A jury convicted Jacqueline Lavertu of driving with a blood alcohol count over 0.08 (“DUI per se”), of driving under the influence of alcohol to the extent that it was less safe for her to drive (“DUI less safe”), and driving with an open alcoholic container. She argues on appeal that the trial court abused its discretion in denying her motion for a new trial on her conviction of DUI less safe because the verdict was strongly against the weight of the evidence, and that she was entitled to a new trial because her trial counsel’s performance was deficient in failing to question witnesses about the handling of her blood sample and failing to object that the State’s closing argument was prejudicial. For the reasons that follow, we affirm.
“On appeal from a criminal conviction, we view the evidence in the light most favorable to the verdict, and the defendant no longer enjoys a presumption of innocence.” Green v. State, 323 Ga. App. 832 (748 SE2d 479) (2013). So viewed, the evidence shows that a trooper with the Georgia State Patrol observed Lavertu cross the yellow centerline several times and stopped her to investigate whether she was impaired. Her pupils were slightly constricted and bloodshot, and her speech was slightly slurred. She initially got out of the car with no shoes on but returned to the car and put them on when the trooper asked if she wanted to do so. Her hand gestures were “sporadic” and nervous. The trooper attempted to conduct field sobriety tests, but Lavertu said an injury prevented her from performing the one-leg stand or walk and turn tests. She performed several counting and dexterity tests, and when asked to count backward from 75 by ones she “butchered it” several times, counting off seemingly random numbers that were not sequential. When asked to perform a finger dexterity test counting from one to four and four to one while touching her thumb to her fingers, Lavertu was unable to count backward. With her head back and eyes closed, Lavertu was unable to touch her nose. While attempting to perform the horizontal *710gaze nystagmus test, the trooper observed a nystagmus at maximum deviation, but Lavertu would not follow his directions to continue focusing on his finger, and so the nystagmus was not sustained sufficiently for him to document its existence. She declined to take a portable breath test to confirm or deny the presence of alcohol in her system. She was very nervous and “antsy,” moving in and out of the camera frame as equipment in the trooper’s car recorded the stop.
A second trooper who was qualified as an expert in alcohol and drug recognition testified that the first trooper asked him to come to the scene and give his opinion about whether Lavertu was under the influence of drugs or alcohol. He also testified that Lavertu would not follow his stimulus on the horizontal gaze nystagmus test, which tested for involuntary jerking of the eyes, nor would she follow his directions for a balance test. When she closed her eyes and leaned her head back with her hands at her side, she estimated that 30 seconds had passed when actually 56 seconds had passed. The expert testified that an estimate of time off by five seconds either way is an indication of impairment. Finally, Lavertu had body tremors and her eyes showed a lack of convergence when tested. In the expert opinion of the second trooper, Lavertu was under the influence of a central nervous system depressant or cannabis, to the extent she was a less safe driver.
While the first trooper thought that Lavertu was not “falling over drunk,” he also thought she was not sober and was a less safe driver than someone who was not under the influence of drugs or alcohol because her reaction time was suppressed. He arrested Lavertu for DUI and after placing her in the back of his patrol car, he and other officers searched the area of her car within reach of the driver’s seat. Inside a pocketbook on the floor on the passenger side, the trooper found a shoe containing a small empty vodka bottle, of the size used on airplanes. A police cadet on scene found one or two more similar empty vodka bottles.1 The trooper read the implied consent notice to Lavertu, and she agreed to give a blood sample, which was drawn at the police station.
A toxicologist with the Georgia Bureau of Investigation (“GBI”) Division of Forensic Sciences testified that upon receipt of evidence, the “evidence receiving technician” creates a file in the GBI database with information such as the subject’s name, the date of receipt, and the incident information. The evidence is assigned a unique identification number, and bar code stickers with that number are printed and applied to all the evidence and paperwork associated with the *711case. The technician then photographs the evidence and forwards it to the toxicologist. After testifying about the specific testing procedures used, the toxicologist testified that the sample labeled with Lavertu’s name contained 0.159 grams of alcohol, plus or minus 0.006 grams, per 100 milliliters of blood, which exceeded the per se blood alcohol content of 0.08. The effect of that level of alcohol on a person would be decreased inhibitions, increased self-confidence and sociability, slower information processing, possible mental confusion, memory lapses, difficulty perceiving sights and sounds, uncoordinated motor skills, poor balance, and slowed reaction to glare. Driving would be affected because it would take longer to process information, and in the toxicologist’s opinion, someone with a 0.159 blood alcohol content would be a less safe driver.
The jury convicted Lavertu of DUI per se (alcohol), DUI less safe (alcohol), and the open container violation, but acquitted her of DUI less safe (alcohol and drugs) and failure to maintain her lane.
1. Lavertu argues that the trial court abused its discretion by denying her motion for new trial on her DUI less safe conviction because the verdict was “decidedly and strongly against the weight of the evidence.” OCGA § 5-5-21 authorizes the trial court to “exercise a sound discretion in granting or refusing new trials in cases where the verdict may he decidedly and strongly against the weight of the evidence even though there may appear to be some slight evidence in favor of the finding.” This statute gives the trial court broad discretion to sit as a “thirteenth juror” and weigh the evidence on a motion for new trial alleging this general ground. See Ricketts v. Williams, 242 Ga. 303, 304 (248 SE2d 673) (1978). When asked to do so, “the law imposes upon the trial court an affirmative duty to exercise its discretion and weigh the evidence to determine whether a new trial is warranted.” Hartley v. State, 299 Ga. App. 534, 540 (3) (683 SE2d 109) (2009).
A trial court’s review of the evidence under OCGA § 5-5-21 differs from its review of the evidence in response to a motion for a directed verdict under OCGA § 17-9-1. In the latter case, the trial court has a duty to grant a directed verdict of acquittal “when there is no conflict in the evidence and it clearly demands a verdict of acquittal as a matter of law.” Merino v. State, 230 Ga. 604, 605 (1) (198 SE2d 311) (1973). A review under OCGA § 5-5-21 also differs from a determination that the evidence was not sufficient to allow a rational trier of fact to find a defendant guilty beyond a reasonable doubt of the offense charged. See Ricketts, 242 Ga. at 304. In that case, “the court is required to approach the evidence from a standpoint most favorable to the government, and to assume the truth of the evidence offered by the prosecution.” (Citation and punctuation omitted.) Id. *712(grant of new trial under predecessor to OCGA § 5-5-21 is not a finding that evidence was legally insufficient and retrial thus is not barred under the Double Jeopardy Clause of the United States Constitution).
Lavertu argues both that the trial court failed to exercise its discretion and weigh the evidence and that it abused its discretion by failing to find the verdict was strongly against the weight of the evidence, contending that “[n]o rational trier of fact could find under the facts in this case, that [she] was guilty beyond [a] reasonable doubt of driving to the extent that she was less safe to drive.” She argues that the following evidence establishes that the trial court erred in denying her motion for new trial based on OCGA § 5-5-21: the officers on scene were unable to discern any definitive evidence of intoxication from field sobriety tests; four officers spent nearly an hour before deciding there was probable cause to arrest her; she consented to a blood test; and her acquittal of the offense of failure to maintain her lane of traffic indicates that the jury believed she was not driving unsafely.
In this case the trial court clearly exercised its discretion under OCGA § 5-5-21 to determine whether the verdict was against the great weight of the evidence. In its order denying Lavertu’s motion for new trial, the court found specifically that her conviction for DUI less safe was “not against the greater weight of evidence.” After thoroughly reviewing the evidence and Lavertu’s arguments, the trial court found that
[m]aking a conclusory allegation that the officer’s testimony is just “some slight evidence” and the remaining evidence weighs “heavily” against the verdict is not enough to overturn the Jury verdict in the instant case. The Defendant has made no showing that the Jury verdicts are against the weight of available evidence. Thus this Court finds no basis to overturn the Jury verdict on the DUI Less Safe [charge].
Likewise, we find no basis for reversing the trial court’s exercise of discretion in denying Lavertu’s motion for a new trial under OCGA § 5-5-21.
2. Lavertu argues that her trial counsel was ineffective for failing “to question the witnesses regarding the handling of the blood sample taken from [her] on the date of her arrest” and for failing to object during the State’s closing argument. Ineffective assistance is a deficient performance by counsel resulting in actual prejudice. Strickland v. Washington, 466 U. S. 668 (104 SCt 2052, 80 LE2d 674) (1984); Dillard v. State, 323 Ga. App. 333, 336 (2) (753 SE2d 772) (2013). “If *713an appellant fails to meet his or her burden of proving either prong of the Strickland test, the reviewing court does not have to examine the other prong.” Wright v. State, 291 Ga. 869, 870 (2) (734 SE2d 876) (2012). “In reviewing the trial court’s decision, [the appellate courts] accept the trial court’s factual findings and credibility determinations unless clearly erroneous, but... independently apply the legal principles to the facts.” (Citation and punctuation omitted.) Id.
(a) Lavertu argues that, although five days passed between the night the blood sample was taken and the day it was received by the lab, defense counsel “made no attempt to discern what happened to the blood” during this five-day window, such as what temperature it was kept at, whether it was handled properly or who possessed it. Because the only theory of defense was that the blood sample must have belonged to someone else, Lavertu argues, failing to raise questions about the test’s reliability constitutes deficient performance.
Lavertu concedes that the State met its burden of proving sufficient chain of custody evidence. She argues instead that her trial counsel should have challenged the authenticity or reliability of the sample itself, because while the trial court decides the legal issue of admissibility, the jury decides whether the evidence is believable. In rejecting this argument, the trial court noted in its order denying Lavertu’s motion for new trial that Lavertu’s trial counsel questioned the State’s witnesses about the temperature of the blood and whether it was handled properly. Citing Lawrence v. State, 274 Ga. 794, 795 (3) (560 SE2d 17) (2002), the court concluded that trial counsel’s decision “to delve into one particular aspect of the evidence more than another is a matter of exercising professional judgment by counsel” and does not show deficient performance. As we have emphasized, “the degree to which an attorney chooses to cross-examine witnesses and the manner in which to attack their credibility fall within the ambit of trial tactics.” (Citation and punctuation omitted.) Id.
Further, Lavertu has not even speculated about what her trial counsel could have asked the witnesses that would have caused the jury to question the test’s reliability, much less proffered evidence about their responses. “Absent a proffer, defendant cannot meet his burden of making an affirmative showing that specifically demonstrates how counsel’s failure would have affected the outcome of his case.” (Citation and punctuation omitted.) Hortman v. State, 293 Ga. App. 803, 807 (e) (670 SE2d 99) (2008) (defendant cannot establish that trial would have ended differently without proffering evidence at the motion for new trial hearing that would have been revealed if counsel had subpoenaed telephone records). As Lavertu established *714neither her trial counsel’s deficiency nor any resulting prejudice, reversal on this ground is not warranted.
(b) Lavertu contends that her trial counsel was ineffective for failing to object when the State personally vouched for a witness’s credibility during closing argument. During her cross-examination of the witness who drew her blood sample at the police station, Lavertu’s trial counsel brought out that the witness also served as the chaplain for Douglas County law enforcement officers. During her closing argument, defense counsel said about this witness:
[S]he is the lady as you’ll recall who owns the private blood taking company that works for Douglas County and 18 other counties. So not only does she do that, but she’s also the chaplain for police officers and inmates here in Douglas County. So she helps put away the people she then later goes and counsels. I’ll let you think about that one.
In response, the prosecutor argued that the jury had to look at the evidence and put it all together, and that he found it “offensive” that defense counsel said it was the “troopers [’] fault” and “the crime lab’s fault” for not doing what they should have done. He continued, “It’s the blood drawer’s fault because she’s a chaplain and she works with inmates and law enforcement.” After discussing the specific charges, the prosecutor said that the per se charge that it was illegal to drive with more than 0.08 grams of alcohol was “pretty straightforward. All right. Unless you get into some conspiracy theories as to where her blood came from and how it got there.” Later, discussing the testimony of the witness about the process of drawing Lavertu’s blood, the prosecutor said:
She said there was nothing to note about it. I’m sure there wasn’t. . . . I’m not sure what’s to be made of the fact that she’s a chaplain and works with inmates. And she works with the sheriff’s office. I’m not really sure what the big deal about that is other than the fact that she’s a good person. I don’t think too many chaplains would be making up evidence. You know, putting someone else’s blood in there. I don’t think she was doing that.
The prosecutor went on to discuss the chain of custody evidence and the toxicologist’s testimony, then noted that Lavertu’s defense seemed to be that the blood tested belonged to someone else. After reviewing *715more of the evidence, the prosecutor argued that the blood test results were consistent with the field sobriety tests and that “[n] o one’s trying to trick anybody here.”
Decided February 7, 2014.
Edward V. C. Silverbach, Katherine L. Dodd, for appellant.
Matthew C. Krull, Solicitor-General, David Holmes, Assistant Solicitor-General, for appellee.
At the hearing on Lavertu’s motion for new trial, her trial counsel testified that unless a prosecutor says something “really egregious,” she tended not to object because prosecutors “have a lot of latitude.” She also considered whether getting up and interrupting someone during closing argument to make an objection that might not be important would make her lose credibility with the jury. In hindsight, the prosecutor’s remarks might have been objectionable, but at the time she thought these remarks were not significant in light of the evidence regarding Lavertu’s blood alcohol level.
“It is improper for counsel to state to the jury counsel’s personal belief as to the veracity of a witness; however, it is not improper for counsel to urge the jury to draw such a conclusion from the evidence.” Metts v. State, 270 Ga. 481, 484 (4) (511 SE2d 508) (1999). The trial court found that, looking at these comments in context, the prosecutor’s closing argument in this case was a permissible response to Lavertu’s attempt to create a question of bias because the witness worked as a chaplain for the county sheriff’s department. The court concluded that the failure of Lavertu’s trial counsel to object to these comments was not ineffective assistance.
Pretermitting whether the prosecutor’s argument was an impermissible statement of his personal belief, “deficient performance, alone, does not amount to ineffective counsel. [Lavertu] must establish that counsel’s inaction was so prejudicial to [her] defense that, but for the deficiency, there was a reasonable probability that the outcome of the trial would have been different.” (Citation omitted.) Lloyd v. State, 280 Ga. 187, 192 (2) (d) (ii) (625 SE2d 771) (2006). “In light of the overwhelming evidence of [Lavertu’s] guilt, [she] has failed to demonstrate that but for the handful of comments in the prosecution’s closing argument discussed above, there is a reasonable probability that the outcome of the trial would have been more favorable to [her].” (Citation and punctuation omitted.) Reed v. State, 291 Ga. 10, 17 (4) (b) (727 SE2d 112) (2012).
Judgment affirmed.
Miller and Ray, JJ., concur.