OPINION OF THE COURT

We are concerned with the constitutional propriety of the investigative stop of defendant’s automobile on a public highway. The pivotal issue is whether the police possessed reasonable suspicion of criminal activity to justify this interference with defendant’s travel and privacy. This case also raises questions regarding the role that race played in the decision to detain defendant. Following the denial of[*617] defendant’s motion to suppress the evidence which was obtained as a result of that seizure, he pleaded guilty to robbery in the first degree and petit larceny.

I

The facts developed at the suppression hearing are undisputed. On April 28, 1982, Sergeant James O’Connor of the Town of Greece Police Department read a newspaper article concerning two convenience store robberies which had occurred in the Towns of Gates and Chili the preceding night. Before going on duty, he was advised by the Gates police that the robberies occurred after 11:00 p.m. within one-half hour of each other, that in both cases the suspect was the same “black male” and that, although no vehicle was seen in the area of the Gates robbery, a black male in a large, dark colored vehicle was seen in the vicinity of the Chili robbery prior to its commission. This information, however, did not “connect the car or the male with the [Chili] robbery”. Thereupon at their 10:00 p.m. roll call, the sergeant told his officers of the pattern of the robberies and asked them to pay special attention to the convenience stores in their area.

At approximately 1:19 a.m. the next morning, while he was operating a police car in the vicinity of Ridgemont Plaza in the Town of Greece, the sergeant received a police radio call “of a robbery at the Stop and Go, Dewey-Latta by a Black male.”^[1] No further description of the robber was broadcast, nor was a vehicle described; in fact, the dispatch did not indicate whether the robber used a getaway vehicle. The sergeant then radioed his police dispatcher to advise the City of Rochester and the Town of Irondequoit Police Departments of the robbery and to dispatch cars to “shut off * * * the main arteries of escape” from the crime scene. He then drove east on Ridge Road to Lake Avenue, a distance of about four miles, and proceeded north on Lake Avenue “trying to figure out the time element of distance from my traveling and the * * * possible suspect getaway route, how long it would take for him or at what point we would possibly meet.” He saw “three or four [cars] around Lake and Ridge” but let them pass because he “didn’t feel[*618] that the suspect could have possibly got that far.” When he reached Hansford Lane, which is the entrance to Kodak Park on Lake Avenue and four or five miles from the crime scene, he slowed down in order to observe the vehicles traveling southbound on Lake Avenue. He directed the spotlight of his police car on the three or four cars which passed him, but none of them “aroused” his suspicion. At approximately 1:25 a.m., however, while he was proceeding northbound in the vicinity of the Holy Sepulchre Cemetery on Lake Avenue, he observed “a large vehicle with a Black male” heading southbound on Lake Avenue. When the sergeant slowed his vehicle down to a stop and directed his spotlight on this individual, later identified as defendant, the defendant looked “startled like” and had a “shocked look” which “through experience” “seemed odd” to the sergeant. The defendant then slowed his car down and the sergeant made a U-turn and followed the car for approximately a mile.^[2] No traffic violations or infractions were observed and a license plate check provéd negative. The sergeant then received an additional radio report that the “suspect was possibly wearing grey colored clothing” and decided to stop the car on the theory that its driver “might possibly be the person [who] robbed [the] store”.^[3] The vehicle was stopped without incident. The sergeant approached it “with caution”^[4] and requested the defendant to produce his license and registration. The sergeant then observed a brown paper bag with a $5 bill protruding from it on the front floor of the vehicle and some Miller beer on the back seat. Because of his “experience” as a police officer investigating robberies, he became “alarmed” and ordered defendant out of the car, frisked him, handcuffed him and then placed him in the back of a police vehicle. Defendant was advised at that time that he was not being placed under arrest but that he was being handcuffed for his safety and the safety of the sergeant. The sergeant then radioed investigators at the crime scene and described the items foiind in the car. The investigators confirmed that[*619] the suspect had placed the money taken in the robbery in a bag and that he had also taken some Miller beer. The sergeant was instructed to hold the defendant until the victim could be transported to the scene. When the victim arrived “a few minutes later”, defendant was removed from the police car and, presumably while handcuffed, “positively” identified by the victim. Defendant was then formally placed under arrest. Approximately 15 minutes elapsed from the time that the vehicle was stopped until the victim arrived at the scene and identified the defendant.

Defendant contends (1) that he was unlawfully stopped by the police without reasonable suspicion and that any tangible evidence seized as well as the showup identification should be suppressed as “fruit of the poisonous tree”, (2) that he was placed under arrest without probable cause when he was handcuffed and placed in the police car, if not before, and (3) that the showup identification was unnecessarily suggestive and that the People failed to establish by clear and convincing evidence that there is an independent basis for in-court identification testimony by the robbery victim.

The hearing court ruled against defendant on each of these points. It said that the stop was authorized under the circumstances, since it “was not undertaken with an intent to harass”, nor based on “mere whim * * * caprice or idle curiosity”, but “was done under exigent circumstances * * * created by the fact that a crime had very recently been committed and an articulable suspicion that the person seized might be engaged in criminal activity by virtue of the area in which he was seen and stopped.” In this regard, the court accorded “due weight” to the reasonable inferences drawn from the facts by Sergeant O’Connor in light of his experience. Furthermore, the handcuffing and securing of defendant was found to be reasonable under the circumstances, although “probably a maximum intrusion.” Defendant’s claim that he was thereby arrested was specifically rejected by the court which found that the arrest did not take place until after he was identified by the victim. Finally, the court found that the victim had an ample and adequate opportunity to observe the robber during the[*620] commission of the crime and that the showup was not impermissibly suggestive. It specifically ruled “that the fact that the suspect may have been in handcuffs and was standing alone at the time of his identification [does not] make [the showup] an impermissible one under the circumstances”.

The People argue that this decision is correct in all respects and that the sergeant’s actions were reasonably tailored to meet the exigencies of this case with minimal intrusion of defendant’s rights. Although we disagree in part with the hearing court’s decision, for the reasons which follow we affirm.

II

In cases such as this, the initial inquiry is always whether the police action was justified in its inception (People v De Bour, 40 NY2d 210, 215). Resolution of this question here requires an understanding of what is sufficient to authorize police to stop a motor vehicle traveling on a public highway and the role that race may play in that decision.

“I saw the car in the area and wasn’t answering any complaints, so I decided to pull them off.” This sentiment, as a justification for an investigative stop of an automobile and its driver by the police, is violative of Fourth Amendment protections (Delaware v Prouse, 440 US 648; People v Ingle, 36 NY2d 413, 420). An investigative stop of an automobile on a public roadway is a seizure within the meaning of constitutional limitations (People v Ingle, supra, p 418) since it interferes with the travel and the privacy of the car’s occupants, and it must be based on “reasonable suspicion of criminal activity” (People v Singleton, 41 NY2d 402, 405; People v Brooks, 88 AD2d 451, 454; see, also, United States v Cortez, 449 US 411, 417; People v Harrison, 57 NY2d 470, 476).

What is “reasonable suspicion”? The United States Supreme Court has said that “the essence of all that has been written is that the totality of the circumstances — the whole picture — must be taken into account. Based upon that whole picture the detaining officers must have a particularized and objective basis for suspecting the particular person stopped of criminal activity.” (United States v [*621] Cortez, 449 US 411, 417-418, supra.) As such, reasonable suspicion is a standard which is less than the probable cause required to arrest or search (United States v Brignoni-Ponce, 422 US 873, 881).

Our Court of Appeals has said that “[reasonable suspicion is the quantum of knowledge sufficient to induce an ordinarily prudent and cautious man under the circumstances to believe criminal activity is at hand * * * To justify such an intrusion, the police officer must indicate specific and articulable facts which, along with any logical deductions, reasonably prompted that intrusion. Vague or unparticularized hunches will not suffice” (People v Cantor, 36 NY2d 106, 112-113; see People v Singleton, 41 NY2d 402, 404, supra; see, also, CPL 140.50, subd 1). What is prohibited are random, arbitrary stops which are “the product of mere whim, caprice, or idle curiosity.” (People v Ingle, 36 NY2d 413, 420, supra; see, also, People v Singleton, supra, p 404.)

Whatever the formulation of the rule, however, it is clear that “reasonable suspicion” deals with probabilities and not with hard certainties. The police may draw inferences and make deductions from not only their own observations but also from, among other things, information obtained from police reports, objective facts and consideration of the type of crime committed and usual modus operandi. “Long before the law of probabilities was articulated as such, practical people formulated certain commonsense conclusions about human behavior; jurors as factfinders are permitted to do the same — and so are law enforcement officers * * * [T]he evidence thus collected must be seen and weighed not in terms of library analysis by scholars, but as understood by those versed in the field of law enforcement.” (United States v Cortez, 449 US 411, 418, supra; see, also, People v Valentine, 17 NY2d 128, 132.)

Race is often a factor which enters into the determination of reasonable suspicion, since witnesses and victims frequently describe criminal perpetrators by the color of their skin (see, generally, Johnson, Race and the Decision to Detain a Suspect, 93 Yale LJ 214). Where a suspect has been described by his race, it is “a characteristic which may properly be used as one element of identification”[*622] (Franklin v State, 374 So 2d 1151, 1154 [Fla]). It is “an identifying factor which * * * assists the police in narrowing the scope of their identification procedure.” (United States v Collins, 532 F2d 79, 82, cert den 429 US 836.)

The color of a person’s skin, however, cannot serve as the sole basis for suspicion. It has been held in this State, as in others, that ethnic identity alone is an insufficient basis upon which to premise reasonable suspicion (People v George T., 39 NY2d 1028, revg 48 AD2d 779; People v La Borde, 66 AD2d 803, 804; see, also, United States v Brignoni-Ponce, 422 US 873, 885-887, supra). A person cannot be stopped by the police solely because he is black, no more than he can be stopped because he is Mexican-American or Puerto Rican. A person’s racial status is neither an unusual circumstance nor probative of propensity to commit crime, and it does not provide a basis to conclude that crime is afoot. This has been held to be true where an individual of one race is observed in an area inhabited primarily by members of another {People v George T., supra), since freedom to travel and associate are fundamental rights, and the suggestion that their exercise can contribute to a lawful seizure of one’s person is both illogical and intolerable (see People v Bower, 24 Cal 3d 638). Race assumes importance in determining the existence of reasonable suspicion only when it is considered in conjunction with other facts which provide an articulable basis for suspicion.

This case implicates these principles — “especially the imperative of recognizing that, when used by trained law enforcement officers, objective facts, meaningless to the untrained, can be combined with permissible deductions from such facts to form a legitimate basis for suspicion of a particular person and for action on that suspicion.” (United States v Cortez, 449 US 411, 419, supra.) The police stop of defendant’s vehicle was, in our view, justified by the facts and was not based on a subjective hunch.

Critically, the crime and subsequent stop took place at an early morning hour when there was little motor vehicle traffic. This is recognized as a significant factor justifying a stop upon much less comprehensive information than[*623] would be adequate were the stop made at midday (3 La-Fave, Search and Seizure, § 9.3, p 91). Illustrative is Orricer v Erickson (471 F2d 1204) holding lawful the stopping of a car bearing out-of-county license plates within one hour of a 3:40 a.m. burglary where, despite the lack of any description of the burglars or their car, the area was so deserted that only two pedestrians and two cars had been seen in that interval (see, also, People v Brown, 88 Ill App 3d 514, cert den 451 US 1019).

Also of crucial importance in determining the propriety of the stop is its timing and location. The proof shows that Sergeant O’Connor, a police officer of 13 years, had stationed himself along one of the likely escape routes at an appropriate distance from the crime scene taking into consideration the time which had elapsed since the crime’s commission. His positioning significantly limited the possible suspects from simply any “black male” to only a black male whose presence was observed at that point and time on the highway (see Williams v State, 261 Ind 547 [where stop of car driven by a black male traveling on most direct escape route from robbery to Chicago was upheld on ground that police had reason to suspect those black males using highway at that time and place]; Franklin v State, 374 So 2d 1151, supra [where stop was held permissible since it occurred at a time and distance from robbery which was consistent with an attempt to flee and three passengers fit general description of the suspects]). Although the sergeant had no knowledge of a getaway vehicle or of the exact escape route taken by the robber, it was reasonable for him under the circumstances to infer that the black man who committed the crime would attempt to escape from the scene in a motor vehicle via one of the main highways leading southerly from the scene to the City of Rochester (see People v Grice, 87 Ill App 3d 718, cert den 450 US 1003 [holding that it was not unreasonable for police to infer that suspects would attempt to escape by car on the main highway leading from scene]). It would ignore reality to suggest that the modus operandi in crimes of this character does not ordinarily involve the use of a getaway car. In any event, the sergeant knew that the type of crime committed, i.e., an early morning robbery of a convenience[*624] store, fit the general pattern of two other recent convenience store robberies which had occurred within one-half hour of each other in the Towns of Chili and Gates, crimes which could not have been committed by the same person unless a motor vehicle was used for transportation. Furthermore, the sergeant was familiar with the area where the crime occurred, its proximity to the City of Rochester and the limited number of direct escape routes from the crime scene to the city. As a police officer he was permitted to make a judgment, based upon his experience, as to the route most likely to have been taken by the perpetrator (see 3 LaFave, Search and Seizure, § 9.3, p 92). It must be remembered in this regard that reasonable suspicion is concerned with probabilities and not hard certainties (United States v Cortez, 449 US 411, 418, supra). The mere possibility that the perpetrator of the crime may have taken another route or escaped on foot, did not require Sergeant O’Connor simply to shrug his shoulders and permit defendant’s car to proceed on its way unimpeded (see People v Finlayson, 76 AD2d 670, 677).

Sergeant O’Connor further narrowed his search for possible suspects to only black males driving “large, dark colored” cars, such as that driven by defendant. This, too, was eminently reasonable given the report of a suspicious “large, dark colored” car seen in the area of the Chili robbery prior to its commission. The sergeant was obviously operating under the assumption that the person responsible for the Gates and Chili holdups was also responsible for the present robbery as well. Whether this assumption was true or false is immaterial, since it was a reasonable one to make, given the pattern of the crimes which he knew had only recently been committed.

Finally, there is no doubt that the sergeant’s decision was influenced by the defendant’s reaction to the police encounter. The sergeant testified that his suspicions were aroused by defendant’s “startled” and “shocked” look when the searchlight shown on him and by the noticeable reduction in speed of defendant’s car when he knew that he was being followed by a police officer. Suspicious conduct may provide the additional basis needed to justify a stop where the police have only a general description of the person[*625] involved (see, generally, 3 LaFave, Search and Seizure, § 9.3, pp 92-93). Obviously, defendant’s conduct was not of such character as to justify a stop on that basis alone, and a strong argument can certainly be made that the conduct was “normal” under the circumstances and hardly indicative of any criminality. Nevertheless, the sergeant testified that the “look” which defendant gave was such, based upon his years of experience apprehending lawbreakers, as to arouse his suspicion.

We conclude, therefore, that the stop of defendant’s vehicle was based on reasonable suspicion and that the sergeant had a particularized and objective basis for suspecting that the defendant was the perpetrator of the robbery which had occurred only minutes before. Defendant’s race was only one of many factors which assisted the police in narrowing the scope of their investigation. The stop was the product of good police work by a sergeant who drew upon his years of experience to make the deductions necessary to quickly apprehend a robber, and mere whim or idle curiosity did not play a part in his decision-making process.

Ill

Having decided that the action of the police was justified in its inception, we must now consider whether the subsequent intrusion was reasonably limited in scopé and intensity within the constitutional structuring against unreasonable searches and seizures (US Const, 4th Amdt; NY Const, art I, § 12; see, also, People v De Bour, 40 NY2d 210, 221, supra). There can be no question that Sergeant O’Con-nor acted reasonably here in ordering defendant out of his motor vehicle, even if the sergeant had his gun drawn. When a car is lawfully stopped, the police may order its occupants to get out without conducting any preliminary inquiry (People v Lathigee, 84 AD2d 918; People v Rosario, 94 AD2d 329, 333). If the individual stopped is suspected of a serious or violent offense, as the defendant was here, then the officer’s resort to his weapon, standing alone, does not transform the confrontation into an arrest, since the drawing of a gun by a police officer under these circumstances is a permissible act of self-protection (People v Finlayson, 76 AD2d 670, 679, 680, supra; cf. People v [*626] Rosario, supra). The difficulty with this case is that Sergeant O’Connor went beyond merely ordering defendant from his car. He took the additional “protective measures” of frisking defendant, handcuffing him and placing him in a police car. Contrary to the conclusion of the hearing court, such an intrusion amounts to an arrest which must be supported by probable cause (People v Brnja, 50 NY2d 366, 372; see, also, People v Battaglia, 56 NY2d 558, revg 82 AD2d 389). The sergeant, however, did not have probable cause to arrest at that time and the restraint was effected without legal authorization. Although the bag with the money and the beer were in plain view in the car, the sergeant did not have probable cause to arrest defendant until the investigators at the crime scene confirmed that the robber had placed the money taken in the robbery in a bag and that he had also taken some Miller beer. Then, since it was more probable than not that defendant was responsible for the crime which had been committed, probable cause existed for his arrest (see People v Carrasquillo, 54 NY2d 248, 254; People v McRay, 51 NY2d 594, 602; People v Gordon, 87 AD2d 636, 637). Thus, the conclusion is inescapable that the defendant was subjected to an illegal arrest when he was handcuffed and placed in a police car, albeit only moments before the sergeant obtained sufficient information to give rise to probable cause.

The question then becomes what are the fruits of that illegal arrest and should they be suppressed. Clearly, the seizure of the bag of money and the beer is unrelated to the illegal arrest, since these items were discovered before defendant was handcuffed and placed in the police car (cf. People v Milaski, 62 NY2d 147). The only possible fruit of the illegal arrest could be the showup identification (see People v Gregory, 90 AD2d 506; People v Gordon, 87 AD2d 636, supra). On its facts, this case is similar to People v Johnson (75 AD2d 715, app dsmd 53 NY2d 839). There the defendant was illegally detained by the police without probable cause and questioned. Nevertheless, before he made any inculpatory statements police obtained a statement from a codefendant which clearly implicated the defendant and they then had the probable cause necessary for a legal arrest. We held that the confession was not[*627] gained by exploitation of the illegal seizure and that defendant’s motion to suppress was properly denied. Here, although defendant was arrested initially without probable cause, within a minute or two of the illegal arrest the sergeant knew that the bag of money and beer were the fruits of the crime. This information was sufficient to supply the necessary probable cause to arrest the defendant. The showup was conducted shortly after this intervening event which was of such significance that it broke any causal connection between the illegal arrest and the otherwise admissible showup identification. Thus, like the confession in Johnson, the showup was not conducted until sometime after the police had independently obtained the probable cause necessary to arrest the defendant. The showup identification, therefore, was neither an exploitation nor a product of the initial illegal detention, and was sufficiently attenuated from the initial detention as to make its suppression unnecessary (see Wong Sun v United States, 371 US 471; see, also, Brown v Illinois, 422 US 590; cf. Dunaway v New York, 442 US 200).

IV

The final issue concerns the propriety of the showup itself. “Speedy viewings, on the scene if possible, benefit both law enforcement and the defendant * * * If the accused is identified as the culprit, the witness’ recollection will be as fresh and reliable as his capacity and the situation permits. If the accused is not identified, he may then be released with a minimum of delay.” (People v Blake, 35 NY2d 331, 337.) Thus, it has been held that one-on-one showups proximate in time and location to the point of arrest, such as occurred here, are not constitutionally infirm (People v Brnja, 50 NY2d 366, 372, supra). Although it may be undesirable to display a suspect while he is handcuffed and standing alone, procedures that are less than ideal are tolerable in the interest of prompt identification (see People v Love, 57 NY2d 1023, 1024; People v Adams, 53 NY2d 241, 249). We, therefore, find no merit to defendant’s contention that the identification procedure in this case was unnecessarily suggestive. In view of this determination, there is no need to address defendant’s[*628] contention that an independent basis was not established for an in-court identification by the victim.

Accordingly, the judgment of conviction should be affirmed.

Dillon, P. J., Denman, Boomer and Green, JJ., concur.

Judgment unanimously affirmed.