dissenting.
I respectfully dissent. I disagree with that portion of the majority opinion analyzing issues three, four and five under “EXTRANEOUS OFFENSES.”
Saxer complains it was error for the trial court to permit Lisa Garcia to testify that appellant had told her almost one week before the murder that if he, Saxer, found out who helped his wife move out, he would kill them; and error for the trial court to permit Patrick Sparks to testify that appellant ingested methamphetamine on the morning of June 1, 2000. Ms. Garcia’s testimony of Saxer’s statement upon learning his wife had moved out of the apartment was never tied to any evidence that Saxer later learned, prior to the murder, that Ms. Thorson, among others, helped his wife move out. The State proffered this evidence as admissible under Rule 404(b) as “motive.” The trial court opined Garcia’s testimony of Saxer’s statement was also admissible, in the face of any hearsay complaint, as “a statement by a party opponent.” See Tex.R. Evid. 801(e)(2)(A). The trial court also performed a Rule 403 analysis finding any prejudicial effect of the testimony “does not outweigh the probative value[.]” The trial court recognized Saxer’s timely objection under Rule 404(b) and Rule 403 and granted him a running objection on the testimony in question.
An objection based upon a violation of Rule 404(b) demands a relevancy analysis. Rankin v. State, 974 S.W.2d 707, 709 (Tex.Crim.App.l996)(opinion on original submission). Evidence is relevant that has “any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Tex.R. Evid. 401. However, as this court noted in Fleming v. State, 987 S.W.2d 912, 921 (Tex.App.-Beaumont 1999), pet. dism’d, improvidently granted, 21 S.W.3d 275 (Tex.Crim.App.2000), relevance is not an inherent characteristic of any item of evidence but exists as a relation between an item of evidence and a matter properly provable in the case. This involves, among other things, making sure that the proffered extraneous bad act/offense evidence has logical relevance to the non-character conformity purpose for which it is offered. See Powell v. State, 63 S.W.3d 435, 438 (Tex.Crim.App.2001)(quoting from Montgomery v. State in observing that extraneous act evidence “has noncharacter conformity relevance where it logically serves to make less probable defensive evidence that undermines an elemental fact.”)
In Rankin, the Court observed that when we say evidence is “relevant,” in terms of Tex.R. Evid. 4023, we are saying *783that the evidence makes a “fact of consequence” in the case more or less likely. Id. at 709. As noted by the Rankin Court, Montgomery explained the relevancy dynamic in the following way:
[A] party may introduce such evidence where it logically serves “to make ... more probable or less probable” an elemental fact; where it serves “to make ... more probable or less probable” an evidentiary fact that inferentially leads to an elemental fact; or where it serves “to make ... more probable or less probable” defensive evidence that undermines an elemental fact.
Rankin, 974 S.W.2d at 710 (quoting Montgomery, 810 S.W.2d at 887). Rankin further warns reviewing courts considering the relevancy of evidence to beware of the following pitfall:
An evidentiary fact that stands wholly unconnected to an elemental fact, however, is not a “fact of consequence.” A court that articulates the relevancy of evidence to an evidentiary fact but does not, in any way, draw the inference to an elemental fact has not completed the necessary relevancy inquiry because it has not shown how the evidence makes a “fact of consequence” in the case more or less likely.
Rankin, 974 S.W.2d at 710.
In the instant case, the State’s proffer of Saxer’s statement to Garcia that he would kill whoever helped his wife move out, at the time it was offered into evidence, had only conditional relevancy with regard to motive. Por the condition to be removed, the State would also have to prove directly, circumstantially, or inferentially, that Saxer eventually became aware that Ms. Thorson did indeed help his wife move out. In the analogous evidentiary circumstance noted above, a substantial line of case-law holds that evidence of a wife’s extra-marital relationship is not admissible to show motive to kill unless it is also shown that the defendant-husband was aware of the relationship. See Jernigan v. State, 585 S.W.2d 701, 704 (Tex.Crim.App.l979)(and cases cited therein). Therefore, at the time the State proffered the testimony in question, the State had not provided the evidentiary link that would show Saxer’s statement “made it more probable” that Saxer had a motive for murdering Ms. Thorson and that the motive was because he knew, or even suspected, that she helped his wife move out of the apartment. Because the record is silent as to this small but vital link tying Saxer’s statement to subsequent knowledge on his part that Ms. Thorson had helped his wife move out, the statement was not relevant to prove an evidentiary fact that inferentially leads to an elemental fact, the identifying of Saxer as the murderer. I believe the trial court abused its discretion in admitting this testimony of Ms. Garcia.
With regard to Issue Five, prior to Sparks’s testimony being presented to the jury, both parties and the trial court discussed its admissibility. The State proffered the testimony as “circumstances of the crime” and to show Saxer’s “state of mind” under Tex.Code CRiM. PROC. Ann. art. 88.36 (Vernon Supp.2003). The bench conference discussed issues implicating Tex.R. Evid. 401, 404(b), and 403. As noted above, the trial court permitted the State to elicit the testimony in question from Sparks but provided a limiting instruction to the jury. In making its ruling, the trial court specifically found the testimony to be relevant, admissible under Rule 404(b) as going to Saxer’s “state of mind,” and that its probative value was not substantially outweighed by the danger of unfair prejudice.
Article 38.36 of the Texas Code of Criminal Procedure is entitled, “Evidence In Prosecutions for Murder.” See Tex. *784Code CRiM. PRoc. Ann. art. 38.36 (Vernon Supp.2003). Article 38.36(a) reads as follows:
In all prosecutions for murder, the state or the defendant shall be permitted to offer testimony as to all relevant facts and circumstances surrounding the killing and the previous relationship existing between the accused and the deceased, together with all relevant facts and circumstances going to show the condition of the mind of the accused at the time of the offense.
In Smith v. State, 5 S.W.3d 673, 679 (Tex.Crim.App.1999), the Court of Criminal Appeals held that evidence admissible under article 38.36(a) may be excluded under Tex.R. Evid. 404(b) and 403. Specifically, the Court stated that if a defendant makes a timely 404(b) or 403 objection, before a trial court can properly admit the evidence under article 38.36(a), it must first find the non-character conformity purpose for which it is proffered is relevant to a material issue. Id. If the proffered evidence is relevant to a material issue, the trial court must then determine whether the evidence should nevertheless be excluded because its probative value is substantially outweighed by the factors in Rule 403. Id.
In the instant case, the State proffered the methamphetamine ingestion testimony for the purpose of showing Saxer’s “mental state” or “state of mind.” The trial court immediately responded by stating: “So, you’re offering it as it being relative or probative [of] the fact that he was high or under the influence of some drug?” The State replied, ‘Tes, Your Honor.” After further discussion of admissibility of the testimony the trial court made the following comment:
Well, yeah, I hear what you are saying. I guess — I assume it’s relevant. You know, it’s certainly probative on what was in his body and how his mind may be working. I don’t know. It’s also prejudicial, and all that is sort of a weighing thing really, I guess, is what it boils down to, is how — -what is the probative value as opposed to the prejudicial impact of it.
It appears both the State and the trial court were under the impression that testimony that appellant ingested methamphetamine in what appears, from other testimony, to be two to three hours before the murder, was relevant to Saxer’s “state of mind.” I have carefully examined the entire record and find no attempt by the State to show how the ingestion of methamphetamine works on the mind of a human being, much less on Saxer’s mind. The trial court suggested that such testimony would be probative to show Saxer was “high or under the influence” of the drug.
Nevertheless, the record does not contain a shred of evidence as to how methamphetamine affects individuals in general or if there is a commonly recognizable symptom or symptoms among individuals who ingest the drug. While it may be logical to infer that ingestion of a contraband substance results in some alteration of the mental or physical state of the person in question, it is impossible to make the inferential leap required by the State in the instant case that Saxer’s ingestion of methamphetamine resulted in a mental state prone to committing acts of intense violence. In terms of the analysis discussed in Rankin, supra, there was a failure by the State, as well as the trial court, to draw any inference between Saxer’s methamphetamine ingestion and an elemental fact in the case. As “common scheme or plan” was not an elemental fact of the charged offense of aggravated sexual assault in Rankin, so too in the instant case “state of mind” is not an elemental fact of the charge of murder. As prof*785fered by the State, Saxer’s methamphet-araine ingestion, standing alone, was not relevant to any “fact of consequence” involved in the instant prosecution. This is so because, like the erroneously admitted testimony of Ms. Garcia described above, Saxer’s methamphetamine ingestion was not an elemental fact that inferentially led to the ultimate elemental fact, the identifying of Saxer as the murderer. Testimony or evidence described in article 38.36(a), including the “condition of the mind” of the defendant, is admissible with the explicit proviso that it first be relevant. See Tex. Code CRim. PROC. Ann. art. 38.36. I would hold, therefore, that the trial court abused its discretion in admitting the testimony of Saxer’s ingestion of the methamphetamine on the morning of the murder. See Rankin, 974 S.W.2d at 719 (opinion on rehearing).
There being trial error, there must now be a “harm” analysis. See Cain v. State, 947 S.W.2d 262, 264 (Tex.Crim.App.1997). In Johnson v. State, 43 S.W.3d 1, 5 (Tex.Crim.App.2001), the Court held that it is the responsibility of the appellate court to assess harm after reviewing the record and that the burden to demonstrate whether the appellant was harmed by a trial court error does not rest on the appellant or the State. In the instant case, the appropriate standard for reviewing for “harm” is to disregard the error unless a substantial right has been affected. Tex. R.App. P. 44.2(b). A substantial right is affected when the error has a substantial and injurious effect or influence in determining the jury’s verdict. King v. State, 953 S.W.2d 266, 271 (Tex.Crim.App.1997). The following language taken from Kotteakos v. United States, 328 U.S. 750, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946), has been adopted by reviewing courts in Texas as authoritative when conducting a substantial rights review under Rule 44.2(b):
If, when all is said and done, the conviction is sure that the error did not influence the jury, or had but very slight effect, the verdict and the judgment should stand, except perhaps where the departure is from a constitutional norm or a specific command of Congress. But if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error, it is impossible to conclude that the substantial rights were not affected. The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence. If so, or if one is left in grave doubt, the conviction cannot stand.
Id. at 764-65, 66 S.Ct. 1239. (citation and footnote omitted). See also Johnson, 43 S.W.3d at 4; Morales v. State, 32 S.W.3d 862, 867 (Tex.Crim.App.2000); Jackson v. State, 17 S.W.3d 664, 672 (Tex.Crim.App.2000); Dorsey v. State, 24 S.W.3d 921, 930 (Tex.App.-Beaumont 2000, no pet.).
In a somewhat recent opinion, the Court of Criminal Appeals re-emphasized the necessity for a reviewing court to examine “the record as a whole” when conducting a harm analysis under Rule 44.2(b). See Schutz v. State, 63 S.W.3d 442, 444 (Tex.Crim.App.2001). As the Court noted, the reviewing court is to consider testimony, physical evidence, the nature of the evidence supporting the verdict, the character of the error and its relationship to other evidence, the trial court’s instructions to the jury, the theories of the respective cases for the State and the defense, arguments to the jury, and relevant portions of the voir dire. Id. at 444 — 45.
While the Schütz, Morales, Johnson, King, and Kotteakos cases discuss stan*786dards by which reviewing courts are to measure harm stemming from non-constitutional trial error, we find useful the following language contained in O’Neal v. McAninch, 513 U.S. 432, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995):
But we consider here the legal rule that governs the special circumstances in which record review leaves the conscientious judge in grave doubt about the likely effect of an error on the jury’s verdict. (By “grave doubt” we mean that, in the judge’s mind, the matter is so evenly balanced that he feels himself in virtual equipoise as to the harmlessness of the error.) We conclude that the uncertain judge should treat the error, not as if it were harmless, but as if it affected the verdict (i.e., as if it had a “substantial and injurious effect or influence in determining the jury’s verdict.”)
Id. at 435,115 S.Ct. 992 4.
In conducting a proper “harm” analysis in the instant case, the most difficult steps will be to fully describe the “nature of the evidence supporting the verdict” as we carefully examine the entire record, and to fully explain the “character of the error and its relationship to other evidence.” Schutz, 63 S.W.3d at 444-45. The character of the error was to provide to the jury what amounted to two “motives” for Saxer to have killed Ms. Thorson in a case where the evidence of motive for the murder rested on the thinnest of thin reeds. Even after all of the evidence was in, the State’s theory of the case was still somewhat murky. The record indicates four specific “motives” advanced by the State: 1) the murder was “methamphetamine-induced”; 2) the murder was based upon Saxer’s wanting sex from Ms. Thorson; 3) Saxer murdered Ms. Thorson based upon his statement to Ms. Garcia that he would kill anyone who helped his wife move; and 4) Saxer murdered Ms. Thorson because she owed him money for marijuana. As I would hold that the testimony as to motives 1 and 3 was erroneously admitted, I will initially examine the strength of the evidence of the remaining two motives in our examination of the nature of the evidence supporting the verdict as well as the character of the error and how it related to other evidence.
Wanting Sex From The Victim.
The State relies on the following direct examination testimony by the State’s witness, John Sayers:
Q. [State] Prior to June the 1st, did the defendant ever mention Emily Thorson to you?
A. [Sayers] He did.
Q. And did he ever indicate whether or not he had a relationship with Emily Thorson?
A. Not a relationship; as a friend, yes, sir.
Q. Did he ever indicate to you whether or not he desired to have a sexual—
[Trial Counsel]: Objection. Leading.
THE COURT: Sustained. Rephrase.
Q. (BY [State]) Did he ever indicate whether or not he wanted to have a relationship, other than a friendship, with her?
A. Yes.
[Trial Counsel]: Objection. Leading.
THE COURT: Overruled.
[Sayers]: Yes, he did.
Q. (BY [State]) And what type of relationship or experience did he indicate he wanted to have with her?
A. An intimate one.
*787Q. He told you that?
A. Sitting around, drinking, yeah.
Q. And this was prior to June 1 st?
A. Yes, sir.
[State]: Pass the witness.
Earlier during Sayers’s direct examination testimony it was elicited from Sayers that at the time of the murder, he, Sayers, was on probation out of Walker County, Texas, for possession of a controlled substance, and that subsequent to June 1, 2000, the Walker County probation was revoked based upon having committed a new offense out of Montgomery County, Texas, which also involved possession of a controlled substance. Sayers further testified he pleaded guilty to the Montgomery County charge and was serving a “jail sentence” in the Walker County Jail. Although Sayers insisted that he had made no deal with the State in exchange for favorable testimony, it was established under cross-examination that both a state jail felony conviction for possession of a controlled substance and a felony conviction for the same offense were to be served in county jail facilities; that in spite of the two recent drug convictions as well as prior convictions for misdemeanor family assault and felony bail jumping, Sayers made trusty at the Walker County jail facility which permitted him to have personal contact visits from relatives and access to phones. Sayers had no explanation for this seeming generosity on the State’s part, nor did the State attempt to explain Sayers’s bit of good fortune.
Sayers’s credibility was no small matter to the defense as he was the key State’s witness in that his testimony was the most solid of any that placed the murder weapon in the hands of Saxer in the early morning hours on the day of the killing. As noted above, Sayers also provided the account of Saxer’s statement of his wish to be “intimate” with Ms. Thorson. As set out in the record, however, this particular testimony comes across as a bit of fantasizing undertaken during a round of drinks with the boys. It certainly does not take on the inference of an expression of demented lust, as the State would have the jury, and this Court, believe. The actual testimony was brief and otherwise innocuous in and of itself. Even taking it together with Spark’s testimony that Saxer admitted to having “F’d” the victim earlier in the day on June 1, along with Saxer’s cryptic “It’s messed up,” a motive for murder still does not immediately leap to mind. There is no other testimony or evidence from any source that there was any enmity between Saxer and Ms. Thor-son at the time of the murder.
The following question and response does appear in the record during the direct examination of Ms. Thorson’s fiance, Paul Morales: “Q.[State] Mr. Morales, did Emily ever express to you that she was afraid or scared of the defendant Norman Saxer? A.[Morales] Yes.” Yet, the question is so global — “ever express ” — that its probative value is somewhat limited as Morales had known Ms. Thorson for approximately six months prior to the murder. There is no other evidence from any source that Saxer was known for violence or had a temper or had ever threatened Ms. Thorson or Morales. Indeed, Morales testified that both he and Ms. Thorson “socialized” with the Saxers and were guests in the Saxer apartment on several occasions. In fact; it appears that Morales and Ms. Thorson socialized with the Saxers and with both Mr. Sayers and Mr. Sparks on more than one occasion. In short, after examining the entire record, the nature of this particular “motive” evidence is quite weak and almost non-probative of motive for murder on Saxer’s part.
Ms. Thorson Owed Saxer For Marijuana.
*788Like the “sex” motive, the testimony developing this motive was also very brief. It is reproduced from the record as follows: “Q.[State] Did you ever owe him [Saxer] money for drugs? A. [Morales] Yes. Q. Okay. What kind of drugs? A. Marijuana.” The record reflects that when the State argued this as a motive for murder by Saxer it would couple it with the argument that because Morales and the victim “shorted” Saxer some “drugs,” this was an additional motive for murder. This appears to be a mistaken reading of Morales’s testimony because the next questions to Morales provided the following responses:
Q. [State] And did you and Emily ever conversely provide the defendant any marijuana?
A. Yes.
Q. And prior to providing him that marijuana, did you take some of it out for you and Emily’s own personal use?
A. Yes. We split it.
Q. You shorted him?
A. No. We split it. It was when I got it, it was already short.
As Morales explains, the amount of marijuana he and the victim were to provide to Saxer was already short when Morales received it, and that he and the victim simply split the shorted amount with Sax-er. Again, this incident, like much of the other “motive,” testimony, appears to exist in a virtual vacuum as there is no further evidence from any source even implying Saxer was angry or upset because of this disappointing transaction. The probative nature of this evidence with regard to motive for murder on Saxer’s part is also very weak. This evidence appears to indicate that Morales, the victim, and Saxer were, at most, recreational drug users, and not in the “business” of dealing drugs. Indeed, among this group of social acquaintances, it appears that Patrick Sparks, by his own admission, was actually in the business of selling illegal narcotics.
I commend the State for the candid observation, contained in its brief under a discussion of the harmlessness of the admission of the extraneous bad act/offense testimony, that the “drug transactions” motive and the “helping his wife to move” motive for killing the victim was much weaker than the “sex” motive. Certainly having found the “helping his wife to move” motive irrelevant, I could not agree more with the State as such evidence would have no probative value at all with regard to motive to murder. In my view, however, the “drug transactions” motive is equally as weak, if not more so, than the “sex” motive. Again, in light of the entire record, the nature of the “drug transactions” motive evidence has virtually no probative value as proof of motive for Sax-er to murder the victim.
Nature Of Evidence Supporting The Verdict/ Character of Error Related to Other Evidence.
An examination of the “nature of the evidence supporting the verdict” and “character of the error related to the other record evidence” in a “harmless” error analysis requires a look at all the evidence admitted before the jury without any “light most favorable,” as there must be an attempt to ascertain “the likely effect of the error on the jury’s verdict.” O’Neal, 513 U.S. at 435, 115 S.Ct. 992, 130 L.Ed.2d at 951. The nature of the incriminating evidence against Saxer was essentially a “mixed-bag” of circumstances, most of which have two reasonable, but opposing, implications. The one solid and unshakable fact was that a spent bullet found near the body of the victim was eventually showed via ballistics tests to have come from the. 22 caliber handgun introduced as State’s Exhibit 12. This handgun was *789identified as having been in the possession of three individuals during the 24-hour period of June 1, 2000.
As noted above, Sayers testified he gave the handgun to Saxer in the early-morning hours of June 1, following Sayers’s “road rage” incident. From Saxer, the handgun was passed onto Sparks sometime in the afternoon of June 1, following the shooting trip to Huntsville. Since the State’s theory of the victim’s time of death was before noon on June 1, the combined testimony of Sayers and Sparks circumstantially placed the murder weapon in Saxer’s possession at the time of the murder. Yet both Sayers and Sparks had their credibility significantly impeached with criminal histories, and the fact that Sparks provided an incomplete written statement to the police when first interviewed, but later provided a second written statement somewhat more consistent with his trial testimony. It was also brought before the jury that Sayers and Garcia had been “lovers” while Sayers was apparently involved in a long-term relationship with Saxer’s ex-wife, Crystal Spikes.
Sparks’s testimony provided a good example of the “mixed-bag” nature of the evidence in the record. During Sparks’s direct examination testimony, he stated that he received a phone call from Saxer sometime between 10:30 a.m. and 11:20 a.m. telling Sparks not to come by Saxer’s apartment because Saxer had “a lady friend” coming over. However, during cross-examination, Sparks was asked about what he wrote in his second statement to the police, only days after the murder, in which he never mentions Sax-er was expecting a “lady,” but instead reads that Saxer “had someone coming over.” Sparks was not asked to explain the disparity between his trial testimony and his written statement. In any other trial, the distinction would probably be of no consequence. In this ease, however, any evidence from which incriminating circumstances could be inferred may have provided the final piece of the puzzle for the jury.
Another significant “mixed-bag” of testimony came from Lisa Garcia, the manager of the apartment complex where the murder occurred. Ms. Garcia was permitted to testify that six days before the murder Saxer’s wife moved out, and that Saxer remarked to Garcia, “[I]f he found out who helped her move out, he was going to kill them.” On cross-examination, Ms. Garcia was presented with the written statement she gave to the police on the day of the murder. The pertinent portion of this exchange reads as follows:
Q. [Trial Counsel] And when did you give this statement? Is there a date?
A. [Garcia] It was the 1st.
Q. June 1, the day that Emily was found dead. And in this statement you say, “On May 26, Emily helped Susan Saxer move out of her apartment leaving her husband, Norman Saxer. Norman Saxer told me — ” and you have your name in parenthesis, correct? “If he found out that there was a man in his apartment helping her move out, helping her — ” and then you have Susan Saxer in parenthesis— “or a man she was having an affair with, he would kill them.” Now, is that what you told the police on June 1st?
A. Yes, sir. I was very confused that day.
Q. But you’re not confused now. Ma’am? You’re not confused now?
A. No, sir.
It would seem that in Ms. Garcia’s written statement, the focus of Saxer’s anger was clearly on the adulterous relationship, if there was one, between his wife and the unknown male paramour. In Ms. Garcia’s *790direct examination testimony, the focus of Saxer’s anger was on any individual who aided his wife in moving. In Garcia’s written statement, no motive for Saxer’s killing of the victim can be inferred from his comment even if he had known the victim aided his wife in moving. In Garcia’s trial testimony, had Saxer become aware of the victim’s participation in the move, the statement would have been laden with motive. Again, small distinctions with huge repercussions appear because of the particular facts and circumstances in the record.
Much was made by the State that testimony from Sparks indicated Saxer was in a hurry to rid himself of the .22 caliber handgun and that when Sparks later learned of the murder and wanted to return the weapon, Saxer “got hostile” and told Sparks not to bring the handgun back. Yet, the entire reason that Saxer came into possession of the weapon in the first place was to help Sayers because Sayers did not want to be caught with it as it had been used by Sayers in his “minor altercation” with another vehicle at which time Sayers “show[ed]” the handgun to the other motorist. Later, Sayers took the weapon to Saxer’s apartment, told Saxer about the altercation, and sold the handgun back to Saxer. That does not end the story, however, as Sayers was stopped by police the following day, June 1, around 11:30 a.m. Sayers was told by the police that they were investigating the altercation of the night before. The police searched Sayers’s car and person. Soon after this encounter with the police, Sayers went to Saxer’s apartment and told him that he had just been pulled over by the police and that the police were looking for the handgun. Sayers testified that Saxer told him to “burn the original receipt” that Sayers had received when Sayers purchased the handgun from Saxer approximately two months before.
Placed in context with the events described in Sayers’s testimony regarding the murder weapon, the State’s position that Saxer was anxious to pass the gun to Sparks because he had used it in the murder is somewhat less compelling. Once Sayers told Saxer that the police had stopped him and were looking for the handgun, it was reasonable for Saxer to want to get rid of it himself. From Sayers’s point of view following his encounter with the police, the weapon was “hot.” Why should Saxer not have thought the same? Ridding himself of the gun by selling it to a third party was certainly reasonable under the circumstances. The incriminating aura the State places on Sparks’s testimony detailing the transfer of the handgun from Saxer is undoubtedly tarnished by the events described by Sayers regarding his experience with the police.
Another of the State’s incriminating circumstances centered on the testimony of Sparks with regard to Saxer throwing items of clothing out of the vehicle window as the two men drove up Interstate 45 toward Huntsville for their shooting practice. Sparks remembered the general description of the clothing. Later that day, Sparks was interviewed by the police and recounted this incident, according to the testimony of Detective Eddie Davis of the Conroe Police Department. Based upon this information from Sparks, several police officers conducted a search of a “7- or 8-mile stretch” of Interstate 45 in hopes of recovering these items of clothing. This search was conducted over two consecutive days. While the officers collected a number of items of clothing, none matched Sparks’s description of the clothes Saxer discarded. Undoubtedly, Saxer’s actions in discarding the clothing was somewhat suspicious. However, Sparks did not testify that he observed blood on the clothing *791nor do we have any testimony tying Saxer to having been seen in the same or similar clothing on the day of the murder. Again, on balance, all of the record evidence surrounding the discarding of the clothes is not strongly probative of Saxer’s complicity in the murder.
Finally, there is the entirety of the testimony of Melissa Lewis McNeely, the representative of Saxer’s church, who happened to be visiting another congregant in the apartment complex when she encountered Saxer. Ms. McNeely had a brief conversation with Saxer in which Saxer stated that he had just gotten home from a doctor’s appointment, and that if he had not been at the doctor’s he would have been in the state pen. Ms. McNeely further testified that until this encounter with Saxer she was unaware that he was employed as a correctional officer at the penitentiary. Once learning this, Ms. McNeely acknowledged that she took Saxer’s previous comment as a joke. From this, the State argues that Saxer’s comment was indicative of his guilt in that “he lied about his whereabouts.” Yet taking Saxer’s “lie” in the context of the entire record, it is only weakly indicative of an attempt to somehow cover his whereabouts or establish an alibi so as to escape suspicion for the murder. This is so because Saxer was openly seen or easily placed at the apartment complex for the entire morning and early afternoon hours of June 1. Mario Henson observed Saxer walking with the victim at about 11:35 a.m. Sparks had been with Saxer from around 8:00 a.m. to about 10:30 a.m. Sayers had contact with Saxer twice that morning — a phone call to tell Saxer of his encounter with the police and that the police were looking for the handgun, and the second time Sayers drove to Saxer’s apartment to get a bottle of whisky left at Saxer’s the night before. Sparks’s testimony then ties Saxer at the apartment complex from about 11:00 a.m. onward, virtually up to the time he and Saxer left for Huntsville to shoot the guns. All this is to say that the State’s use of Ms. McNeely’s testimony as proof that Saxer was trying to seriously provide an alibi for himself at the time the murder was supposed to have taken place is weak when taken in context with the entirety of the record evidence. Any serious probative value from Ms. McNeely’s testimony as to Saxer’s guilt is, again, thin at best.
When all of the record evidence is examined, it consists of what I have termed “mixed-bag” evidence which could have either an incriminating or innocent inference as to Saxer’s guilt depending upon how the factfinder cared to resolve credibility or other circumstances surrounding the testimony; and the evidence includes facts and circumstances the inferences from which are weakly probative of Saxer’s guilt.
I conclude by examining the jury arguments contained in the record. Significantly, the State begins and ends its argument by emphasizing the improperly admitted “methamphetamine ingestion” evidence. Specifically, the State begins its closing argument with the following two points:
You heard from Dr. Shrode, and Dr. Shrode told you about the results of his autopsy. And that autopsy tells you the story, the story of what this defendant, Norman Saxer, in his methamphetamine-induced state of mind did to Emily Thorson the late morning hours of June the 1st, 2000.
[[Image here]]
The evidence shows that the defendant, in his drug-induced state — you’ll recall the testimony of John Sayers, that the defendant wanted to have sex with her.
Following this brief presentation, trial counsel for Saxer made their closing re*792marks, and then the State closed with its final remarks to the jury, concluding its remarks as follows:
Where is the testimony that the defendant had motive? Well, we know that the defendant wanted to have sex with her. We know that the defendant — we know that the defendant had a relationship with her and her boyfriend, her fiancé previously which involved drugs and that they had shorted him and that they owed him money for drugs. And we also know that Emily, out of the goodness of her heart, helped the defendant’s wife move the preceding Friday and we know that the defendant was upset about this and we know that he threatened to kill people who had helped her move. Motive.
[[Image here]]
So, Pat [Sparks], yes, he was a drug dealer at the time. He was all too willing to provide it, and he came over and he and the defendant shared some methamphetamine.
[[Image here]]
You have the evidence before you to find the defendant guilty of this offense. You have the evidence showing that he had this gun in his possession, that he had the motive, that he had the opportunity, and that he had the mental state, that drug-induced mental state to intentionally and knowingly kill her, kill her on June the 1st in Montgomery County, Texas. And if Emily Thorson were here today, she could tell you what the defendant did to her, but she’s not.
Ladies and gentlemen, I ask you to return a verdict of guilty in this case. Thank you.
The record reflects the State emphasized the erroneously admitted drug ingestion evidence on three occasions; at the beginning and at the end of its closing remarks, and also emphasized the erroneously admitted “kill-anyone-who-helps-wife-move” evidence. Each time the State summarized the “motives,” it was essentially summarizing its theory of the prosecution. The core of Saxer’s defense appeared to be an attempt to undermine the credibility of the State’s key witnesses— Sayers, Sparks, and Garcia — and to raise the alternative theory that it was Sayers who was the murderer.5
At the conclusion of my exhaustive examination of the record in this case I am where the “conscientious judge” finds him/herself when he or she is in grave doubt about the likely effect of the trial court’s errors on the jury’s verdict. I certainly find myself with “grave doubt” as we find ourselves past “virtual equipoise.” See O’Neal, 513 U.S. at 435, 115 S.Ct. 992, 130 L.Ed.2d at 951. That two clearly irrelevant “motives” were presented to the jury and emphasized during final argument by the State, coupled with the tenuous nature of the remaining two “motives,” leads me to the belief that the errors could have had a substantial and injurious impact on the verdict. My “grave doubt” is reinforced by the fact that, because the key State’s witnesses had their credibility seriously undermined, the nature of their testimony in support of the verdict, being entirely circumstantial in the first place, had a dark shadow cast over it. While weighing credibility of the witnesses and the weight to give to their testimony is entirely within the factfinder’s prerogative, an appellate court is mandated to examine the nature of the incriminating evidence *793under a “harmless” error analysis. Obviously, if the incriminating nature of the evidence is strong and solid, the impact of the error on the jury may be slight. In this case, however, testimony is significantly weakened by credibility questions and is dominated by the relating of circumstances that at times requires strained inferences to reach an incriminating conclusion. I can only conclude that the errors did affect Saxer’s substantial rights. Points of error three, four and five should be sustained. Consequently, the judgment should be reversed and the cause remanded to the trial court. Because the majority is of the opposite opinion, I respectfully dissent.
. Texas Rule of Evidence 402 reads: "All relevant evidence is admissible, except as otherwise provided by constitution, by statute, by these rules or by other rules prescribed pursuant to statutory authority. Evidence which is not relevant is inadmissible.”
. Because Rule 44.2(b) is based upon Federal Rule of Criminal Procedure 52(a) we look to the Supreme Court for guidance in interpreting our "harmless error” rule. See Johnson, 43 S.W.3d at 4.
. Statements made at the bench, though not heard by the jury, are still part of the record if recorded. Tex.R.App. P. 13.2(b)(3). See Hayes v. State, 85 S.W.3d 809, 816 (Tex.Crim.App.2002).