OPINION OF THE COURT
Daniel P. FitzGerald, J.
The defendants move under CPL article 440 to vacate their conviction, after a jury trial, of two counts of sexual abuse in the first degree. This motion follows their discovery that a police officer who testified at trial, Craig McKernan, had been suspended from the force for 30 days and put on departmental probation for one year as a result of an off-duty incident that took place in 1991. The defendants claim that the prosecutor’s failure to turn over this information prior to trial prevented them from impeaching McKernan’s credibility at trial, and that they are therefore entitled to a new trial.
A threshold issue is whether a defendant’s boilerplate request for the discovery of prior disciplinary actions involving any police officer who participated in the investigation of the case triggers a duty for the prosecutor to investigate each officer’s confidential personnel file. If a prosecutor has such a duty, the decisive question then presented is whether, under the facts of this case, the details of the 1991 incident constituted Brady material. Because I find that such a general request does not impose a duty on a prosecutor to make a wholesale review of every officer’s confidential personnel file, and that, in any event, the prior bad acts of McKernan were not Brady material in this case, the defendants’ motion is denied.
Relevant Facts at Trial
The defendants were tried on charges of sodomy, attempted sodomy, sexual abuse and unlawful possession of two handguns. The main witness at trial was the 20-year-old victim who testified that the defendants sexually assaulted her in their hotel suite. She said that she first met the defendant Shakur, a noted actor and rap singer, several days before the assault at a nightclub called Nell’s here in New York City. The star-struck victim was attracted to Shakur, and their meeting culminated shortly thereafter with the two returning to Shakur’s hotel suite to have consensual sex. Later that week, the victim was telephoned and again invited to Shakur’s hotel. She accepted, and even bought a new dress for the occasion. When she arrived at the hotel, she met the defendant Fuller as well as two [963] other men who were watching television with Shakur. She also saw two guns in the living room of the suite which, at Shakur’s direction, Fuller moved to an adjoining bedroom. After having a drink and watching television for awhile, she and Shakur retired to the bedroom, lay on the bed, and began to kiss. Their activity had not gone very far when the door burst open and Fuller and the two other men entered. The victim testified that Shakur then held her down and said words to the effect that "what you do for me you have to do for them.” Thus began the sexual attack for which the defendants have been convicted.
After the attack ended and she was permitted to leave, the victim immediately called hotel security. Security Officer Gary Thomas testified that he responded to the call and found the victim on the floor near the elevator bank outside the defendants’ suite. Her clothes were disheveled, her stockings torn and she was sobbing and obviously distraught. When she told Thomas that she had been sexually assaulted, he took her to the hotel security office. Once there, she provided him with more details of the attack and also told him of the two guns she had seen in Shakur’s suite. Hotel security then telephoned the police.
McKernan’s Trial Testimony
Several officers responded to the call, including Craig Mc-Kernan. McKernan, one of 10 police officers who testified at trial, gave the jury an overview and general narrative of the police activity so that the jury had an understanding of the chronology of events that night. His testimony did not supply the critical details for any contested issues; that role was reserved for witnesses with more in-depth firsthand knowledge.
McKernan testified that when the police arrived, he and his fellow officers first talked to Gary Thomas. Thomas told them of the victim’s allegations and also that Shakur had recently been involved in a shooting with off-duty police officers in Atlanta, Georgia. McKernan and the other officers then interviewed the victim. McKernan testified to the victim’s appearance, thus corroborating Thomas’s description, and he further agreed with Thomas that the victim mentioned the two guns she had seen in the hotel suite. On hearing about the guns, McKernan and his Sergeant called for an Emergency Services Unit (ESU).
McKernan also provided other miscellaneous testimony. For instance, he testified that he saw another police officer arrest [964] Shakur, and that he, along with two other officers, initially detained Fuller in the hallway outside the hotel suite. He said that he was present when the victim had an opportunity to view Fuller before his arrest. McKernan was also present when the ESU team went into the hotel suite to look for suspects, and his testimony was consistent with the trial testimony of ESU Officer Daniel Masterson.1 Later that night, when the defendants were in the holding cell at the station house, Mc-Kernan heard them say that they should have taken the victim downstairs and put her in a limousine. Finally, McKernan vouchered some of the clothing that the defendants wore that evening.
The defendants’ cross-examination of McKernan was brief and uncontentious. The defendants brought out that McKernan and the other officers learned that evening that Shakur had allegedly shot an off-duty officer in Atlanta and that this information was communicated to the ESU team. McKernan also acknowledged that he did not observe all the activities of the ESU officers when they were in the hotel suite. This was relevant to the defendants’ claim, albeit unsupported by any evidence, that the police planted the guns in the hotel suite to retaliate for the Atlanta incident. The defendants also successfully used McKernan to elicit some allegedly inconsistent statements of the victim. Finally, counsel for Fuller elicited from McKernan that Fuller was at all times cooperative with the police.
Although on its face McKernan’s testimony appears to have added little to the case against the defendants other than to complete the narrative of events on the evening of their arrest, the defendants are nonetheless correct when they claim that they were unable to question him about any personal misconduct he committed in 1991. Thus, in order to evaluate any impact this may have had on the trial, it is necessary to briefly review the facts which gave rise to McKernan’s disciplinary sanction.
McKernan’s Prior Misconduct
At about 3:30 in the morning on November 27, 1991, Mc-Kernan tried to order food at the drive-through window of a [965] White Castle restaurant even though the restaurant was closed. McKernan, who was off duty at the time and had consumed "a few beers,” argued with two employees who were still at the restaurant. Although he then left, he returned a few minutes later and knocked over a garbage can with his car. As he was again leaving, one of the employees threw a bottle at McKernan’s retreating car; it landed either on his car or in the street. In response, McKernan made a U-turn and brandished his service weapon as he drove by the restaurant. The employees called the police, and although the police initially took McKernan into custody, the employees declined to press charges. The police department, however, filed internal disciplinary charges against McKernan.
The departmental proceedings brought against McKernan took almost two years to resolve. On November 23, 1993, following a departmental disciplinary hearing, the Police Commissioner approved the Hearing Commissioner’s recommendation that McKernan be suspended for 30 days and put on disciplinary probation for one year. McKernan subsequently challenged the sanction, but to no avail (see, McKernan v Kelly, 216 AD2d 17). Defendants first learned of the disciplinary sanction when the First Department’s decision was officially reported approximately seven months after the defendants’ trial. The prosecutor first learned of it when the defendants made the instant motion.2 Presumably, however, if the People had reviewed McKernan’s personnel file in preparation for the defendants’ trial, they would have been alerted to the facts of his misconduct. Therefore, one of the issues raised by the [966] defendants’ motion is whether the People had a duty to review that file before the trial.3
Prosecutor’s Duty to Investigate Confidential Files Both the courts and the Legislature have narrowly limited the ability of a defendant to discover the information contained in a police officer’s confidential personnel file (People v Gissendanner, 48 NY2d 543; Civil Rights Law § 50-a). In Gissendanner, the Court recognized the State’s legitimate interest in maintaining the confidentiality of the internal records relating to performance and discipline of its police officers. The Court noted that “the maintenance of police morale and the encouragement of both citizens and officers to co-operate fully without fear of reprisal or disclosure in internal investigations into misconduct” supported keeping such records confidential (48 NY2d, supra, at 548). And while there exists an inevitable tension between the right of an accused to confront adverse witnesses on the one hand and the legitimate interests of law enforcement in maintaining the confidentiality of a police officer’s personnel file on the other, the Court held that “it is not to be assumed that, in striking the balance between the two, police confidentiality must always yield to the demands of a defendant in a criminal case” (48 NY2d, supra, at 548). Thus, the Court held that before a defendant is even entitled to have a Judge sign an order authorizing the delivery of a confidential file to the court for in camera review, he must set forth in good faith a factual predicate which would provide some basis fo believe that the file will establish the "unreliability of either the criminal charge or of a witness upon whose testimony it depends” (at 550). The purpose of the factual predicate is to [967] insure that the defense request is not "merely a desperate grasping at a straw.” (People v Gissendanner, 48 NY2d 543, 550, supra.)
This rule is also supported by section 50-a of the Civil Rights Law. Although the Court in Gissendanner (supra) did not rely on section 50-a because the statute became effective after the trial of that case, it nonetheless stated in dicta that section 50-a codified the judicial decisions in this area by making it plain that a police officer’s personnel records must be kept confidential unless a party shows that their delivery to the court is justified. The statute formally provides that before a court may breach the confidentiality of these protected records, it must hold a hearing and find that the defendant has made a "clear showing of facts” warranting disclosure (Civil Rights Law § 50-a [2]); only then may the court order that the personnel files be produced to it under seal. And even then the court must not disclose their contents to the defense until it has first reviewed the records itself in camera and determined disclosure is warranted (Civil Rights Law § 50-a [2]).
In the present case, although Fuller’s demand for discovery4 asked for an in camera review of the personnel files of "any law enforcement agent who participated in the investigation of this case,” he neither named a particular officer or officers whom he believed warranted such a review, nor gave any reasons to believe that any disciplinary matters existed in those files. Indeed, it is clear he had no basis to believe that the personnel file of any police officer, including McKernan, contained helpful information. His boilerplate discovery request was therefore a fishing expedition and the "desperate grasping at a straw” plainly disfavored by the Court in Gissendanner (supra, 48 NY2d, at 550).
That seven months after the trial the defendants were able to set forth facts to support Fuller’s initial request does not change the analysis that under Gissendanner (supra) or section 50-a the defendants were not entitled to pretrial disclosure of the confidential files. It has long been understood that under Gissendanner and section 50-a it is a defendant’s obligation to make a timely prima facie showing in order to warrant court review (People v Valentine, 160 AD2d 325; People v Harris, 121 AD2d 788; People v Norman, 76 Misc 2d 644). The defendants, however, now seek to in effect change this rule and replace it [968] with a much broader pretrial discovery standard. In essence, they ask that the prosecutor routinely review the personnel files of all the police officers who participated in the investigation of the case, whether or not there is reason to believe that any of those files will contain information helpful to the defendant. The defendants contend in their moving papers that it is the prosecutor’s duty to first review the personnel file and then supply the defendants with information to support the factual predicate required under Gissendanner and section 50-a. In other words, the defense hopes to accomplish by indirection what Gissendanner and section 50-a prohibit the defense from doing directly. They argue that Brady mandates this rule. I disagree.
In Brady v Maryland (373 US 83), the Supreme Cpurt held that prosecution’s failure to turn over evidence favorable to an accused upon request violates due process "where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution” (373 US, at 87). This duty is not absolved where the prosecutor does not personally know about the favorable information so long as the information is known to those acting on the prosecutor’s behalf: "[negligent, as well as deliberate, nondisclosure may deny due process” (People v Simmons, 36 NY2d 126, 132; see also, Giglio v United States, 405 US 150, 154; People v Vasquez, 214 AD2d 93, 99, supra). The normal application of this rule can be seen in those cases which hold that a prosecutor has constructive knowledge of information the police have gathered during their investigation of the crime the defendant is accused of having committed (People v Wright, 86 NY2d 591; People v Vilardi, 76 NY2d 67; People v Clausell, 182 AD2d 132). It is simply no excuse that a police investigator failed to inform the prosecutor of information he has gathered in the case.
But as information in various other law enforcement files becomes more and more removed from the case on trial, and when it therefore becomes more speculative to think that any relevant information even exists, the duty of the prosecutor to investigate far flung police files must, of necessity, diminish. A prosecutor is not constructively aware of police files unrelated to the case on trial unless there exists some reason to believe a file contains relevant information. Otherwise, a conscientious prosecutor would have to search every police department file, whether related to the case on trial or not, in order to be certain of fulfilling his Brady obligation. That is clearly not required (see, e.g., United States v Joseph, 996 F2d 36).
[969] It begs the question, though, whether it is appropriate to deem a prosecutor to be constructively aware of what is maintained in administrative police files that both the courts and Legislature have deemed confidential. In other areas where the courts have had an opportunity to discuss the role of the prosecutor in disclosing protected files, the courts have uniformly held that a prosecutor has no duty to disclose such material. For instance, in the Rosario context, it has long been recognized that the prosecutor is under no duty to turn over prior statements that for public policy reasons are deemed confidential (People v Rosario, 9 NY2d 286, 289; People v Ranghelle, 69 NY2d 56, 59; People v Tissois, 72 NY2d 75, 78). Further, in the analogous context of CPL 240.45, a prosecutor is not required to disclose any record of conviction of a prosecution witness unless the "record * * * of conviction * * * is known by the prosecutor to exist.” (CPL 240.45 [1] [b].) The prosecutor need not affirmatively search for the prior convictions of his witnesses. To the contrary, the statute specifically provides that a prosecutor is not required "to fingerprint a witness or otherwise cause the division of criminal justice services or other law enforcement agency or court to issue a report concerning a witness” (CPL 240.45 [1] [c] [emphasis added]). If a witness has a prior conviction unknown to the prosecutor, but which is in fact reflected in some unrelated police or prosecution file, the defendants here would argue that a prosecutor could rely on CPL 240.45 only at his peril. Yet, this would be wrongly equating "access” with "constructive possession” (see, e.g, People v Washington, 86 NY2d 189), and would be unwarranted in light of the plain legislative directive that prosecutors are not required to investigate the criminal history of their witnesses. It would effectively compel the prosecutor to routinely search for prior convictions of all their witnesses and would thus eviscerate CPL 240.45. It would thus improperly allow a defendant to do an end run around the statute merely by couching a boilerplate discovery request in the language of Brady5
[970] In a similar vein, the Legislature has provided, that the personnel records of police officers are exempted from general discovery. The Legislature allows the production of these files to a Judge only in the limited case where a defendant can make a legitimate good-faith offer of proof that specific records contain information helpful to the defense. Permitting a routine request to trigger a corresponding search of these files by a prosecutor, then, would be contrary to the clear legislative declaration that the road to discovery of a police officer’s personnel file in a criminal case must go through section 50-a.* *6 Indeed, indiscriminate searches of these files would turn section 50-a on its head by shifting to the prosecutor the duty to set forth the defendant’s "clear showing” that the personnel files will provide relevant and material information. In the future, every defendant who lacks a good-faith basis to believe any helpful information exists in a personnel file would simply make a bare bones request that the prosecutor ferret out any helpful information by reviewing all testifying officers’ confidential files. Thereafter, if something helpful to the defendant fortuitously surfaces, the onus would shift to the prosecutor to ask the court for in camera review.
To adopt a rule that permits roving searches of confidential files would also be a significant change in New York’s discovery practice. In this era of word processors, voluminous and generic omnibus motions often contain a whole host of boilerplate provisions regardless of their specific relevance to the case. This type of motion practice has become common, as the voluminous demands in this case well shows. If the prosecutor is required to investigate all administrative and confidential files upon a mere request, however unsupported, then every officer who makes an arrest will have his personnel file reviewed by the prosecutor and perhaps the court. If every negative entry in these files will routinely be disclosed, it could have a chilling affect on internal investigations by the police department.
[971] It will also place a drain, not only on prosecutorial resources, but on court resources as well. Well over 100,000 criminal prosecutions were commenced last year in New York County alone. Each of these cases involved at least one if not several police officers. Indeed, the instant case involved the testimony of 10 officers and the involvement of many more. All the police personnel files would have to be routinely examined by the prosecutor during the course of every case. Further, prosecutors will likely be reluctant to withhold even slightly negative comments contained in the file for fear at some later point it may spawn a Brady complaint. Yet, in order to insure that the confidential nature of these files is not lightly breached, the prosecutor would not be permitted to disclose any information from the file without a court order (Civil Rights Law § 50-a). Thus, prosecutors will make routine applications to the court for in camera review of any information in the files, even where it relates only to insignificant areas affecting, at best, general credibility. The court would then become the routine repository for police officers’ personnel files during the normal discovery phase of every criminal case where a police officer has any allegation of wrongdoing or minor blemish on his record.
I agree that prosecutors should be encouraged to reveal evidence even marginally helpful to the defendant. But to institutionalize the wholesale review of personnel files in all cases without any initial basis to believe the files contain Brady material is not only unwarranted, but would allow the defense to achieve by indirection what it could not do directly under either Gissendanner (48 NY2d 543, supra) or Civil Rights Law § 50-a. I agree with the position of one court, which aptly noted, "the personnel records of a police officer should not be examined by a Magistrate, prosecutor or defense counsel each time an arrest is made” (People v Lugo, 93 Misc 2d 195, 199; see also, People v Norman, 76 Misc 2d 644, 651, supra). It is not required by Brady, and flies in the face of both Gissendanner and section 50-a. I therefore hold that a prosecutor is not required to investigate the statutorily protected personnel files at the mere whim of the defendant.7
[972] The Files as Brady Material
In any event, even if I were to assume that the failure to disclose McKernan’s disciplinary record is attributable to the People, their failure did not rise to the level of a Brady violation. Here, contrary to the defendants’ lengthy and tortured suggestions about the use they could have made of McKernan’s three-year-old misconduct, the defendants’ inability to use this information clearly had no impact on the outcome of the trial.
The United States Constitution is not violated every time the prosecutor fails to disclose evidence favorable to the defense (United States v Bagley, 473 US 667, 675; see also, People v Vilardi, 76 NY2d 67, supra). Brady requires the disclosure of only relevant and material exculpatory evidence. And in deciding whether favorable information is "material”, the information must be evaluated in the context of the defendant’s request. If the defendant has made only a general request for favorable information, he must show that the undisclosed exculpatory material " 'create[d] a reasonable doubt that did not otherwise exist’ ” (People v Vilardi, 76 NY2d 67, 73, supra, citing United States v Agurs, 427 US 97, 112). If he has made a specific request narrowly tailored to identify the relevant information he seeks, he must show that there is a "reasonable possibility” that the failure to disclose the information contributed to the verdict (People v Vilardi, 76 NY2d 67, 77; supra; cf. and contrast, United States v Bagley, 473 US 667, supra [adopting a single "reasonable probability” standard in Federal cases]).
Here, the defendant and the prosecutor clash over whether the defendant’s boilerplate request should be considered a general or specific request. Although the defendant is correct that his discovery paragraph specifically requests all information regarding disciplinary actions against all police officers, the People’s rejoinder is well taken. The defendant did not limit his request to a specific officer or even a group of officers critical to the case. Instead he made a general, open-ended request. It included at least the 10 officers who testified as well as any others who "participated in the investigation.” This is the type of request that, although nominally "specific,” is actually a general request to search a "sprawling mass of records” on the [973] speculative hope that something helpful will be uncovered; there reaches a point where wholesale "specific” requests become meaningless and are the equivalent of a general request for exculpatory material (see, e.g., United States v Joseph, 996 F2d 36, 41, n 5, supra). Additionally, it is not unreasonable for a prosecutor to assume chat a blunderbuss request for nondiscoverable information in confidential files would not qualify as a specific request because it fails to alert a prosecutor to a reasonable belief that helpful information will be found there. That the prosecutor indeed held this position is plain in light of her original response, which asserted that Fuller was not entitled to the discovery of the files under Gissendanner (supra) or Civil Rights Law § 50-a because he had not provided the requisite factual predicate to believe the files contained exculpatory information. The People thus have a colorable claim. In the end, though, I need not decide what standard should be applied in this case because I find that even under the more stringent standard of "reasonable possibility,” McKernan’s disciplinary matter would have had no impact on the verdict.8
It is true that when the reliability of a given witness may well be determinative of guilt or innocence, nondisclosure of important evidence affecting credibility falls within the rule of Brady (Giglio v United States, 405 US 150, 154, supra, citing Napue v Illinois, 360 US 264, 269; see also, People v Savvides, 1 NY2d 554). This principle applies, however, where the impeachment material is directly probative of a critical witness’s veracity about matters on which he testified at the trial (see, e.g., Giglio v United States, 405 US 150, supra [witness offered a benefit for his testimony against defendant]; People v Steadman, 82 NY2d 1 [same]; People v Cwikla, 46 NY2d 434 [same]; People v Wallert, 98 AD2d 47 [complainant filed civil suit against defendant for his role in crime]; People v Wright, 86 NY2d 591, supra [victim’s status as police informant provided motive for police officers to corroborate her testimony]). Evidence which merely impeaches a witness’s general [974] credibility as to a collateral matter, on the other hand, is rarely material. And the unexcused failure to disclose such evidence to the defendant does not provide a basis for a new trial (People v Battee, 122 AD2d 526; see also, People v Richards, 184 AD2d 222; People v Irvin, 180 AD2d 753; People v Ramos, 147 Misc 2d 672). Here, by no stretch was McKernan’s off-duty incident for brandishing a weapon directly probative of any issue at the defendants’ trial.9
Moreover, even as to general credibility, while prior bad or immoral acts may affect credibility, not all such acts are equally probative. For instance, acts of impulsive violence are unlike those involving theft or dishonesty, and will seldom have any logical bearing on a person’s general credibility, especially where remote in time (People v Sandoval, 34 NY2d 371, 376-377; People v Scott, 118 AD2d 881). Here, McKernan’s personal misconduct in wrongfully brandishing a weapon as he drove by two civilians three years before the defendants’ trial is precisely the type of behavior involving such an act of remote and impulsive violence. This prior bad act thus had little, if any, probative value on the issue of McKernan’s general credibility. Moreover, it was wholly collateral to any issue relevant to the defendants’ trial, and the defendants would therefore have been bound by any answer McKernan gave to their questions (Prince, Richardson on Evidence § 6-305 [Farrell 11th ed]). Indeed, because it was so collateral it would have even been well within my discretion to curtail all inquiry into it in the first place (People v Sorge, 301 NY 198).10
[975] Finally, the subject matter of McKernan’s trial testimony was hardly of the sort where his reliability "may well be determinative of guilt or innocence.” He gave little, if any, noncumulative testimony relating to the sexual abuse charges, and defendants’ arguments to the contrary are unpersuasive. McKernan’s testimony regarding the victim’s physical appearance and her initial report that she saw two guns in the suite was wholly cumulative to the more detailed testimony of Gary Thomas, as well as Detectives Slimak and Lewis, who both interviewed the victim at length. His testimony regarding the defendants’ arrests was narrative background information. It was unimportant because identification was not a contested issue, nor could the defendants have hoped to make it one given the victim’s prior relationship with Shakur and her knowledge that Fuller was Shakur’s business manager. To the extent that the defendants would have hoped to use McKernan’s prior misconduct to cast a general cloud over the police in an effort to somehow show they planted the weapons in the suite, it is sufficient to note that the defendants were acquitted of all gun charges. Similarly, although McKernan helped voucher the defendants’ clothing for later forensic testing, this evidence, hardly critical to the prosecution, related solely to the sodomy counts for which the defendants were also acquitted.
In this case there was simply nothing in McKernan’s testimony that made him significant to the People’s case against the defendants or a focus of the defendants’ attack on the victim. In the end, despite his role as an "arresting officer,” McKernan was a peripheral witness to the People’s case. The defendants’ stated intent to have somehow used the three-year-old event of this marginal officer in the same fashion "Johnnie Cochran did in connection with Mark Furhmann” in the O.J. Simpson case borders on the bizarre. Hyperbole is rarely an effective form of persuasion. Nor is there any constitutional right to courtroom theatrics. This trial focused on the victim’s story: the defense was that the victim made up the account of a sexual assault because she was scorned by Shakur. Any attempt to impeach McKernan with an unrelated 1991 incident of misconduct, minimally probative even on general credibility, [976] simply would not have contributed to the jury’s fact-finding on any relevant, contested issue. Thus, I find there is absolutely no possibility that the defendants were harmed by the nondisclosure of McKernan’s misconduct.
Conclusion
For the foregoing reasons, the defendants’ motion to vacate the judgment is denied. To the extent that the defendant Fuller has raised any other arguments not specifically dealt with here, I have considered them and believe them to have no merit.