OPINION
The State of Texas filed a petition to civilly commit Darryl Wayne Day as a sexually violent predator. See Tex. Health & Safety Code Ann. §§ 841.001-.150 (West 2010). A jury found Day suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence. Id. § 841.003(a). The trial court entered final judgment and an order of civil commitment under the Act. We affirm the trial court’s judgment.
OFFENSE DETAILS AND OTHER BAD ACTS
In issue one, Day complains that the trial court allowed the State to develop testimony concerning the details of Day’s past offenses, the prejudicial effect of which substantially outweighed the probative value of that evidence. See Tex.R. Evid. 403. Day identifies seven points in the record where he contends the State presented details of prior misconduct from unidentified prosecution records. He includes the State’s opening statement and closing argument. The lawyer’s statements in opening and closing argument are not evidence. See McCain v. NME Hosps., Inc., 856 S.W.2d 751, 757 (Tex.App.-Dallas 1993, no writ).
Day also complains that the State read his responses to requests for admissions into the record, but he did not object to that action at trial. See Tex.R.App. P. 33.1. Likewise, Day complains that his examination by the State regarding the predicate offenses and disciplinary infractions occurring during his imprisonment added little to the proof of any relevant fact, but Day did not object during the trial. To preserve error concerning the admission of evidence at trial, the appellant must make a timely objection that states the specific ground of objection. Tex.R. Evid. 103(a)(1); see also Tex.R.App. P. 33.1.
Through three expert witnesses, the State developed testimony about the details of the offenses for which Day was imprisoned, the commission of other crimes and bad acts for which he was not convicted, and his behavior while in prison. One expert, forensic psychologist Walter Y. Quijano, was retained by Day. Quijano stated that he reviewed Day’s records and partly relied on those records in evaluating Day for a behavioral abnormality. When the State asked Quijano if Day forced one of his victims into his car and told her to take off her jeans and panties, Day objected that “these are hearsay statements taken from the records.” Compare Tex.R. Evid. 802 with Tex.R. Evid. 403. Because Day did not make a Rule 403 objection to the testimony elicited from Quijano, he failed to preserve error on appeal. See Tex.R. Evid. 103(a)(1); Tex.R.App. P. 33.1.
The State asked its forensic psychologist, Jason D. Dunham, “And from your review of the records what happened in the sexual assault of a child[?]” Day objected that the question called for hearsay and that the probative value of the evidence would be substantially outweighed by the danger of unfair prejudice. See Tex.R. Evid. 802; see also Tex.R. Evid. 403, 705(d). The trial court overruled the objections, and at the request of counsel instructed the jury “hearsay normally is not admissible in trial. However, certain hearsay information contained in records reviewed by the expert is allowed into evidence through expert testimony. Such *198evidence is admitted only for the purpose of showing the basis of the expert’s opinion.” Without further Rule 403 objections, Dunham testified about the details of the offenses, as revealed in the records he reviewed, and explained for the jury the significance of the information he read in the records and how that information related to the actuarial instruments used as predictors for rearrest and reconviction. Dunham expressed his opinion that Day was at very high risk to reoffend when he entered prison and while imprisoned Day did nothing through treatment to statistically mitigate that risk, so in Dunham’s opinion Day is “a very high risk when he leaves as well.”
Another State’s expert, forensic psychiatrist Michael R. Arambula, testified that he reviewed Day’s records and personally interviewed Day in conducting his evaluation of Day for a behavioral abnormality. Counsel asked Arambula if Day discussed his childhood during the interview. Aram-bula responded that Day minimized any physical abuse that may have occurred during his childhood. Counsel asked Ar-ambula whether he saw “anywhere in the record that there was any evidence of abuse?” Day objected, but the trial court overruled Day’s hearsay objection. See Tex.R. Evid. 802. Counsel for Day then objected that “its prejudicial value is outweighed by its probative.” See Tex.R. Evid. 403. The trial court overruled that objection. Day next objected that “the danger that hearsay may be used for a purpose other than an explanation or support of the expert’s opinion outweighs its value and explanation or support.” See Tex.R. Evid. 705(d). The trial court overruled that objection but provided a limiting instruction to the jury that was substantially the same as the instruction given to the jury when Dunham testified. Counsel then asked whether, outside of the interview with Day, Arambula saw any evidence of childhood physical abuse in Day’s records. Arambula replied, “There were some notations in past records where he had divulged that to someone.” Arambula testified on direct examination without further objection. Arambula diagnosed Day with “paraphilia not otherwise specified with features of sadism,” alcohol abuse, and “personality disorder not otherwise specified with features of antisocial personality.” Arambula explained that he used that diagnosis, rather than a diagnosis of sadism, because while the behavior occurred over a period greater than six months, it only repeated itself three times within four years. Arambula also explained that in evaluating cases he determines what the details of the rape are and then places the act on a continuum of violence, considering the intensity and degree of sexual violence. According to Ar-ambula, Day minimized what happened in the assaults.
Day argues that his own testimony established that his account of the incidents varied from that contained in the records, and he suggests that the repeated recounting of the details of the offenses and other bad acts that are revealed in his records served to create outrage in the jury without adding to the determination of a behavioral abnormality. But Dunham diagnosed Day with antisocial personality disorder, while Arambula made a diagnosis of personality disorder not otherwise specified with features of antisocial personality. Arambula explained that at the time he evaluated Day the information available to him did not meet the criteria for determining that Day had conduct disorder during his youth, while Dunham considered Day’s non-sexual criminal history to be very important in determining that Day has a personality disorder. Dun-ham further explained that the details of the offense reveal risk factors, such as *199lack of empathy, and behavior while incarcerated helps reveal that Day continues to victimize others. Although both Dunham and Arambula conclude that Day has a behavioral abnormality, the information each expert considered and the analytical process each expert employed differed from each other and from Quijano. Having each expert to explain which facts were considered and how those facts influenced his evaluation assisted the jury in weighing each expert’s testimony and the opinion each offered regarding the ultimate issue in the case. In re Commitment of Wilson, No. 09-08-00043-CV, 2009 WL 2616921, *9-10 (Tex.App.-Beaumont Aug. 27, 2009, no pet.) (mem. op.). We also presume the jury followed the trial court’s limiting instruction. Id. at *9 We hold the trial court acted within its discretion in allowing the experts to discuss the details of the offenses and other bad acts committed by Day that are contained in the records they reviewed. We overrule issue one.
EXCLUDED EVIDENCE
In issue two, Day contends the trial court erred in sustaining the State’s objection to Day’s testimony concerning the nature of major and minor disciplinary actions during his incarceration. The State called Day as a witness during its case-in-chief. During that examination, counsel asked Day about the disciplinary actions he received during his incarceration. Counsel mentioned the type of infraction involved in thirteen of the thirty-nine disciplinary actions. On the defense’s case-in-chief, Day’s counsel asked Day to describe a disciplinary infraction. Day explained that breaking a rule while in the correctional system would result in a disciplinary infraction, and that there are minor and major disciplinary infractions. Day gave an example of a major infraction and explained that the Texas Department of Criminal Justice classes infractions by three levels, that is, major, discretionary major or minor, and automatic minor. When counsel asked Day to provide an example of a minor infraction, the State objected on grounds of relevance, and the trial court sustained the objection. Counsel explained that the State had developed testimony about major and minor infractions and he wanted Day “to clear up what they are.” The trial court noted that “I don’t think Mr. Day can clear it up for them.” Later in Day’s testimony, counsel for the State asked Day about a number of infractions that counsel identified as “major” cases and Day’s counsel asked Day about a number of infractions that counsel referred to as “minor” infractions.
On appeal, Day argues that the proffered testimony was relevant because the State repeatedly referred to “major” infractions in their experts’ testimony. See Tex.R. Evid. 402. Day spent twenty years as an inmate and incurred thirty-nine disciplinary infractions, and on appeal he argues that he demonstrated his lay witness knowledge of the prison disciplinary system. See Tex.R. Evid. 601(a), 602. The State argues that the error, if any, was harmless because the jury heard what the disciplinary actions were for and whether they were considered major or minor infractions. See Tex.R.App. P. 44.1(a).
When a ruling excludes evidence, to preserve error the appellant must have made the substance of the evidence known to the trial court through an offer of proof, unless the substance of the evidence was apparent from the context within which the question was asked. See Tex.R. Evid. 103(a)(2); Tex.R.App. P. 33.1(a)(1)(B). “To properly pass on the question of the exclusion of testimony, the record should indicate the questions that would have been asked, what the answers would have *200been and what was expected to be proved by those answers.” Lopez v. S. Pac. Transp. Co., 847 S.W.2d 330, 336 (Tex.App.-El Paso 1993, no writ). Day has not shown that he perfected a bill of exception or made an offer of proof. See Smith v. Smith, 143 S.W.3d 206, 211 (Tex.App.Waco 2004, no pet.). Because error has not been preserved, we overrule issue two.
LEGAL SUFFICIENCY
In issue three, Day contends legally insufficient evidence supports the jury’s verdict that Day suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence. We apply the appellate standard of review adopted in Mullens to a legal sufficiency challenge in an appeal from an order of commitment as a sexually violent predator. In re Commitment of Mullens, 92 S.W.3d 881, 885 (Tex.App.-Beaumont 2002, pet. denied). We review all of the evidence in the light most favorable to the verdict, and we consider whether a rational jury could have found, beyond a reasonable doubt, that Day suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence. Id. (citing Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979)).
Conclusory or speculative expert testimony does not make the existence of a material fact more probable or less probable. Wal-Mart Stores, Inc. v. Merrell, 313 S.W.3d 837, 839 (Tex.2010). Day compares his case to Merrell, a wrongful death suit in which the plaintiffs’ expert attributed a late-night living room fire to an exploding halogen lamp. Id. at 838. In Merrell, the expert ruled out the use of lit smoking materials as a cause of the fire because the report did not show evidence of burnt cigarettes at the point of the fire, but the expert failed to account for why he did so when there was also no evidence of charred or exploded glass to support his own theory. Id. at 839. The expert’s failure to explain or adequately disprove alternative theories of causation made the expert’s own theory speculative and con-clusory. Id. at 840. The expert laid a general foundation for the dangerousness of halogen lamps, but only speculated that the halogen lamp had caused the fatal fire. Id. The expert’s testimony lacked objective, evidence-based support for his conclusions. Id. Holding that the expert’s testimony was legally insufficient to support causation, the Supreme Court rendered judgment that the plaintiffs take nothing. Id.
Day contends that the expert opinion testimony of Arambula and Dunham provide no evidentiary support for the jury’s verdict. Day argues that these experts failed to identify whether his psychological condition was congenital or acquired, and further failed to explain their analytical processes or describe any condition from which he suffers. According to Day, Dun-ham gathers unverified information from the Special Prosecution Unit, conducts one interview with the subject of the evaluation, performs tests without the subject being present, develops the case in his mind and formulates his opinion based on a supposedly scientific theory that past behavior is a good predictor of future behavior. Day notes that Arambula also reviews records, meets once with the subject, and determines whether the individual has a behavioral abnormality based on a professional manual that does not describe “behavioral abnormality.”
Day describes an analytical process that involves reviewing records, interviewing the subject, and formulating an opinion. Day’s own expert, Quijano, testified that a thorough evaluation requires him to review all of Day’s records and to conduct a thor*201ough interview in which he determines whether the victims were strangers, whether Day had contradicted himself in the records, and whether Day suffers from a mental illness that would make the interview invalid. He evaluates whether the subject has a behavior abnormality today, not whether he would benefit from sex offender treatment. According to Quijano, actuarial testing is a significant part of what forensic psychologists do in evaluating a person for behavioral abnormality. The Static-99 is accepted in the field of psychology, as is the Minnesota Sex Offender Screening Tool Revised (“MnSOST-R”), although the MnSOST-R is not as accepted as the Static. The Psychopathy Checklist Revised (“PCL-R”) is also accepted in the field of psychology. Both Day’s and the State’s forensic psychologists employed the same methodology and the same tests, made similar findings regarding their diagnoses of Day’s psychological condition, but they differed in their conclusions regarding whether Day’s psychological condition predisposes him to engage in a predatory act of sexual violence.
Dr. Dunham’s Methodology
Dunham explained for the jury that as a forensic psychologist he conducts different types of evaluation. Some evaluations, such as cases in which a defendant has entered an insanity plea, are past-based and concern evaluating someone’s mental state at the time of the crime. Other evaluations, such as determinations of competency to stand trial, are based on the current time. Future-based evaluations fall in the area of risk assessments. Forensic psychologists typically testify regarding whether a person has a behavioral abnormality, or as it is known in some jurisdictions, a mental abnormality. According to Dunham, behavioral abnormality is a legal term that describes a congenital or acquired condition that affects someone’s emotional or volitional capacity such that it predisposes someone to commit an act of sexual violence to the extent that it causes a menace to society. Most psychological disorders that place a person at risk to commit a sex offense in the future have been acquired over time. Volitional capacity refers to one’s desires, while emotional capacity refers to a person’s impulses. A condition that affects emotional or volitional capacity may place someone at risk or make them susceptible, but it does not mean they have a total lack of control.
Dunham explained that the records he reviews include police reports and sex offender treatment records. Based on the information in the records, which in this case covered the individual from age ten to the present, Dunham formulates a list of questions to ask the person being evaluated. Dunham always asks the subject about childhood, alcohol use, and drug use. The comprehensive interview lasts two to three hours, and includes background, education history, medical history, relationships, sexual offenses, school behavior, nonsexual criminal history, therapy, and prison behavior to the current time. He also inquires about current attitudes and beliefs and plans for the future. If necessary, Dunham conducts a collateral source interview, such as a sex offender treatment provider. He conducts actuarial tests that are based on information in the records of the person being evaluated. Dunham used that methodology in Day’s case, in accordance with Dunham’s training in psychology and the accepted standards in the field of psychology.
Dunham explained that he looks at lifelong behavior patterns. According to Dunham, “when you’re talking about certain sexual deviancy and personality disorders, if they’re well established from early time in life all the way up through adult*202hood I think that’s a pretty good indicator of what will happen in the future.”
Dunham stated that actuarial tests are an important tool used by forensic psychologists performing sexual violence risk assessments. The tests used in evaluating sex offenders include the Static-99, the MnSOST-R, and the PCL-R. The actuarial tests include factors that are known in the literature to correlate with future offending, and they add to the psychologist’s clinical interpretation. The MnSOST-R provides a risk level based on data obtained from the Minnesota prison population, and the test uses rearrest data. Because the reoffense rate cannot be reliably measured due to problems inherent in detecting offenses that do not lead to arrest, the MnSOST-R and the Static-99 under represent the true risk of reoffending. On the MnSOST-R, a score of “8” or above places the offender in the high risk category. The Static-99 measures the reconviction rate for sex offenders. Dunham stated that the PCL-R is not an actuarial test, but is a measure of psychopathy that provides a range as to how psychopathic someone is.
On cross-examination, Dunham admitted that the Static-99 uses normative data that separates incarcerated sex offenders into groups that includes a routine sample, non-routine sample, treatment needs sample, and high risk sample. Dunham compared Day to the high risk sample, but he admitted that was a subjective decision. In Dunham’s opinion, the best test to use is the Static-99 because it is very well established, has the best peer review, the best statistical properties, and is used by the most people. The Static-99 authors also have a larger statistical sample than the MnSOST-R. The Static-99 has had a more solid inter-rater reliability because the MnSOST-R provides more choices for the rater and thus has more subjectivity. A new test, the Static-99R, based upon a study that concluded that the importance of a person’s age has been underrepresented, lowered Day’s risk of reoffense at “5 years to 25.2” and at “10 years [to] 35.5.”
Dr. Dunham’s Evaluation
Based on his review of Day’s records, Dunham formed an opinion that Day suffers from a behavioral abnormality. Day’s penitentiary packet includes convictions for sexual assault. Day was nineteen at the time of the first offense, which involved the abduction at gunpoint of a fourteen-year-old girl. During his interview, Day admitted to sexually assaulting the child but he claimed the child was a promiscuous drug-user and he did not mention the use of a weapon. Day received deferred adjudication probation for that offense but was convicted when he received another sexual assault conviction four years later. That assault involved the abduction at gunpoint of a twenty-three-year-old woman. In his interview, Day claimed the encounter was consensual. A third sexual assault charge was no-billed about the same time and the assault that resulted in his second conviction. While in prison, Day was disciplined for masturbating in front of female officers. Dunham determined Day had eight to ten victims.
Day also exhibited a pattern of nonsexual criminal behavior. According to Dunham, Day had fifteen arrests and thirteen convictions. At nineteen, Day committed an assault by shooting at his former girlfriend’s new boyfriend. Day was serving probation on that assault case and the sexual assault of a child case when he committed aggravated sexual assault. Other offenses included resisting arrest, assault on a police officer, another assault and vehicle theft.
On the MnSOST-R, Day scored two points for previous convictions, three *203points for having a sex offending history of one to six years, two points for being charged with an offense while on probation, two points for abductions occurring in a public pace, zero points for use of force, one point for multiple acts in a single event, three points for victimizing in two age groups, two points for age of victim, three points for offending against a stranger, two points for adolescent antisocial behavior, one point for substantial drug or alcohol use, negative two points for employment, one point for discipline history while incarcerated, no points for chemical dependency treatment while incarcerated, no points for sex offender treatment, and negative one point for age at release, for a total of at least seventeen points. Dun-ham placed Day at the highest risk level. Persons at the highest risk level on average are rearrested “at about 72 percent within six years for a hands-on sex offense.” Day’s Static-99 score of six correlates to a “31.2 percent” chance of being reconvicted within five years and a “41.9 percent” chance of being reconvicted within ten years. Additionally, Dunham found Day to show concern for others to be “pretty equal with other prison inmates” and in Dunham’s opinion, Day did not reach the level of a typical psychopath based on the PCL-R.
In determining whether Day has a behavioral abnormality, Dunham considered protective factors, which statistically reduce risk, and positive factors, which might increase chances of success in the community. In making his determination, Dunham also considered Day’s risk factors. According to Dunham, an offender’s risk increases exponentially "with additional offenses, and Dunham saw seven or eight to ten victims, two sex offense convictions, and prison behavior that included sexual misconduct. Dunham found Day used weapons above what is necessary to gain compliance, and that his commission of sexual offenses over time indicated an established deviancy. The presence of weapons and blindfolds, and the similarity in the way the offenses were committed, indicated a degree of planning was involved. Additionally, committing the offenses while on probation indicated an inability to stay out of trouble. The fact that his victims were strangers from different target groups placed more people at risk to be potential victims, and abducting victims from a public place indicated a lack of concern about consequences and indicated Day has difficulty controlling himself. Day’s sexual misconduct included having sex with prostitutes, cheating on his fian-cée, and engaging in sexual conduct while in prison. Day did not participate in anger management or sex offender treatment while in prison. He lacks empathy and remorse for his victims. Day also has a poor appraisal of his own risk, as he claims he is at no risk to reoffend after his release, and therefore may not be aware of the triggers for an offense.
Dunham stated that in his opinion Day has antisocial personality disorder manifested in a chronic, lifelong failure to follow the norms of society. It may remit with age but will never fully go away. Day’s non-conforming behavior started at age eleven with stealing and fighting, and continued throughout the twenty years of his incarceration. Regarding positive factors, Day is not a psychopath, he has a stable employment history, is likeable, and has good social skills. But his antisocial personality order and sexual deviancy continued in prison with actions such as cursing at officers and masturbating toward female officers. Using the Diagnostic and Statistical Manual of Mental Disorders, or DSM-IV, Dunham diagnosed Day with paraphilia not otherwise specified, noncon-sensual type. The paraphilia and the antisocial personality disorder are the psy*204chological conditions that make up Day’s behavioral abnormality. In Dunham’s opinion, Day’s behavioral abnormality predisposes him to engage in a predatory act of sexual violence.
Dr. Dunham’s Opinion
Contrary to Day’s argument, Dunham did not simply look at Day’s past behavior and make his best guess as to whether Day would reoffend. Dunham researched records that reached well into Day’s childhood and personally interviewed Day to gain further insight into Day’s condition, then employed the tests and actuarial tools that are recognized and utilized in his profession to determine whether Day has a psychological condition and to measure the amount of risk to reoffend that Day presents. Dunham’s opinion testimony does not lack objective, evidence-based support for his conclusions. Dunham’s testimony presents a reasoned judgment based upon established research and techniques for his profession and not the mere ipse dixit of a credentialed witness. See Merrell, 813 S.W.3d at 840.
Dr. Arambula’s Methodology
Arambula is a physician specializing in general psychiatry with a subspecialty in forensic psychiatry. Forensic psychiatrists subspecialize in the overlap of law and mental illness. Typically, in conducting a behavioral abnormality evaluation, Arambula is sent a packet of records concerning the individual being evaluated. He interviews the individual and determines whether the individual has a behavioral abnormality. He relies on the facts and data contained in the records in forming the basis for his opinion. According to Arambula, mental conditions develop over time and the physician must determine the patient’s history in order to arrive at a more accurate diagnosis. Arambula uses the DSM-IV in making his diagnoses. The DSM-IV “lists all of the mental conditions in which there’s been considerable research done to where symptoms of an illness that render a diagnosis valid and reliable and predicable, those are all included as criteria.” The criteria listed in order to form the basis of a diagnosis have already been determined in the research to be significant.
The term “behavioral abnormality” is used by the Texas Legislature to describe sexual dangerousness. Arambula looks to the components of the legal definition and determines if the individual has a behavioral abnormality. Paraphilia is a section of the DSM-IV that concerns deviant sexual behavior. The more specific diagnoses within the paraphilic section of the DSM-IV describe conditions that impair an individual’s life and occur with a non-consenting victim. Arambula determines the degree of excitement or gratification and how unusual the behavior is. The behavior must be repeated, because the conditions are chronic and cause impairment to the individual’s life. A diagnosis of not otherwise specified with features of sadism means the individual has a history that does not meet the specific criteria for a diagnosis but involves gratification coupled with the suffering of another individual. The behavior must be present over six months.
Dr. Arambula’s Evaluation
Arambula conducted a three hour interview with Day and reviewed records that started when Day was nineteen. Arambu-la inquired about childhood experiences with Day because behavioral abnormality paraphilias are chronic conditions. Day indicated he had a decent childhood with his mother and grandmother but also stated that there was some harsh treatment. Day’s father was absent from his life, so *205the cause of Day’s condition is at least environmental. Arambula diagnosed Day with personality disorder because of his history.
Arambula gave Day diagnoses of para-philia not otherwise specified with features of sadism, alcohol abuse, and personality disorder not otherwise specified with features of antisocial personality. Arambula used the DSM-IV in making his diagnoses. In Day’s case, while the crimes were committed over more than a six month period, the behavior was repeated just three times within those four years, so Arambula resorted to the more general diagnosis of paraphilia with features of sadism and not specifically sadism.
In evaluating for behavioral abnormality Arambula considers the severity of the assault. The details of the assault determine the intensity and degree of sexual deviance. In the interview, Day minimized what had happened in the assaults. The records showed Day pulled one victim away from her brother, blindfolded her, took her to the woods, and assaulted her sexually, and that the other victim was at a pay phone when Day forced her to get into a car and drove to a secluded place where he assaulted her sexually. A man who is functionally normally would not sustain physical arousal from this behavior, but the humiliation, fright, and resistance is stimulating to a rapist. Those are the marks of abnormal behavior. From a clinical perspective, volitional capacity means that the condition, whether congenital or acquired, affects the individual’s decision-making.
Arambula explained that someone who has features of antisocial personality disorder has a lifestyle that is reckless. The individual does things to excite himself, including exploiting people, and an individual who does not have a strong conscience will engage in denial or minimization, or will not care that his behavior victimizes others. Based on the information Aram-bula had at the time, he did not have enough criteria to say that Day had conduct disorder during his youth, although there were numerous details in his history showing a continuum of antisocial behavior.
Arambula also reasoned that the commission of a second sexual offense while on probation substantiates the antisocial aspect of Day’s personality. Knowing that he was being closely watched did not affect Day’s decision-making. Day’s paraphilia combined with his antisocial personality affected how he was thinking so that he offended again. Paraphilia is a chronic condition that will not go away unless it is addressed. Day’s prison disciplinary record concerned Arambula because he would expect an offender who is incarcerated for a long period of time to accumulate cases early on, and then have the number decrease as the offender adjusts to incarceration. Day’s sex-related infractions occurred later in his incarceration. Day’s violence has not decreased during his imprisonment. Arambula stated his opinion that Day has a behavioral abnormality.
Arambula explained that the diagnosis of alcohol abuse means that at some point in life the person drank more than he should and it caused problems in his life. That condition was determined by Day’s history and was in remission due to his twenty-year incarceration. Day received alcohol treatment in prison. Other positive features include the education Day received in prison and his prior work as an automobile mechanic. Day also has a stable relationship with his grandmother.
During his interview, Day stated that he was willing to participate in sex offender treatment, but at trial Day testified he did not think he needed sex offender treatment. To Arambula, this inconsistency in*206dicated significant denial and a lack of insight. Although Day stated he went through a phase when he was arrogant and macho, Arambula thought Day’s behavior in prison has not shown that he changed much.
On cross-examination, Arambula agreed that in his practice he sees more sexual assault than many of the other paraphilias listed in the DSM-IV, but sexual assault is not specifically listed in the DSM-IV. According to Arambula, close to everyone who commits sexual assault has sadistic features.
Contrary to the argument presented by Day, the basis of Arambula’s underlying methodology is not completely absent. Arambula discussed the criteria for para-philia in the DSM-IV, applied it specifically to Day, and explained for the jury what factors in Day’s history and interview Ar-ambula considered in reaching his conclusion. Arambula’s testimony presents a reasoned judgment based upon established research and techniques for his profession and not the mere ipse dixit of a credentialed witness. See Merrell, 313 S.W.3d at 840.
Considering all of the evidence in the light most favorable to the jury’s verdict, a rational jury could have found, beyond a reasonable doubt, that Day suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence. See Mullens, 92 S.W.3d at 885. We overrule issue three.
FACTUAL SUFFICIENCY
Standard of Review
In issue four, Day contends the evidence is factually insufficient to support the jury’s determination that Day has a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence. Because the beyond-a-reasonable-doubt burden of proof applies in a sexually violent predator commitment proceeding, we adopted the standard of review established by the Court of Criminal Appeals for criminal cases. See In re Commitment of Barbee, 192 S.W.3d 835, 839 (Tex.App.-Beaumont 2006, no pet.). Under the standard for factual sufficiency review previously adopted by this court, “we view all of the evidence in a neutral light and ask whether a jury was rationally justified in finding guilt beyond a reasonable doubt.” In re Commitment of Gollihar, 224 S.W.3d 843, 846 (Tex.App.-Beaumont 2007, no pet.). “To reverse a case on a factual sufficiency challenge, we must be able to say that the great weight and preponderance of the evidence contradicts the jury’s verdict or that the verdict is clearly wrong or manifestly unjust.” Id. The Court of Criminal Appeals recently abandoned factual sufficiency review in criminal cases where the State’s burden of proof is beyond a reasonable doubt. Brooks v. State, 323 S.W.3d 893, 894-95 (Tex.Crim.App.2010). At our request, the parties provided arguments concerning the standard of review in supplemental briefs.
In criminal cases, the federal due process standard of review of the legal sufficiency of the evidence requires the appellate court to defer to the jury on matters of weight and credibility. Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979). If the appellate court reverses a criminal conviction because the evidence is legally insufficient to establish the defendant’s guilt beyond a reasonable doubt, the Double Jeopardy Clause bars a retrial. Burks v. United States, 437 U.S. 1, 18, 98 S.Ct. 2141, 57 L.Ed.2d 1 (1978); Greene v. Massey, 437 U.S. 19, 24, 98 S.Ct. 2151, 57 L.Ed.2d 15 (1978). But “the Double Jeopardy Clause [does not] bar[ ] retrial after a state appellate court sets aside a conviction on the *207ground that the verdict was against ‘the weight of the evidence[ ]’ ” under state law. Tibbs v. Florida, 457 U.S. 31, 32, 102 S.Ct. 2211, 72 L.Ed.2d 652 (1982). Tibbs recognized that when the appellate court reverses a case because it disagrees with the jury’s resolution of the conflicting testimony, after the State has presented sufficient evidence to support the conviction and has persuaded the jury to convict, the appellate court’s disagreement with the jurors’ weighing of the evidence does not require the special deference accorded verdicts of acquittal. Id. 457 U.S. at 42, 102 S.Ct. 2211.
Almost two decades after the United States Supreme Court established the standard of review for the legal sufficiency of the evidence in criminal cases, the Court of Criminal Appeals recognized state-law factual sufficiency review of the evidence used to obtain a conviction. Clewis v. State, 922 S.W.2d 126, 133 (Tex.Crim.App.1996). Over the next decade, the Court of Criminal Appeals struggled to craft a workable standard of review. See Brooks, 323 S.W.3d at 894. The development of a deferential standard of review contributed to the Court’s eventual decision to overrule its precedent and make the legal-sufficiency standard of federal due process the only standard that a reviewing court applies in determining whether the evidence is sufficient to support each element of a criminal offense that the State is required to prove beyond a reasonable doubt. Id. at 894-95 (overruling Clewis factual-sufficiency standard and following Jackson v. Virginia legal-sufficiency standard). A few years before, the Court had recognized that the only apparent difference between legal and factual sufficiency review was that the appellate court views the evidence in a neutral light under a factual-sufficiency standard and in the light most favorable to the verdict under a legal-sufficiency standard. Brooks, 323 S.W.3d at 898-99 (citing Watson v. State, 204 S.W.3d 404, 415 (Tex.Crim.App.2006)). Watson reasoned that, unlike the deferential standard of Jackson v. Virginia, the neutral light analysis adopted in Clewis authorized a reviewing court “to act in the capacity of a so-called ‘thirteenth juror.’” Watson, 204 S.W.3d at 417. However, in Watson the Court noted that under the Clewis standard “it is not enough that the appellate court harbor a subjective level of reasonable doubt to overturn a conviction that is founded on legally sufficient evidence.” Id. When the Court of Criminal Appeals reexamined the factual sufficiency standard of review for criminal cases in Brooks, the Court decided that because Clewis required the reviewing court to defer to the jury on matters of credibility and weight, the reviewing court was not viewing the evidence in a “ ‘neutral light.’ ” Brooks, 323 S.W.3d at 900. As a result, the distinction between legal and factual sufficiency had become indistinguishable. Id. at 902.
In reviewing civil cases, the appellate courts are likewise mindful of the role of the jury. Golden Eagle Archery, Inc. v. Jackson, 116 S.W.3d 757, 761 (Tex.2003) (“[T]he jury is the sole judge of the credibility of witnesses and the weight to be given to their testimony.”). The purpose of factual sufficiency review is not for the appellate court to substitute its judgment for that of the jury, but to determine whether in a particular case the jury functioned correctly in resolving disputed issues of fact. See generally Cropper v. Caterpillar Tractor Co., 754 S.W.2d 646, 651-52 (Tex.1988); Bailey v. Haddy, Dallam 376, 379 (Tex.1841) (To reverse a judgment on the facts, “the illegality or abuse of the verdict ought to be manifest.”).
The Supreme Court established the appellate court’s role in reviewing a jury verdict for factual sufficiency in the years *208before the establishment of the Courts of Civil Appeals. For instance, in Edrington v. Kiger, the former partner of a deceased businessman claimed a debt was owed by the estate over a plea of payment. 4 Tex. 89, 89-90 (1849). A note in evidence had been signed by the deceased and was payable to the plaintiff. Id. at 94. Edrington, the former partner, appealed to the Supreme Court after the jury found for the defendant and the trial court denied a motion for new trial. Id. at 90-91. Edrington’s clerk had testified that Edrington had taken possession of the partnership’s books, but the note was found in the deceased’s papers. Id. at 90. On appeal, the court reasoned that the jury could conclude that the debt had been settled from the unexplained presence of the note in the possession of the maker instead of in Edr-ington’s possession with the rest of the partnership’s papers. Id. at 94. “The Court will never set aside a verdict as against the evidence, merely because they might upon an examination of the evidence, have arrived at a result, different from that found by the jury.” Id. at 94.
In Houston & T.C. Ry. Co. v. Schmidt, the plaintiff testified that the train moved suddenly as he boarded the car at the station, causing him to misstep and catch his foot between two cross-ties. 61 Tex. 282, 283-84 (1884). The other witnesses testified that Schmidt was intoxicated, and some witnesses placed Schmidt on the freight platform rather than the passenger platform. Id. at 284. One witness stated that Schmidt fell off the freight platform. Id. Other witnesses claimed Schmidt gave inconsistent accounts of how he came to be injured, though Schmidt denied making the statements. Id. at 285. One doctor testified that the injury might have occurred either in the manner Schmidt claimed, or by jumping from a height. Id. The treating surgeon testified that in his experience it is more than probable that the injury was received from a leap. Id. The jury found for the plaintiff, and the trial court denied the railway’s motion for new trial. Id. On appeal, the Supreme Court held that if Schmidt was injured in the manner he claimed, it was due to his contributory negligence. Id. “While the verdict of a jury is entitled to great weight when rendered on evidence reasonably sufficient to sustain it, yet, when rendered contrary to evidence, or against the great preponderance of the evidence, and it is most likely that injustice has been done, trial courts should not hesitate to grant new trials.” Id. at 285-86. The Supreme Court remanded the case to the trial court for a new trial. Id. at 286.
In a suit for debt, the Supreme Court reversed a judgment on a jury verdict and remanded the case for a new trial. Dimmitt v. Robbins, 14, Tex. 441, 12 S.W. 94, 99 (1889). Robbins owed $2,500 to his landlord, Dimmitt. Id. at 97. Dimmitt had acquired the property from a man named Smith at a forced sale. Id. Robbins arranged to pay Dimmitt from the sale of some livestock. Id. Robbins claimed that he kept the money from the sale on his person while he and Dimmitt traveled on a route selected by Robbins. Id. at 97-98. As they approached the area where they were to camp, Dimmitt noticed two men in the distance hurriedly mount their horses and disappear. Id. at 98. Robbins told Dimmitt that he had heard that Smith had threatened Dimmitt in a letter to someone named Sterling. Id. Two armed men appeared and demanded a $5,000 ransom from Dimmitt. Id. Robbins offered to give Dimmitt the envelope that supposedly contained $2,500. Id. Dimmitt took the envelope and gave it to the robbers, who refused to count the money. Id. Robbins sued on the debt and Dimmitt claimed duress. Id. at 95. The jury gave a verdict in favor of Rob*209bins. Id. at 96. Examining the evidence from the trial, the Supreme Court noted the conflict in the evidence between Dim-mitt and Robbins. Id. at 98-99. Rather than defer to the jury, however, the court ordered a new trial because “the entire evidence shows a knowledge of and participation in the restraint by appellee exercised upon Dimmitt[.]” Id. at 99.
In Missouri Pacific Railway Company v. Somers, the Supreme Court reversed and ordered a new trial twice. Missouri Pac. Ry. Co. v. Somers, 78 Tex. 439, 14 S.W. 779 (1890) (“Somers II ”); Missouri Pac. Ry. Co. v. Somers, 71 Tex. 700, 9 S.W. 741 (1888) (“Somers I”). The plaintiff worked as a brakeman. Somers I, 9 S.W. at 741. Because the brake was defective, Somers would climb to the lower step at the back of the caboose and throw off the brakes using a tool. Id. Stooping over one day to throw the brakes, Somers struck his head on a cattle-guard as the train passed. Id. The jury found for the plaintiff on an allegation that the cattle-guard was improperly constructed by reason of being in dangerous proximity to the track. Id. On appeal, the court noted that the evidence was conflicting but sufficient to warrant the jury’s finding that the cattle-guard was dangerous. Id. The Supreme Court reversed and remanded for a new trial. Id. at 742. The evidence showed that all of the cattle-guards were the same distance from the track and Som-ers was not a new hand on the track. Id. The court reasoned that Somers “must be held to have taken the risks incident to their general condition.” Id.
In the retrial, Somers tried to explain his testimony from the first trial. Somers II, 14 S.W. at 779. Somers claimed his comments about the condition of the cattle-guards referred to a time period several years before the accident, and that he did not know what struck him but merely repeated what he had been told by the conductor and brakeman. Id. Those men no longer worked for the railroad at the time of trial, and both stated that Somers told them he saw the cattle-guard but thought he could pull his head back in time. Id. The Supreme Court stated that the trial court should have granted the defendant’s motion for new trial. Id. The court reasoned that Somers’s story was contradicted by the conductor and brakeman, both of whom were disinterested. Id. According to the court, Somers relied completely on his own uncorroborated testimony, he gave inconsistent versions of the most material point in the two trials, and he failed to adequately explain that inconsistency. Id. The court concluded, “It seems to us, therefore, that the jury must have been controlled in their verdict more by their sympathy for a hard-working zealous man, injured in a dangerous employment, than by the law of the ease as applied to the evidence.” Id.
Choate v. San Antonio and Aransas Pass Railway Company illustrates the development of factual sufficiency review under the factual conclusivity clause. See San Antonio & A.P. Ry. Co. v. Choate, 22 Tex.Civ.App. 618, 56 S.W. 214 (Tex.Civ. App.-San Antonio 1900, writ ref'd) (Choate VII); Choate v. San Antonio & A.P. Ry. Co., 91 Tex. 406, 44 S.W. 69 (1898) (Choate VI); San Antonio & A.P. Ry. Co. v. Choate, 43 S.W. 537 (Tex.Civ.App.-San Antonio 1897, writ dism’d) (Choate V); Choate v. San Antonio & A.P. Ry. Co., 90 Tex. 82, 37 S.W. 319 (1896) (Choate IV); Choate v. San Antonio & A.P. Ry. Co., 90 Tex. 82, 36 S.W. 247 (1896) (Choate III); San Antonio & A.P. Ry. Co. v. Choate, 90 Tex. 81, 35 S.W. 472 (1896) (Choate II); San Antonio & A.P. Ry. Co. v. Choate, 35 S.W. 180 (Tex.Civ.App.-San Antonio 1896, writ granted) (Choate I). Choate was a passenger on a train that stopped at a station. Choate I, 35 S.W. at 180. Choate *210claimed the train started again suddenly as he was walking between cars to retrieve a crutch he had left in another car. Id. A wheel caught his foot, severing his toes. Id. The trial court had set aside the verdict in two previous trials of the case. Id. Choate claimed he could not recall what happened after he was injured. Id. Other witnesses testified that after the accident Choate stated that he had stepped off the car and slipped as he stepped back on the car. Id. at 180-81. The jury returned a verdict of $2,050 for Choate. Id. at 180. The court of appeals held that “the verdict has no reasonable support in the evidence.” Id. at 181. The court noted that no evidence supported the jury’s verdict and further noted:
It is readily understood how he might, in attempting to get back on the car when in motion, after getting off, have the end of his foot run over by the wheel. Also that in falling off the platform, how, in catching hold of the railing or steps to avoid being hurt, the same thing might happen. This latter theory is advanced by appellee as the one the jury may have adopted.
Id. But Choate had stated that he did not catch himself in the fall, and he was not bruised in a manner that would have been expected if he had been thrown from the train. Id. “[Rjegard for the ordinary laws of nature,” caused the court to conclude that “a jury is not warranted in finding plaintiffs injury came about as alleged.” Id. In remanding the case to the trial court, the court stated that “[i]n the event of another trial, the court should, upon the same testimony, direct a verdict for the defendant.” Id.
On writ of error, the Supreme Court initially held that the evidence, viewed in the light most favorable to the plaintiff, would not justify a jury in finding that the railway was negligent. Choate III, 36 S.W. at 249-50. On motion for rehearing, the court noted that
if the verdict rendered thereon is against the preponderance of the evidence, to that degree which shows that manifest injustice has been done, the trial court may and should grant a new trial. The judge should not invade the province of the jury, and take from it the decision of a question which properly belongs to it. Neither should he abdicate the functions of his office, and permit the prerogatives of the jury to be perverted to the accomplishment of wrong.
Choate IV, 37 S.W. at 319. The Supreme Court held that the Court of Civil Appeals erred by instructing the trial court to direct a verdict. Id.
On retrial, the jury returned a verdict of $5,000 for Choate. Choate V, 43 S.W. at 537. The Court of Civil Appeals expressed its surprise that the Supreme Court had exercised jurisdiction over the issue of fact raised in the previous appeal. Id. at 538. According to the court, “[t]he result of the difference of opinion on a question of fact, which under the constitution should not have arisen, has been that on practically the same facts a jury has again returned a verdict for the appellee, and the trial court necessarily refused a new trial, and the case is again before this court.” Id. Once again, the court noted Choate’s contradictory testimony and the lack of corroborating witnesses. Id. The court noted its opinion that “there is a total lack of testimony to sustain the verdict of the jury[ ]” but reversed and remanded to the trial court without instructions, in accordance with the earlier opinion of the Supreme Court. Id.
On writ of error, the Supreme Court recalled that it had held that “the determination of the court of civil appeals that the verdict should be set aside was conclusive, *211and we therefore affirmed their judgment in so far as it reversed the judgment of the trial court, and remanded the eause[.]” Choate VI, 44 S.W. at 69. The purpose of the factual conclusivity clause was not to enlarge the power of the Court of Civil Appeals, but to restrict the jurisdiction of the Supreme Court. Id. Factual sufficiency review does not “substitute the judgment of the appellate courts upon the facts of a case in place of that of the jury[.]” Id. at 70. In setting aside the verdict, the Court of Civil Appeals remanded the cause, so that a new trial could be had and the right to a jury trial remained inviolate. Id. The court dismissed the writ of error. Id. On retrial, two additional witnesses corroborated Choate’s story about being thrown from the train by an unusual jerk of the train, not incident to the normal operation of the train. Choate VII, 56 S.W. at 215-16. The court affirmed the judgment in favor of Choate. Id. at 216.
It is notable that these early examples of factual sufficiency review did not view the evidence in the light most favorable to the verdict. The deference exercised by the appellate court was to the process of the jury trial and the right to have disputed fact issues determined by a jury. When the case rested on disputed evidence, the appellate court reversed a judgment on the verdict only after considering all of the evidence. If the appellate court determined the evidence did not support the verdict, the court reversed the judgment and remanded for a new trial. The appellate court’s power to order a new trial in a case of manifest injustice acted as a check on what would otherwise be unfettered power placed with the jury. The appellate court thereby maintained the integrity of the court system while the right to trial by jury remained inviolate.
More recently, the Supreme Court clarified that the intermediate appellate courts do not have limitless power to order a retrial. When a Court of Appeals reverses a judgment on insufficiency grounds, the court must “detail the evidence relevant to the issue in consideration and clearly state why the jury’s finding is factually insufficient or is so against the great weight and preponderance as to be manifestly unjust; why it shocks the conscience; or clearly demonstrates bias.” Pool v. Ford Motor Co., 715 S.W.2d 629, 685 (Tex.1986). The court must also “state in what regard the contrary evidence greatly outweighs the evidence in support of the verdict.” Id. The jurisdiction of the Court of Appeals “extends to all fact questions in the case[.]” Cropper, 754 S.W.2d at 651. The constitutional authority of the Court of Appeals and the right to trial by jury are “of equal constitutional stature.” Id.
When an issue of factual sufficiency is raised in a civil case, the factual conclusivity clause of the Texas Constitution requires the Court of Appeals “to consider the fact question of weight and preponderance of all the evidence” and to order a new trial if the jury’s verdict is clearly unjust. In re King’s Estate, 150 Tex. 662, 244 S.W.2d 660, 662 (1951); see also Tex. Const, art. V, § 6. The reviewing court must “consider and weigh all of the evidence in the case” both supporting the verdict and contrary to it. King’s Estate, 244 S.W.2d at 661. When conducting a factual sufficiency review, “the court does not indulge inferences or confine its view to evidence favoring one side of the case. Rather, it looks at all the evidence on both sides and then makes a predominantly intuitive judgment: is the evidence — already identified as ‘some evidence’ ... — in satisfactory harmony with the fact finding it supports?” William *212Powers, Jr. & Jack Ratliff, Another Look at “No Evidence” and “Insufficient Evidence,” 69 Texas L. Rev. 515, 525 (1991). In its most recent opinions on the subject of factual sufficiency, however, the Supreme Court has utilized language suggestive of a deferential standard of review similar to that adopted by the Court of Criminal Appeals in Clewis and rejected by that court in Brooks. See Sw. Bell Tel. Co. v. Garza, 164 S.W.3d 607, 625 (Tex.2004) (“Issues of credibility that depend on appearance and demeanor cannot be weighed by the appellate court; the witnesses are not present. And even when credibility issues are reflected in the written transcript, the appellate court must defer to the jury’s determinations, at least so long as those determinations are not themselves unreasonable.”).
The Supreme Court has considered the effect of the heightened standard of review when the burden of proof is by clear and convincing evidence. See In re J.F.C., 96 S.W.3d 256, 266 (Tex.2002) (legal sufficiency); In re C.H., 89 S.W.3d 17, 25 (Tex.2002) (factual sufficiency). Weighing conflicting evidence and inferences to determine whether a verdict should be vacated as manifestly unjust is appropriately part of a factual sufficiency review. In re J.O.A., 283 S.W.3d 336, 347 (Tex.2009). But in C.H., the intermediate appellate court disregarded much of the evidence supporting the finding that termination would be in the child’s best interest and failed to clearly explain why it concluded that a reasonable jury could not form a firm conviction or belief from all of the evidence that termination would be in the child’s best interest. C.H., 89 S.W.3d at 28-29. The Supreme Court held that the intermediate appellate court conducted its factual sufficiency review incorrectly because the court did not properly adhere to the proper standard, which in that ease meant the appellate court “should not reverse the judgment unless the jury could not reasonably have formed a firm conviction or belief that terminating [the father’s] parental rights was in [the child’s] best interest.” Id. at 27.
The jury charge and instructions generally provide the starting point for understanding what evidence is pertinent to the issue of factual sufficiency under review. Golden Eagle Archery, 116 S.W.3d at 762. A higher quality of evidence is necessary to “tip the scales” in a factual sufficiency review with an elevated burden of proof. Garza, 164 S.W.3d at 625. For instance, “when proof of an allegation must be clear and convincing, even evidence that does more than raise surmise and suspicion will not suffice unless it is capable of producing a firm belief or conviction that the allegation is true.” Id. at 621. The question of whether a person is a sexually violent offender must be proven beyond a reasonable doubt. Tex. Health & Safety Code Ann. § 841.062(a). Because proof of the allegations must be established beyond a reasonable doubt, even evidence that is capable of producing a firm belief or conviction would not suffice unless that evidence is capable of producing a high degree of certainty that is beyond a reasonable doubt. See In re Winship, 397 U.S. 358, 362-364, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970).
In the appeal of a conviction in a criminal case, “[i]f, given all of the evidence, a rational jury would necessarily entertain a reasonable doubt as to the defendant’s guilt, the due process guarantee requires that we reverse and order a judgment of acquittal.” Swearingen v. State, 101 S.W.3d 89, 95 (Tex.Crim.App.2003). By analogy, in the appeal of a commitment as a sexually violent predator, if a rational jury would necessarily entertain a reasonable doubt as to whether the *213respondent suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence, due process requires that we reverse and render judgment. On the other hand, if upon viewing all of the evidence in the light most favorable to the jury’s verdict, if the appellate court finds a rational jury could have found, beyond a reasonable doubt, that the respondent suffers from a behavioral abnormality that predisposes him to engage in a predatory act of sexual violence, then the evidence is legally sufficient.
The State argues that we should apply Brooks to this and other SVP commitment cases because the burden of proof dictates the standard of review, the burden of proof in a criminal case is beyond a reasonable doubt, and the Court of Criminal Appeals has ruled that legal sufficiency is the sole standard of review in criminal cases, where the burden of proof is also beyond a reasonable doubt. See Brooks, 323 S.W.3d at 895. This court has a constitutional duty to review factual sufficiency when the issue is raised on appeal. See Tex. Const, art. V, § 6; Louisianctr-Pacific Corp. v. Andrade, 19 S.W.3d 245, 248 (Tex.1999). The Supreme Court, not the Court of Criminal Appeals, construes the constitution as it is applied in civil cases. See Tex. Const, art. V, § 3 (The Supreme Court’s appellate jurisdiction “shall be final and shall extend to all cases except in criminal law matters and as otherwise provided in this Constitution or by law.”); see also Dallas Area Rapid Transit v. Amalgamated Transit Union Local No. 1338, 273 S.W.3d 659, 666 (Tex.2008) (“[Tjhis Court has never lacked jurisdiction to prevent an intermediate appellate court from conflicting with one of this Court’s decisions.”). Unless the Supreme Court overrules or distinguishes its precedent, we shall continue to follow In re King’s Estate and its progeny. 244 S.W.2d at 662. Finally, we note that although the possibility that the evidence in a particular case will be legally sufficient but factually insufficient essentially decreases as the burden of proof increases, the public policy reasons for retaining factual sufficiency review are greater in a case where an unincarcerated (or soon-to-be-released) person’s liberty is affected. These commitment proceedings are decided on evidence that concerns the application of a “soft” science that calls for the exercise of a considerable amount of intuitive judgment on the part of experts with specialized training. The consequences of an incorrect judgment are great enough that the legal system should retain a factual sufficiency standard of review to minimize the risk of an injustice. In reviewing the factual sufficiency of the evidence in a civil case in which the burden of proof is beyond a reasonable doubt, an appellate court weighs the evidence to determine whether a verdict that is supported by legally sufficient evidence nevertheless reflects a risk of injustice that would compel ordering a new trial. When the burden of proof is beyond-a-reasonable-doubt, the risk of injustice is essentially slight. Nevertheless, if in the view of the appellate court after weighing the evidence, the risk of an injustice remains too great to allow the verdict to stand, the appellate court may grant the defendant a new trial. Accordingly, we shall address Day’s fourth issue, in which he claims that there was factually insufficient evidence to support the jury’s determination that he has a behavioral abnormality.
Application
In issue four, Day repeats his argument from issue three that the testimony of the State’s experts provides no more than bare opinions. Day contends that the experts failed to establish a nexus between the old records of Day’s offenses and their *214conclusions that Day suffers from a congenital or acquired condition. According to Day, the experts “both think about it and come up with a conclusion.” Day argues this methodology does not “constitute evidence legally sufficient to support a judgment of civil commitment.” We have determined, in addressing Day’s third issue, that both Dunham and Arambula provided sufficient bases for their professional opinions. Day contends that Dunham’s and Arambula’s diagnoses include “neither tangible medical findings (defects in the brain or something similar) nor mental conditions recognized in the DSM-IV.” He argues that Dunham and Arambula failed to account for the positive and protective factors in Day’s case. Day further argues that the opinion of his expert, Walter Quijano, shows that the State’s experts’ opinions are too weak to support the judgment.
Dr. Quijano
Walter Y. Quijano is a forensic psychologist who was retained by Day. In his clinical practice, Quijano has performed forensic evaluations of competency, insanity, and future dangerousness. He is a licensed sex offender treatment provider. Quijano testified that he was retained to review the data and offer his professional opinion on whether Day has a congenital or acquired condition that, by affecting a person’s emotional or volitional capacity, predisposes a person to commit a sexually violent offense to the extent the person becomes a menace to the health and safety of others. Because the Health and Safety Code defines a “sexually violent predator” as a person “likely to engage in a predatory act of sexual violence” and other sections of the applicable chapter require a determination of whether the person is “likely to engage in a predatory act of sexual violence,” Quijano sought to determine whether there is at least a fifty percent likelihood that Day will engage in such acts; that is, to determine whether there was a probability of sexual violence and not just a possibility.
Dr. Quijano’s Methodology
In a psychological evaluation of this type, Quijano first reads the records, then reads the other experts’ opinions about the presence or absence of a behavioral abnormality before conducting a clinical interview. In this case, Quijano reviewed the offense reports of three sexual assault cases, including the two offenses for which Day was convicted, the police report for the aggravated assault, and Day’s prison disciplinary history. Quijano confirmed these are the types of records that psychologists use in their evaluations. He also read Dunham’s and Arambula’s depositions. Quijano met with Day once, for one and one-half hours. Quijano investigated Day’s family support because, as he testified, such support is important because it is difficult for one to start over after living in prison for twenty years. In his interview with Day, Quijano was trying to determine whether Day had a mental condition that might invalidate the interview, was attempting to clarify whether the victims in the offenses committed prior to Day’s imprisonment were strangers, and was attempting to verify the reports of sexual misconduct in prison. He was particularly interested in whether Day’s free world victims were strangers because some of Day’s Static-99 forms showed conflicting scores on that issue.
Quijano agreed that to answer the legal question of whether Day has a behavioral abnormality that predisposes him to commit a sexually violent offense, it is necessary to conduct a thorough investigation that includes reviewing all of Day’s records and conducting an interview. In particular, Quijano looked for contradictions in *215the records and asked Day if his victims were strangers. Quijano noted that Day gave an inconsistent account about whether his victims were strangers. He also tried to determine whether Day was mentally ill and whether there were psychological elements that would invalidate the interview. Quijano noted that Day achieved different scores on the Static-99 because he told different stories to different people when they conducted the test. He noted that Day contradicted himself in his statements.
Dr. Quijano’s Evaluation
For eight years, Quijano had worked in prisons and reviewed sexually violent incidents. Quijano expressed confusion about Day’s prison disciplinary cases from 2008 regarding sexual misconduct. Day had pled guilty to them, but Day’s explanation to Quijano sounded reasonable. One of Day’s incidents involved his use of a gesture that appeared to be motivated by anger and the other incident involving exposure apparently occurred due to the lack of privacy in prison. Quijano reviewed Day’s Static-99 and MnSOST-R which were administered by Dunham. Because he did not view Day’s prison offenses as truly sexual, Quijano scored Day differently than Dunham did on the Static-99, but Day was still in the moderate or moderate high range. Quijano did not give much weight to the MnSOST-R administered by Dunham because he is not trained with regard to that actuarial study. He made note of Dunham’s score of 72 or 75 that rated Day as high, but he did not use the test himself. Quijano stated that the research instruments and the literature suggest that the Static-99 is the more “robust” instrument, meaning it has been studied, and he felt it is the most reliable and valid predictor. The Static-99 has more recent data from 2008 to 2010, and has local norms for Texas. He stated the Texas norms correlated to a recidivism range between three to six percent. The test must be used carefully because one point can change the result from moderate to high. The old Static-99 had ten questions, while the new test has fourteen questions that have been found to be the best statistically derived predictors based on “static” circumstances that do not change over time. The new questions include the age of the person at discharge from prison. Quijano indicated that age is a good predictor of recidivism. The older the person is at time of release, the lower the person’s score will be. The new scoring has more age ranges that will possibly lower the final score achieved.
Quijano also reviewed a scoring of the Hare Psychopathy (“Hare-PCL-R”). Qui-jano noted that Day did not measure as a psychopath on the Hare-PCL-R, which was computer scored by the Department of Criminal Justice. Quijano also reviewed the Personality Assessment Inventory, or PAI, which is scored and then associated with certain crime groups. In his opinion, Day did not fit the profile for rapist on the PAI. Other positive factors that Quijano used in forming his opinion were that Day took vocational courses while incarcerated, including industrial cleaning and machine shop. Earned certificates for taking vocational courses are designed to help former inmates obtain employment after release.
Day has never taken any sex offender treatment. According to Quijano, the recidivism rate for persons who receive sex offender treatment in a probation setting is five percent, the recidivism rate for offenders who receive treatment while in prison is fifteen percent, and the recidivism rate for offenders who receive no treatment is twice that amount.
Quijano testified that he never calculated Day’s sexual misconduct disciplinary *216actions into his evaluation because there were no formal charges. Because the 2002 version of the Static only considers actual convictions, not prison disciplinary actions, he excluded Day’ disciplinary actions, but Day still scored moderate to high. Quija-no observed that Day had occasional conflict with some guards but in general, Day had an overall regard for the prison system. Additionally, Quijano dismissed Day’s major prison cases in forming his opinion as to whether Day suffers from a behavioral abnormality, because those cases did not involve violent sexual acts and would not predict sexual violence. Quijano conceded that Day might have a personality disorder but personality disorders are not related to reoffending. He was led to that conclusion because he administered the MMPI, an accepted personality test, to two hundred sex offenders in Conroe and they all “came out normal” on the test.
Quijano explained the DSM-IV is the official classification system of the American Psychiatric Association. The system uses five axes: (1) clinical; (2) personality; (3) health condition that may affect mental health; (4) stressors; and (5) global assessment of functioning, or GAF. Quijano usually uses Axis I and Axis II. Quijano gave Day an Axis I diagnosis ranging from paraphilia NOS to a more firm diagnosis of sexual abuse of a child and sexual abuse of an adult. Day clearly meets those criteria for sexual abuse of a child and an adult, and may meet the requirement of paraphi-lia not otherwise specified. Quijano felt a paraphilia diagnosis was debatable because four years elapsed between Day’s sexually violent offenses and that indicated some ability for self-control. He stated that Day’s particular Axis I diagnosis had no direct significance in formulating his opinion in the case. He felt Day does not meet the criteria for either sadism or features of sadism because a diagnosis of sadism requires pleasure from the infliction of pain, and in this case, it cannot be shown that Day derived any enjoyment from the infliction of pain during his assaults. According to Quijano, an alcohol abuse diagnosis should be given any significance because Day was in remission because of institutionalization. Antisocial personality disorder is a possible diagnosis because Day engaged in criminal activity before age seventeen. Clearly Day meets the definition of adult antisocial behavior, because he committed criminal acts in violation of the rights of others. Adult antisocial behavior affects the behavioral abnormality evaluation because it affects the number of times the subject goes to court, and is convicted, and that affects the score on the Static-99 risk assessment.
On cross-examination, Quijano admitted that he did not ask Day if he had any current sexual fantasies, if he had any more victims, or if he felt a constant urge to commit sexual assault. Also on cross-examination, Quijano testified that rape is not an act that would fall under sexual sadism, but admitted that rape is in the DSM-IV as an example of sexual sadism. However, he explained that Day is a power rapist, not a sadistic rapist. Quijano testified that he believed Dunham’s testimony showed that Day is capable of controlling himself.
Dr. Quijano’s Opinion
In Quijano’s opinion Day clearly met the criteria for adult antisocial behavior for Axis II, but Quijano did not take issue with a diagnosis of personality disorder NOS. Day’s risk factors are having two convictions and use of drugs and alcohol during the time he offended. Given Day’s age and family support, Quijano opined Day is now more likely to be a more stable person than he was twenty years ago. Positive factors and protective factors re*217duce the risk of reoffending. In Day’s case, age and vocational education are positive contributing factors. In Quijano’s opinion, Day does not meet the criteria for behavior abnormality, and the gaps of time between Day’s offenses support his opinion.
Battle of Experts
The three experts who testified in this case disagreed about the most appropriate diagnosis for Day under the DSM-IV. Dunham decided Day has paraphilia NOS, nonconsensual type, and antisocial personality disorder. In scoring the MnSOST-R, Dunham considered Day’s sexual misconduct in prison and placed him at the statistically highest risk level. Arambula made a diagnosis of paraphilia NOS with features of sadism and personality disorder NOS with features of antisocial personality. The main difference between Dunham and Arambula appeared to be Arambula’s assumption that Day engaged in criminal behavior sporadically. Quijano diagnosed Day with an Axis I diagnosis ranging from paraphilia NOS to a more firm diagnosis of sexual abuse of an adult and sexual abuse of a child, and adult antisocial behavior. Quijano felt that Day did not show a lack of volition, but Dunham determined that Day exhibited a lifelong pattern of antisocial behavior. Dunham considered that Day engaged in violent behavior in his youth, then went to prison, where his sexual deviancy and antisocial personality were expressed as violations of the prison’s rules. Quijano stated that he never considered Day’s prison disciplinary infractions in scoring the Static-99 because they were handled administratively and did not lead to criminal convictions. But as Quija-no admitted, even without considering such administrative actions, Day would still be in the moderate to high risk range. Aram-bula determined that the commission of a sexual offense while on probation showed that the combination of paraphilia and antisocial personality affected Day’s decision making. Ultimately, all three experts agreed that Day was at risk to reoffend, and Day never received sex offender treatment that could have lowered that risk. Although Quijano criticized Dunham’s use of the older version of the Static-99, he also admitted that it was the most studied predictor and that using the 2002 Static “did not make a big difference in the out-comet.]”. The subjective aspects of the experts’ analyses and the points of disagreement between the experts were developed for the jury. On this record, we cannot say that the jury could not have discredited Quijano’s professional opinion. Likewise, Dunham’s and Arambula’s opinions were based on accepted techniques and the bases for their opinions were explained to the jury. Their opinions were not so weak that is was unreasonable for the jury to consider them in reaching a verdict. Having weighed all the evidence, we see no injustice in the jury’s verdict that requires a new trial. The evidence is factually sufficient. We overrule issue four.
Excluded Expert Testimony
Issue five contends the trial court erred in refusing to permit Quijano to refer to sections of the Health and Safety Code in explaining his opinion regarding a behavioral abnormality. The trial court excluded testimony from Quijano regarding his interpretation of sections 841.001 and 841.003(a)(2) of the Health and Safety Code. See Tex. Health & Safety Code Ann. § 841.001 (“The legislature finds that a small but extremely dangerous group of sexually violent predators exists and that those predators have a behavioral abnormality that is not amenable to traditional mental illness treatment modalities and that makes the predators likely to engage *218in repeated predatory acts of sexual violence.”), § 841.003(a)(2) (“A person is a sexually violent predator for the purposes of this chapter if the person: (2) suffers from a behavioral abnormality that makes the person likely to engage in a predatory act of sexual violence.”). The trial court also excluded testimony from Quijano regarding two decisions of this court. See In re Commitment of Almaguer, 117 S.W.3d 500, 506 (Tex.App.-Beaumont 2003, pet. denied) (“The definitions and question make clear that a person must have a behavioral abnormality that predisposes the person to commit sexually violent acts, and that the abnormality makes it probable that the person will commit such acts in the future and be a menace to health and safety.”); In re Commitment of Martinez, No. 09-05-493 CV, 2006 WL 2439752, *4 (Tex.App.-Beaumont Aug. 24, 2006, no pet.) (mem. op.) (quoting Kansas v. Crane, 534 U.S. 407, 413, 122 S.Ct. 867, 151 L.Ed.2d 856 (2002)) (“The inability to control behavior ‘must be sufficient to distinguish the dangerous sexual offender whose serious mental illness, abnormality, or disorder subjects him to civil commitment from the dangerous but typical recidivist convicted in an ordinary criminal case.’ ”).
Outside the presence of the jury, Quija-no testified that he used these statutes to help him acquire his definition of behavioral abnormality and formulate his opinion in this case. Quijano stated that “likely” has a meaning of “probability” not just “possibility” and that to him “probable” means “[a]t least the probability should be 50 percent plus one.” See Tex. Health & Safety Code Ann. § 841.003(a)(2).
The jury did hear Quijano testify that in addition to the statutory definition of behavioral abnormality contained in section 841.002 of the Health and Safety Code, he used that part of section 841.003(a)(2) which states a sexually violent predator “suffers from a behavioral abnormality that makes the person likely to engage in a predatory act of sexual violence” in formulating his opinion. The jury also heard Quijano testify that “[tjhere’s a very small difference except the word ‘likely’ is present in 003.”
On appeal, Day argues that as a forensic psychologist Quijano was qualified to testify about legal questions related to mental health. We review the exclusion of expert testimony for abuse of discretion. K-Mart Corp. v. Honeycutt, 24 S.W.3d 357, 360 (Tex.2000). “A trial court abuses its discretion when its ruling is arbitrary, unreasonable or without reference to any guiding rules or legal principles.” Id. Day argues that the trial court failed to follow the guideline established by the Court of Criminal Appeals. See Morris v. State, 301 S.W.3d 281, 287 n. 14 (Tex. Crim.App.2009) (“We are not asked to decide whether these medical experts’ opinions concerning the legal standards by which to measure competency under Texas law were admissible under Tex.R. Evid. 702-705. They may certainly give their opinions concerning the specific facts and factors set out in [the statute governing competency to stand trial], but the legal definition and standard by which competency is measured is a question of law, not defined by medical experts.”).
The trial court cautioned that it would instruct the jury on the law applicable to the case. The court’s ruling was intended to prevent the witness from confusing the jury with the witness’s interpretation of a question of law. In addition to excluding Quijano’s construction of the legislative intent behind the use of the word “likely” or “probable” in the statutory definition, the trial court also excluded the psychologist’s testimony regarding the level of probability that he considered *219would be necessary to establish that Day is predisposed to commit a sexually violent offense. Quijano’s statement was in the context of his discussion of passages in Al-maguer and Martinez. He explained that the standard for forensic psychology encourages practitioners to consider all statutes and laws when rendering an opinion. Quijano’s comments about probability were not clearly offered as facts or data upon which he was offering his opinion as a forensic psychologist. See Tex.R. Evid. 703, 704. Moreover, Quijano was able to explain to the jury that he evaluated Day with reference to both section 841.002, which defines “behavioral abnormality” as a condition that predisposes an individual, and section 841.003(a)(2), which refers to a behavior abnormality that makes a person likely to engage in a predatory act of sexual violence. Under the circumstances, the trial court’s ruling was not an unreasonable application of the rules of evidence governing the admissibility of expert opinions. See Tex.R. Evid. 702-705. We overrule issue five and affirm the trial court’s judgment.
AFFIRMED.