MEMORANDUM OPINION
This matter comes before the Court on Charles J. Moore’s petition pursuant to 28 U.S.C. § 2255 to vacate, set aside, or correct his sentence. Moore contends: (1) that his conspiracy conviction was not supported by the evidence presented at trial; (2) that his court-appointed lawyer at trial provided ineffective assistance in fading to object to the prosecutor’s remarks labeling Moore and his brother as “liars” during closing argument; and (3) that his court-appointed lawyer on appeal provided ineffective assistance by fading to argue that Moore’s brother’s gudty plea estopped the government from attacking the brother’s trial testimony exculpating Moore from involvement in the conspiracy. For the reasons expressed here, petitioner’s contentions fad.
I
On August 11,1992, a grand jury returned a two-count indictment against Moore and his brother. Count I of the indictment charged both defendants with conspiracy to possess with intent to distribute more than fifty grams of “crack” cocaine, in violation of 21 U.S.C. §§ 841(a)(1) and 846. Count II charged Moore with using and carrying a firearm during and in relation to a drug trafficking offense, in violation of 18 U.S.C. § 924(e). The eases were severed, and whde Moore’s case went to trial, his brother entered a gudty plea. On October 22, 1992, the jury found Moore gudty as to Count I and not gudty as to Count II. With respect to Count I, Moore was sentenced on January 22, 1993, to 121 months of imprisonment, five years of supervised release, and a $50 special assessment. He then promptly filed and pursued a direct appeal, which was ultimately unsuccessful. See United States v. Moore, 11 F.3d 475 (4th Cir.1993), cert. denied, — U.S. -, 114 S.Ct. 1864, 128 L.Ed.2d 486 (1994) (affirming Civil Action No. 92-0035-01-AM) (hereinafter referred to as “Moore I”).
II.
Moore first argues that the government faded to adduce sufficient evidence to establish his participation in a conspiracy to distribute crack cocaine. He made this same argument on direct appeal, where it was rejected by a unanimous Fourth Circuit panel, which held that the government’s evidence of Moore’s gudt met both the sufficiency of evidence standard set forth in Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 2789, 61 L.Ed.2d 560 (1979), as wed as the essential elements of conspiracy standard established under United States v. Giunta, 925 F.2d 758, 764 (4th Cir.1991). See Moore I at 480. Because this claim of error is an issue previously raised and rejected on direct appeal, it must fad here, for it is wed settled that a criminal defendant is forbidden “to recast, under the guise of a collateral attack, questions fidly considered” on direct review. Boeckenhaupt v. United States, 537 F.2d 1182, 1183 (4th Cir.1976).
III.
Moore’s second contention is that his court-appointed lawyer provided constitutionady defective assistance at trial when the lawyer faded to object to the prosecutor’s statement in closing argument that Moore and his brother had “lied” on the witness stand regarding Moore’s involvement in the conspiracy. To merit relief for ineffective assistance of counsel, Moore must first establish that his “counsel’s representation fed below an objective standard of reasonableness.” Strickland v. Washington, 466 U.S. 668, 688, 104 S.Ct. 2052, 2064, 80 L.Ed.2d 674 (1984). Second, Moore must also “show that there is a reasonable probability that, but for *727counsel’s unprofessional errors, the result of the proceeding would have been different.” Id. at 694, 104 S.Ct. at 2068.1 Moore can make neither showing here.
Because the trigger for an ineffective assistance claim is an unprofessional error or mistake by counsel, the Strickland analysis of the claim properly begins with the identification of the error or mistake. Here Moore points to his counsel’s failure to object to a remark made in the government’s closing statement. Specifically, the prosecutor said that it was a “tragic case” in which the defendant “takes the oath on the [B]ible, and lies____” Moore I at 481. The prosecutor, speaking of the defendant’s brother who testified, then said to the jury “what the government knows and what you ladies and gentlemen know is that [Kirk Moore] lied when he took the stand.” Id.
Although the line separating permissible from prohibited argument is not always clear, there is little doubt that these remarks fell on the foul side of the line.2 According to the Fourth Circuit panel, they were “improper and unnecessary” and served to “skirt the precipice of reversible error.” Moore I at 481. But merely because these remarks were objectionable and presented defense counsel with an opportunity to object does not end the analysis. Constitutionally effective assistance does not require the assertion of every possible valid objection. Many objections, while not lacking in technical merit, are either not worth making or may reasonably be viewed as likely to cause more harm if made than if foregone. In these circumstances, an objection may have no effect beyond calling further attention to a damaging conflict in testimony and the inevitable conclusion that someone has lied. Courts therefore sensibly recognize that trial counsel must be granted latitude in making tactical and strategic decisions of this sort. Jones v. Barnes, 463 U.S. 745, 103 S.Ct. 3308, 77 L.Ed.2d 987 (1983); Bunch v. Thompson, 949 F.2d 1354, 1364 (4th Cir.1991) (stating that in habeas corpus petitions, the court should “credit plausible strategic judgments” to trial counsel).
Inge v. Procunier, 758 F.2d 1010 (4th Cir.1985) is particularly instructive in this regard. There, a unanimous panel held that a defense counsel’s failure to object to shotguns brought into the courtroom, but not admitted into evidence, did not amount to ineffective assistance. Id. at 1016. The panel reasoned that “whether the defense attorneys simply didn’t think to object or deliberately did not object ... is immaterial.” Id. Similarly, in the instant case, it does not matter that counsel either forgot to object or decided not to object to the government’s remarks in closing. In either event, the record plainly reflected the conflicting testimony of Moore’s brother and the police officers, from which, as the Fourth Circuit noted, “the jury obviously recognized someone was lying.” Moore I at 482. Indeed, all experienced practitioners recognize that not infrequently, it is better to remain silent than to draw attention to a matter by offering an objection. Courts, too, recognize counsel’s autonomy in strategic trial behavior. See Strickland, 466 U.S. at 689, 104 S.Ct. at 2065 (stating that it would be improper to “interfere with the constitutionally protected independence of counsel and restrict the wide latitude counsel must have in making tactical decisions”). Consequently, the silence of Moore’s counsel during the government’s closing argument may be viewed as a reasonable trial tactic. As such, counsel’s failure to object in Moore’s trial falls within a standard of objective reasonableness, and survives the first prong of Strickland.
Moore’s claim also fails Strickland’s second prong: Moore cannot show that his counsel’s failure to object at trial directly *728affected the trial outcome or lessened his chances of success on direct appeal. Because Moore’s counsel did not raise the objection at trial, Moore was forced to bear the burden on appeal of proving that the government’s statement constituted “plain error,” ie. that there was error, that it was plain, and that the error affected the defendant’s substantial rights. See United States v. Olano, 507 U.S. 725, 732-36, 113 S.Ct. 1770, 1776-79, 123 L.Ed.2d 508 (1993).3 The failure to object, even assuming it amounted in context to ineffective assistance, was harmless in Moore I beyond a reasonable doubt. There, in disposing of the appeal, the panel recognized the impropriety of the remarks, but then went on to conclude that in the context of the whole ease, Moore’s substantial rights were not affected and “the remarks did not affect the outcome of the trial.” Moore I at 481. Consequently, in light of the fact that the failure to object was not in itself ineffective assistance, and that Moore’s rights were not affected by the burden of proving plain error, Moore’s second claim also fails.
IV.
At this point, a brief digression, obiter dicta, is in order. There is, it appears, more than a little confusion in this district (and perhaps elsewhere) over the propriety of a prosecutor’s use of the words “liar” and “lying” in closing argument. Some have been heard to speculate that Fourth Circuit precedent banishes these emotive words from the prosecutor’s lexicon. A careful review of Fourth and other circuit authority on this point suggests this conclusion is too severe. The words themselves, while they have significant emotive potential, seem not to be the crux of the Fourth Circuit’s concern. Rather, the line between permissible and impermissible use seems to have a contextual basis.
At least four published Fourth Circuit opinions have expressed concern over a prosecutor using forms of the term “liar” in closing argument, the most recent of these occurring in the instant case. See Moore I. In 1983, a panel found objectionable a prosecutor’s statement that a defense witness had “insulted” the jury by “walking] in here with a straight face [and telling] the most incredible lies.” United States v. Moore, 710 F.2d 157, 159 (4th Cir.1983) (hereinafter referred to as “Moore II ”). The panel stated that in making this remark, “the prosecutor strayed close to, if not beyond, the outer limits of proper argument.” Id. Even so, the panel ruled it was not reversible error. The second occasion arose four years later, when a Fourth Circuit panel chastised a prosecutor for calling the defense witnesses “liars.” United States v. Cooper, 827 F.2d 991, 995 (4th Cir.1987). While observing that the prosecutor had engaged in “an improper excess of advocacy,” the panel went on to note that the remark “was most certainly not reversible error” because it “was provoked by defense counsel” and was “well-founded.” Id. Finally, a third case arose when a prosecutor referred to the defendant as a liar, thief, and crook who could not be believed, and told the jury that defendant was “a disease on society [a]nd you are the cure.” United States v. Rogers, 853 F.2d 249, 251 (4th Cir.1988). While the panel referred to the remarks as “inflammatory” and “excessive and uncalled for,” the remarks were “harmless overkill, not plain error.” Id. at 253.
Carefully read, these cases suggest that the Fourth Circuit’s chief concern over the use of the word “liar” and its variants in closing statements is not attributable solely to the terms’ potentially emotive effects, but is more likely a function of the context in which they are used. And this is as under*729standable as it is sensible, for many commonly used words have at least as much emotive effect as the word “liar.” Consider, for example, such terms as “thief,” “murderer,” “killer,” or “rapist,” all of which are no less offensive or emotive than “liar,” but none of which are, or should be, banned per se from a prosecutor’s lexicon. Of course, “liar” or indeed any emotive term may, if used excessively and intemperately, amount to improper argument. See Darden v. Wainwright, 477 U.S. 168, 181, 106 S.Ct. 2464, 2471-72, 91 L.Ed.2d 144 (1986) (quoting Donnelly v. DeChristoforo, 416 U.S. 637, 641, 94 S.Ct. 1868, 1870-71, 40 L.Ed.2d 431 (1974) (stating that closing arguments could become “so infected with unfairness as to make the resulting conviction a denial of due process”)). But in sum, despite pointed expressions of concern, it is significant that no Fourth Circuit panel has ever found use of the word “liar” in closing arguments to be reversible error.4
So, if it is not the emotive effect of the terms “liar” and “lying,” what then lies at the heart of courts’ concern over a prosecutor’s use of these terms? The answer, it seems, is the way in which the words are used. The line between permissible and impermissible use of the terms is drawn by reference to whether they are used in a context intended either to convey a prosecutor’s personal opinion or to suggest the government’s possession of extra-judicial information regarding the credibility of a defendant or witness. This conclusion is borne out by the Fourth Circuit’s two central objections to the practice in Moore II. First, the panel pointed to the danger that “because of his position, the jury may give weight to United States Attorney’s personal views____” Moore II at 159 (citations omitted). Second, the panel warned that a jury might “infer, from the form of counsel’s language, that he had access to extrajudicial information, not available to the jury.” Id. (citations omitted).
The panel’s first concern regarding the context in which prosecutors use the word “liar” stems from the long and sensibly settled rule that it is improper for a prosecutor, or indeed any lawyer, to assert what could be construed as her personal views of a witness’s credibility.5 Yet, this is, regrettably, an all too common occurrence, owing perhaps to the fact that common speech patterns in closing arguments are often inadvertently sprinkled with such phrases as “I think,” “I believe” and “I submit,” all of which, in context, may serve to convey a personal opinion on witness credibility. Perhaps because it is both common and inadvertent, the Fourth Circuit has charitably noted that using the phrase “I think” in closing argument does not necessarily “suggest an attempt to replace the evidence with the prosecutor’s personal judgments____” United States v. Adam, 70 F.3d 776, 780 (4th Cir.1995). Still, all counsel, prosecutors and defense attorneys alike, should avoid using these phrases, especially when commenting on witness credibility. It is the record evidence that should *730speak to witness credibility, not counsel’s personal opinion. Thus, counsel should use forms of expression that tell the jury that it is “the trial record,” not the counsel’s opinion, that “compels the conclusion,” “leaves no doubt” or “points persuasively to the fact” that witness X is lying, or is a liar.
But, as Moore II makes clear, a prosecutor must do more than simply avoid expression of a personal opinion; she must also avoid formulations that invite the impermissible inference that the government has extrajudicial information that casts a shadow on a witness’s credibility.6 Precisely this occurred in the instant case. The prosecutor’s statement in Moore I that “the government knows” the defendant lied on the witness stand was clearly impermissible, as using the phrase could easily be interpreted by jurors to mean that the government knew facts about the defendant’s testimony that were not presented at trial. Because the prosecutor called the defendant’s witness a liar in a manner that both expressed his personal opinion as to the witness’s credibility and suggested that the government had access to extra-judicial information, the Fourth Circuit panel could , reach no other conclusion but that the prosecutor had crossed the line and erred in making the statement.
In sum, there is no per se judicial bar against using the term “liar” or its variants in closing arguments. Nor should there be, given that most criminal trials and many civil trials involve contests between differing testimonial versions of past events where many of the versions reasonably appear to be deliberately false. When this occurs, the commonly used and well understood word “liar” is tailor-made to describe the situation. To preclude its use in such circumstances would be unnecessary, if not silly.7 Thus, where the record justifies the inference that a witness has lied, prosecutors do not cross the line into impermissible argument in calling the witness a “liar,” provided, of course, that the prosecutor does not suggest that this is a personal opinion or that it is based on extra-record government information.8
Y.
Finally, Moore’s third claim asserts that counsel on appeal was ineffective in that counsel failed to argue that Moore’s brother’s guilty plea estopped the government from attacking the brother’s trial testimony exculpating Moore from involvement in the conspiracy. The short answer to this claim is that no such rule of estoppel exists. In general, when the government enters into a plea agreement, it vouches only for the credibility and validity of those facts it offers as the factual basis for the defendant’s guilt. The government does not sponsor or vouch for every one of defendant’s assertions made regarding the crime. Moore’s brother pled to the indictment without an agreement, and while the Court found a factual basis for the plea, including the fact that Moore’s brother conspired with others, it was not required to identify the brother’s co-conspirators. See *731Rule 11, Fed.R.Crim.P. And significantly, the government’s acquiescence to the brother’s plea involved no concession or agreement concerning the brother’s credibility. In short, the plea in no way precluded the government from later attacking the brother’s credibility as it did, and successfully so, it appears.
In summary, there can be no elaim of ineffective assistance where, as here, counsel is alleged to have failed to raise a merit-less argument. Failure to raise a meritless argument can never amount to ineffective assistance.
Moore is therefore not entitled to reliéf and his petition must be denied.
An appropriate order will issue.