, initials, and date as the envelope. Gooden followed all of the lab’s standard operating procedures that were in place at the time she analyzed Appellant’s blood. Her analysis revealed a blood alcohol concentration (BAC) of 0.193, which is above the legal limit of 0.08. The jury found Appellant guilty of driving while intoxicated.[7] Then after a special issue hearing, the jury found that Appellant’s BAC was 0.15 or more at the time the analysis was performed.[8] The jury was dismissed and the court proceeded to the punishment phase. The judge pronounced that Appellant was convicted of a Class A misdemeanor, sentenced her to five days in jail, and assessed a $2,000 fine. Post-Conviction Writ Proceedings Prior to Gooden testifying in Appellant’s case, a police officer in an unrelated case turned in an evidence envelope containing a blood sample and a case submission form to the Houston Police Department Crime Lab. Although the officer labeled the blood vials with the correct name and incident number, he put the wrong case information on the submission form.[9] The analyst who received the sample noted the discrepancy in the case folder: The name of the submission form and LIMS is “Roman-Reyna, Francis Javier.” The name on the envelope & blood tubes is “Hurtado, David.” The tubes have the incident # “124607913″, which is not on LIMS.[10] This first analyst contacted the submitting officer and requested that he provide a corrected submission form. While waiting for the revised form, Gooden worked on the case. Pursuant to common lab practice, Gooden analyzed the blood sample, drafted a blood alcohol report, and set the case aside without signing the report until the officer could resolve the name discrepancy. Like the first analyst, Gooden noted the discrepancy. First, she labeled the blood evidence to indicate that the analysis was complete but it was being held pending the revised submission form. Second, she corrected the batch-technical-review report by crossing out Roman-Reyna’s name and handwriting Hurtado’s name. However, about a month later, Gooden mistakenly signed the “certificate of analysis” on the blood alcohol report she had previously set aside. She then placed the report in the queue for review. The report corresponded to the incorrect defendant listed in the case submission form (Roman-Reyna). But the case folder correctly noted the name of the defendant who had provided the blood sample (Hurtado). That same day, William Arnold, the lab’s interim toxicology manager, conducted both the technical and administrative reviews of the case. Arnold did not catch that the report had been submitted under the wrong defendant’s name. And he approved it, causing the erroneous report to be released in LIMS. A few weeks before Appellant’s trial, Gooden noticed that a blood sample belonging to Hurtado had been set aside with a note in her handwriting: “Waiting on Officer Reply already analyzed. —Andrea.” Right away, Gooden investigated. And after researching in LIMS, she discovered that the blood alcohol report had been released to prosecutors handling the case involving (Roman-Reyna). Gooden immediately notified her superiors. The next day, Arnold emailed Gooden telling her to document what went wrong: Until further notice you are to focus solely on documenting the issues surround[ing] the [errors] in the case we discussed yesterday. Do not handle any evidence, process any data or generate any reports or documentation that is unrelated to your research on this case. *** Generate a document with your findings in memo format in as much detail as you can accurately recall and/or demonstrate via existing documentation. Gooden took this email to mean that she was temporarily removed from casework to draft the memo and ensure the related case documentation was placed in the file. Arnold did not otherwise document this action or disclose it to the District Attorney’s Office. The following day, Gooden sent Arnold her memo and assumed she could return to her casework. But a few days later Arnold told her that although the issue was resolved, she could not return to casework because he and others up the chain of command needed to review her memo. Gooden therefore worked on discovery orders in other cases but did not analyze blood. At the time, Arnold was aware that Gooden was due to testify in Appellant’s upcoming case. The following week, Gooden testified in Appellant’s trial. It was the first time she had testified in a case. Arnold observed Gooden’s trial testimony but did not testify himself. After Appellant’s trial, Arnold made a number of claims related to Gooden’s work status and competency as an analyst. Specifically, during his habeas testimony, Arnold claimed that at the time Gooden testified in Appellant’s trial, she was “suspended” from casework due to the Hurtado incident and his concerns about her overall knowledge base. But the habeas court did not believe Arnold. It found that at the time of Appellant’s trial, Gooden had been temporarily removed from casework to document the Hurtado error. And Arnold’s use of the term “suspended” or “under suspension” to describe Gooden’s work status was “suspect” and “unpersuasive.” Moreover, the investigations by the OIG and TFSC showed that Arnold provided inconsistent reasons for removing Gooden from casework, long before his habeas testimony.[11] Before Appellant’s trial, Arnold told Gooden that it was so she could focus solely on documenting the issues surrounding the Hurtado incident. Yet after Appellant’s trial, Arnold told Gooden (twice) that it was due to her trial testimony. Then, after Gooden had contacted the American Society of Crime Lab Directors (ASCLD), self-disclosed to the TFSC, and communicated with the lab’s HR director about returning to casework,[12] Arnold said that her “suspension” was due to the Hurtado incident coupled with his concerns about her knowledge base.[13] But this was the first time that Gooden was labeled as “suspended.” And despite his claims that he had been having doubts about her ability to testify, Arnold let Gooden testify in Appellant’s trial a few weeks after he had allegedly formed these concerns.[14] The investigation by the TFSC also showed other reasons for discounting Arnold’s description of events. Among them, Arnold attempted to blame Gooden for the reporting error in the Hurtado case. Yet all interviewees who participated in the TFSC’s investigation believed Gooden to be a competent analyst who was “unfairly blamed.” The prevailing view was that Arnold was supposed to catch the errors in the report, such as the wrong name, during his technical and administrative reviews of the case.[15] He did not. Specifically, Arnold had not reviewed the case in sufficient detail: he failed to notice that the incident number on the report did not match the blood vials; that the alleged suspect in the submitted incident form only had a breath test rather than a blood draw; that the first analyst had noted the name discrepancy in the case folder; and that this case was from an earlier batch-technical-review report from December (on which Gooden had hand-written Hurtado’s name as a correction). Arnold also failed to make a connection between previous inquiries about missing evidence in the Hurtado case and the information noted in the case folder.[16] Moreover, only days after approving the erroneously certified report, Arnold received an email from the submitting officer about the missing blood evidence for Hurtado. In it, the officer explained that “the [Hurtado] case was mixed up with another case” due to his error on the submission form.[17] This email did not trigger any follow-up or investigation in LIMS by Arnold. The first analyst said it was her understanding that Arnold was taking over any necessary follow-up on the case. About one month before Appellant’s trial, Arnold received another email from the submitting officer. It stated that the blood evidence belonged to Hurtado—just as the first analyst and Gooden had noted in the case folder. Once again, Arnold did nothing about it. Because of Arnold’s inaction, Hurtado’s blood-alcohol results remained incorrectly assigned to Roman-Reyna in LIMS. Then, a few weeks before Appellant’s trial, Gooden told Arnold about the mistake in the released report. Although he withdrew the report from LIMS before it had been accessed by the prosecution, Arnold took no additional steps to fix the report. The report remained uncorrected for months. The trial court considered this evidence and denied Appellant’s habeas application. APPEAL On appeal, Appellant argued that the habeas court erred in concluding that the undisclosed evidence was neither favorable nor material.[18] In its substitute majority opinion, the court of appeals agreed and reversed the habeas court’s denial of Appellant’s writ application, granted habeas relief, and remanded the case for further proceedings.[19] The court concluded that the undisclosed evidence was favorable because Appellant could have used it on cross-examination to impeach portions of Gooden’s testimony or to get the trial court to exclude her testimony entirely. In making this determination, the court of appeals relied heavily on Arnold’s testimony that he lacked confidence in Gooden’s overall knowledge base. The court also concluded that the undisclosed evidence was material because Gooden’s testimony was necessary for the jury to make an affirmative finding on the special issue of whether Appellant’s BAC level was 0.15 or more. Justice Donovan dissented. He concluded that the undisclosed evidence was not favorable because it would not impeach the evidence that Appellant’s blood was analyzed correctly and that Appellant had a BAC level of 0.193. This was especially true given the habeas court’s unchallenged findings of fact that the blood samples were labeled as Appellant’s and that there was no evidence of any errors in Gooden’s analysis of Appellant’s blood. Moreover, the court was obligated to defer to the habeas court’s credibility assessment of Arnold—which included the finding that Arnold’s claims disparaging Gooden were not credible. Thus, according to the dissent, the undisclosed evidence was not material because the likelihood of a different result was not great enough to undermine confidence in the outcome of the trial. STANDARD OF REVIEW An appellate court reviewing a habeas judge’s ruling in an Article 11.072 application for writ of habeas corpus must view the evidence in the record in the light most favorable to the judge’s ruling and must uphold that ruling absent an abuse of discretion.[20] An appellate court affords almost total deference to a habeas court’s factual findings when they are supported by the record, especially when those findings are based on credibility and demeanor.[21] This degree of deference also applies to any implied findings and conclusions supported by the record.[22] However, if the resolution of the ultimate question turns only on the application of legal standards, the appellate court reviews those determinations de novo.[23] The reviewing court will uphold the habeas court’s ruling if it is correct under any theory of applicable law.[24] Determining whether particular evidence was “material” as part of a claimed Brady violation is a mixed question of law and fact.[25] We give deference to the habeas court’s factual findings underlying its decision but review the ultimate legal conclusion of materiality de novo.[26] We have previously addressed a significant distinction between the posture of article 11.07 habeas cases and article 11.072 habeas cases when it comes to the standard of review.[27] In article 11.07 habeas cases, the habeas court is the original fact finder but this Court is the ultimate fact finder.[28] The habeas court’s findings are not automatically binding upon us, although we usually accept them if they are supported by the record.[29] But in article 11.072 habeas cases, the trial judge is the sole fact finder.[30] The court of appeals and this Court are truly appellate courts.[31] We have less leeway in an article 11.072 context to disregard the habeas court’s findings.[32] BRADY CLAIM An applicant for a post-conviction writ of habeas corpus bears the burden of proving her claim by a preponderance of the evidence.[33] To demonstrate that she is entitled to post-conviction relief on the basis of reversible error under Brady, a habeas applicant must show that (1) the State failed to disclose evidence, regardless of the prosecution’s good or bad faith; (2) the withheld evidence is favorable to her; and (3) the evidence is material.[34] As we stated in Ex parte Miles, “Favorable evidence is that which, if disclosed and used effectively, ‘may make a difference between conviction and acquittal.’”[35] Favorable evidence includes both exculpatory evidence and impeachment evidence.[36] Exculpatory evidence is that which may “justify, excuse, or clear the defendant from fault,” while impeachment evidence is that which “disputes, disparages, denies, or contradicts other evidence.”[37] The nondisclosure of favorable evidence violates due process only if it is “material” to guilt or punishment.[38] Evidence is material only if there is a reasonable probability that, had it been disclosed, the outcome of the trial would have been different.[39] “A ‘reasonable probability’ is a probability sufficient to undermine confidence in the outcome.”[40] Materiality is determined by examining the alleged error in the context of the entire record and overall strength of the state’s case.[41] The suppressed evidence is considered collectively, not item-by-item.[42] ANALYSIS After viewing the evidence in the light most favorable to the habeas court’s ruling, we hold that the habeas court was within its discretion to conclude that the undisclosed evidence was not material.[43] The writ record generally supports and confirms this conclusion.[44] The undisclosed evidence Appellant complains about is (1) Gooden’s certification of a report in another case that contained a labeling error by the submitting officer; and (2) Gooden’s temporary removal from performing her regular job duties to provide documentation regarding that error. At the outset, there was overwhelming evidence of Appellant’s intoxication to support the jury’s finding of guilt regardless of Gooden’s testimony.[45] Appellant was speeding and made several unsafe lane changes without signaling. Despite the officer’s illuminating lights, she took a long time to stop her vehicle. When she exited her vehicle, Appellant was visibly intoxicated: her eyes were red and glassy, her speech was slurred, and she appeared confused. She and her car smelled strongly of alcohol. She admitted that she had been drinking. She had one open can and two cold, unopened cans of beer in her vehicle. She had lost the normal use of her mental faculties: she was incoherent and unable to remember where she was coming from or the medicine she had taken. She had also lost the normal use of her physical faculties: she staggered, could not keep her balance when she stepped out of her car, and failed both of the field sobriety tests. The undisclosed evidence impeaching Gooden would not have impeached the testimony describing Appellant’s intoxicated state. The habeas court’s unchallenged findings of fact regarding Gooden’s analysis and testimony also supports the conclusion that the undisclosed evidence was not material to the jury’s special-issue finding that Appellant’s BAC was 0.15 or more. First, the habeas court found, and Appellant does not challenge, that there was no evidence of any error in the labeling of Appellant’s blood or Gooden’s analysis of Appellant’s blood. This finding is supported by the record. Appellant’s blood vials were labeled with the officer’s initials, Appellant’s name, and the case number. The vials were placed into an envelope labeled with Appellant’s name and correct case number, sealed with tape, and delivered to a secure lockbox. Gooden retrieved the envelope, which showed no signs of tampering. Prior to testing Appellant’s blood, Gooden verified that the name on the envelope matched the name on the blood vials inside; it was Appellant’s. Gooden followed all of the lab’s standard operating procedures when analyzing Appellant’s blood. Her analysis of Appellant’s blood sample revealed a blood alcohol level of 0.193. Moreover, Gooden’s error in the Hurtado case was a protocol error regarding the certification of the report as complete. It was not a mislabeling or analysis error. The officer, not Gooden, had mislabeled the submission form accompanying the blood evidence. The correctness of Gooden’s analysis of that evidence in the Hurtado case was never in question. Finally, Gooden’s certification only moved the report to the next stage; the report still had to be administratively and technically reviewed, and then approved, before it was released. Furthermore, the habeas court, as the exclusive judge of the credibility of the witnesses, determined that Arnold’s description of events and his views regarding Gooden’s performance were not credible.[46] The habeas court rejected Arnold’s purported reasons for removing Gooden from casework, including his alleged concerns about her knowledge base and inability to answer basic questions. Instead, it found that at the time of Appellant’s trial Gooden “had simply been removed to focus solely on documenting the issues surrounding an unrelated mislabeled blood case.”[47] The habeas court also rejected Arnold’s characterization of Gooden’s work status. It found Arnold’s use of the term “suspended” or “under suspension” to describe Gooden’s work status at the time of Appellant trial “suspect and unpersuasive.” The habeas court’s credibility assessment of Arnold, and the findings based upon it, are supported by the record.[48] In its report introduced at the hearing on the habeas application, the TFSC found Arnold professionally negligent, not Gooden.[49] All interviewees participating in the TFSC’s investigation believed Gooden to be a competent analyst who was unfairly blamed for the reporting error in the Hurtado case. In contrast, Arnold provided inconsistent explanations regarding why Gooden was removed from casework. Given the habeas court’s findings of fact and determinations of credibility, we agree with the State that the undisclosed evidence would not have been material impeachment evidence against Gooden sufficient to undermine her testimony regarding the accuracy of Appellant’s blood results.[50] Gooden’s certification of the Hurtado report was a one-time incident in an unrelated case.[51] Appellant’s blood sample and subsequent report contained no labeling errors. And Gooden’s temporary removal from casework at the time of Appellant’s trial does not overcome the fact that there was no evidence of any error in Gooden’s analysis of Appellant’s blood. Finally, Arnold’s claims regarding his concerns about Gooden’s knowledge base were not credible and therefore do not undermine the reliability of her testimony that Appellant’s BAC was 0.193. Whatever impeachment value this new information had to undermine Gooden’s testimony, it was nevertheless not material. We agree with the State that the court of appeals placed too much weight on Arnold’s testimony in its analysis even though the habeas court found Arnold lacked credibility. While the court of appeals did not have to defer to the habeas court’s conclusion regarding the legal determination of materiality, the factual findings underlying that legal conclusion support the habeas court’s denial of relief. Viewing the habeas court’s factual findings with the proper degree of deference, we hold that the habeas court properly concluded that Appellant failed to carry her burden to establish that the undisclosed evidence was material. CONCLUSION Based upon the habeas court’s factual findings, the undisclosed evidence at issue in this case was not material. Gooden’s analysis and testimony in Appellant’s case was proper and reliable. Her error in the unrelated case had to do with an improper certification of a report rather than a failure to catch a mislabeling of a blood sample. And there was overwhelming and uncontested evidence of Appellant’s intoxication to sustain Appellant’s conviction for Class A misdemeanor DWI. Consequently, we reverse the judgment of the court of appeals and affirm the habeas court’s ruling. Delivered: June 10, 2020 Publish
