POOLER, Circuit Judge:
Joseph Freeman appeals from a judgment of the United States District Court for the Southern District of New York (Paul A. Crotty, /.) convicting him of one count of possession of a firearm after a felony conviction, in violation of 18 U.S.C. § 922(g)(1), following a bench trial on stipulated facts. Freeman moved to suppress the firearm discovered by the police on the grounds that he was stopped without reasonable suspicion, as the stop was primarily based upon a pair of anonymous 911 calls from the same caller. The district court denied the motion to suppress and proceeded to trial on stipulated facts. We hold that there was not reasonable suspicion to support the stop of Freeman and vacate his conviction. The suppression decision is reversed, and the case is remanded to the district court.
BACKGROUND
I.
On April 27, 2011, at approximately 1:40 a.m., the New York City Police Department (“NYPD”) responded to two 911 calls from the same caller reporting that a man matching a certain description had a gun. The caller refused to identify herself, and the 911 operator could not re-contact her on multiple attempts. The 911 calls were recorded, and the number was identified as coming from a cell phone, but the caller was never identified. Her identity remains unknown. The caller told the 911 operator that a Hispanic male, wearing a black hat and a white t-shirt had a gun, near the Chase Bank on East Gun Hill Road in the Bronx, New York. The radio dispatch received by the police officers indicated that “a person is possibly armed with a firearm” and was “arguing with a female” near a particular intersection in the vicinity of the Chase Bank. NYPD Officers Joseph Walsh and Ryan Conroy responded to the call from “seven to eight” blocks away. While en route to the location, over the police radio, which was audible to Walsh and Conroy, another officer repeatedly asked the dispatcher to verify whether the 911 caller “actually saw a firearm.” Each time, the dispatcher was unable to confirm if that was the ease. Walsh and Conroy arrived at the scene within minutes of the first call: the first call came into the 911 system at 1:36 a.m., the.updated description from the second call was received at 1:38 a.m., and Freeman was stopped at 1:40 a.m. 911 Incident Record, Exhibit E, Declaration of Sarah Baumgartel, United States v. Freeman, [95]*95No. 11-er-567, 2011 WL 5419739 (S.D.N.Y. Nov. 8, 2011), ECF No. 15. As they approached the location, Walsh heard another radio dispatch indicating that the suspect was actually a “male black” wearing a white du-rag, black hat, and a long white t-shirt. The dispatcher indicated that the original caller had called back and stated that the male was “walking towards” and then “standing on the corner of Burke [Avenue].”
From their unmarked police vehicle, Walsh and Conroy canvassed the area on East Gun Hill Road between Burke Avenue and Young Avenue. The officers observed Joseph Freeman walking eastbound on East Gun Hill Road and observed that he fit the most recent reported description. The officers drove past Freeman, stopped their unmarked car along the side of the road, and waited for Freeman to approach their car. As Freeman walked past the vehicle, Conroy got out of the passenger side and approached him. Conroy attempted to speak to Freeman, but Freeman did not stop walking, so Conroy placed his hand on Freeman’s elbow. Freeman “shrugged [Conroy] off’ and kept walking. Walsh, who had since exited the car, also placed his hand on Freeman’s elbow, and Freeman again “shrugged [Walsh] off and kept walking.”- By the officers’ own admission, Freeman “never ran from [the officers] that night.” After Freeman continued walking upon being touched by Walsh, Walsh grabbed him around the waist in what the government now describes as a “bear hug.” After Walsh grabbed Freeman around the waist, Freeman never broke away. He attempted -to continue walking, but Walsh tripped him to the ground. After a short struggle, and with the assistance of NYPD Officer Humberto Morales, who arrived on scene with his partner after Walsh tripped Freeman, the police handcuffed Freeman and removed a gun from his waistband.
II.
On August 17, 2011, Freeman moved in district court to suppress the gun discovered on the ground that the police lacked reasonable suspicion to stop him. The district court concluded that the police stop was supported by reasonable suspicion and denied the motion to suppress. To preserve his suppression argument for appeal, Freeman waived his right to trial by jury and agreed to a bench trial on stipulated facts. On December 13, 2011, after Freeman stipulated to the elements of the offense charged in the indictment, the district court found him guilty of being a felon in possession of a firearm in violation of 18 U.S.C. § 922(g)(1). Freeman now appeals.
DISCUSSION
In an appeal from a district court’s ruling on a motion to suppress, we review legal conclusions de novo and findings of fact for clear error. United States v. Ferguson, 702 F.3d 89, 93 (2d Cir.2012). Mixed questions of law and fact — including the determination as to reasonable suspicion — are reviewed de novo. United States v. Lucky, 569 F.3d 101, 105-06 (2d Cir.2009). This Court reviews the underlying “findings of historical fact only for clear error and ... give[s] due weight to inferences drawn from those facts by resident judges and local law enforcement officers.” Ornelas v. United States, 517 U.S. 690, 699, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996).
Under the long-established rule of Terry v. Ohio, 392 U.S. 1, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968), police may only [96]*96stop someone when they have “reasonable suspicion supported by articulable facts that criminal activity may be afoot.” United States v. Sokolow, 490 U.S. 1, 7, 109 S.Ct. 1581, 104 L.Ed.2d 1 (1989) (internal quotation marks omitted). Reasonable suspicion must be “based on specific and articulable facts” and not on “inchoate suspicion or mere hunch.” United States v. Bayless, 201 F.3d 116, 132-33 (2d Cir.2000) (internal quotation marks omitted). This Court will “look at the totality of the circumstances of each case to see whether the detaining officer has a particularized and objective basis for suspecting legal wrongdoing.” United States v. Arvizu, 534 U.S. 266, 273, 122 S.Ct. 744, 151 L.Ed.2d 740 (2002) (emphasis added) (internal quotation marks and citation omitted). While we evaluate this totality of the circumstances “through the eyes of a reasonable and cautious police officer on the scene, guided by his experience and training,” we do “not merely defer to the police officer’s judgment.” Bayless, 201 F.3d at 133 (internal quotation marks omitted). Moreover, such a stop must be “justified at its inception.” Terry, 392 U.S. at 20, 88 S.Ct. 1868. Any events that occur after a stop is effectuated cannot contribute to the analysis of whether there was reasonable suspicion to warrant the stop in the first instance.
A.
As an initial matter, we must first determine when exactly the police seized Freeman, in order to assess whether there was reasonable suspicion for the stop “at its inception.” Terry, 392 U.S. at 20, 88 S.Ct. 1868. The government urges this Court to conclude that Freeman was not seized until the police had placed handcuffs on him. This argument cannot stand. A seizure triggering the protection of the Fourth Amendment occurs once an officer has “by means of physical force or show of authority, ... in some way restrained the liberty of a citizen.” Id. at 19 n. 16, 88 S.Ct. 1868. The government, perhaps attempting to minimize the restraint placed upon Freeman, now refers to Walsh grabbing Freeman around the waist as placing him in a “bear hug.” This ursine description does nothing to lessen the restraint placed upon Freeman. The contention that Walsh grabbing Freeman around the waist in a “bear hug” was not “in some way restraining]” his liberty is simply wrong. Though there was a brief struggle in bringing Freeman to the ground, Freeman never broke away once Walsh had placed his arms around Freeman’s waist. Freeman was physically restrained as soon as Walsh grabbed him in a “bear hug,” and thus the justification for the stop must have preceded Walsh’s grabbing of Freeman. See id. at 20, 88 S.Ct. 1868.
The government likely seeks to push the moment of the seizure forward in time in order to reap the benefits of the decisions in California v. Hodari D., 499 U.S. 621, 111 S.Ct. 1547, 113 L.Ed.2d 690 (1991), United States v. Muhammad, 463 F.3d 115 (2d Cir.2006), and United States v. Swindle, 407 F.3d 562 (2d Cir.2005). In each of these cases, the police ordered an individual to stop but the person did not comply and attempted to flee. Further, in each case, the reviewing court concluded that because the seizure was not effectuated at the mere command to stop, the ensuing flight provided the police with reasonable suspicion for the seizure once they had caught up with the fleeing individual. None of these cases are applicable in the instant case, where Freeman never broke away from the police or tried to flee prior to being stopped. Just as we previously recognized in United States v. Simmons, 560 F.3d 98 (2d Cir.2009), the cases Ho-[97]*97dari D., Muhammad, and Swindle, have no applicability where the initial seizure is neither broken away from or where the individual does not flee before he is seized. See Simmons, 560 F.3d at 106. The government’s attempted reliance on this line of cases in the instant matter rests on the same misapplication of the precedent that we rejected in Simmons. Id. at 106-07.
The rule from Hodari D. — the grounds for a stop may ... be based on events that occur after the order to stop is given, — has been applied only in cases where the suspect attempts to flee from police after being ordered to stop. Ho-dari D., Swindle, and Muhammad all involved a police show of authority, a defendant who refused to comply by fleeing, police pursuit, and a ‘seizure’ occurring at the moment the defendant was physically restrained.
Simmons, 560 F.3d at 106 (internal quotation marks citations omitted). The facts of this ease, by contrast, involve an initial approach by the police, followed almost immediately by a physical seizure. Freeman’s short struggle and “arguably suspicious,” id. at 107, movement of his hands toward his waist came after the seizure had occurred. As .the seizure occurred once Walsh grabbed Freeman, we will not consider anything that occurred after that moment in assessing whether there was reasonable suspicion to support the stop. Having concluded that Freeman was seized when grabbed around the waist by Walsh, we now consider whether there was reasonable suspicion to support that stop “at its inception.” Terry, 392 U.S. at 20, 88 S.Ct. 1868.
B.
Anonymous tips, without further corroboration by the police to demonstrate that the tip has sufficient indicia of reliability, are insufficient to provide the reasonable suspicion necessary for a valid Terry stop. See Alabama v. White, 496 U.S. 325, 329-32, 110 S.Ct. 2412, 110 L.Ed.2d 301 (1990). The Supreme Court further refined this holding in Florida v. J.L., 529 U.S. 266, 120 S.Ct. 1375, 146 L.Ed.2d 254 (2000). In J.L., the police received a tip that “a young black male standing at a particular bus stop and wearing a plaid shirt was carrying a gun.” Id. at 268, 120 S.Ct. 1375. Police arrived at the bus stop shortly thereafter, observed a young black male in a plaid shirt, approached him, frisked him, and discovered he was carrying a gun in his pocket. Id. There, the Court held that an anonymous tip — even one that proved accurate in both the description of an individual’s appearance and location — is an insufficient basis for a Terry stop. Id. at 272, 120 S.Ct. 1375. The Court determined that the stop was unconstitutional and the gun had to be suppressed.
While “there are situations in which an anonymous tip, suitably corroborated, exhibits sufficient indicia of reliability to provide reasonable suspicion to make the investigatory stop,” such corroboration did not exist in J.L. Id. at 270, 120 S.Ct. 1375 (internal quotation marks omitted). There, the anonymous call lacked corroboration because it “provided no predictive information and therefore left the police without means to test the informant’s knowledge or credibility.” Id. at 271, 120 S.Ct. 1375. A stop based upon an anonymous tip is only warranted where the tip is established to be “reliable in its assertion of illegality, not just in its tendency to identify a determinate person.” Id. at 272, 120 S.Ct. 1375. Anonymous tips differ from those for which the source is known on two determinative grounds: (1) ability to assess the credibility and reputation for honesty of the tipper and (2) holding the informant accountable for false reporting. [98]*98Id. at 270, 120 S.Ct. 1375 (citing Adams v. Williams, 407 U.S. 143, 146-47, 92 S.Ct. 1921, 32 L.Ed.2d 612 (1972)). Information from a known informant can be assessed for reliability in a way that information from an unknown one simply cannot.
1.
In this case, the pair of anonymous calls to 911 lacked any indicia of reliability and did not provide the police with the reasonable suspicion needed to stop Freeman. The district court below and the government attempt to distinguish the call in the instant case from that in J.L. to no avail. The district court determined that the call was not “truly anonymous” because the cell phone number was automatically recorded by the 911 system, the individual twice called 911, and based upon the information conveyed in the call, the caller was an eyewitness. The government now similarly aims to distinguish J.L. by arguing that, among other things, the calls had sufficient indicia of reliability because the physical description was more detailed here than that in J.L. and because the calls reflected Freeman’s “precise and changing location.”
While the proffered distinctions are indeed factual differences between J.L. and the instant case, they are not ones that undermine the reasons why anonymous phone calls must have sufficient indicia of reliability to support a finding of reasonable suspicion, nor do they provide those indicia for this caller or do anything to alter the reliability analysis laid out in J.L. While in J.L. the Court did indeed note that the call was not recorded, this was not the determinative question. Id. at 268, 120 S.Ct. 1375. Just as in J.L., here “nothing is known about the informant.” Id. As the caller has not placed her “anonymity at risk,” id. at 276 (Kennedy, J., concurring), this call is no different than that in J.L. The fact that the call was recorded and that the caller’s apparent cell phone number is known does not alter the fact that the identity of the caller is still unknown, leaving no way for the police (or for the reviewing court) to determine her credibility and reputation for honesty — one of the main reasons tips from known sources are afforded greater deference than anonymous ones. Id. at 270, 120 S.Ct. 1375. Moreover, while the government argues that the. fact that her number is known would now allow police to track her down, and thus she could be open to the consequences of false reporting, she never has been tracked down, so there is no way for this Court to determine that the number actually would trace back to the individual who made the phone call. There is nothing offered to suggest, for example, that the phone was not a prepaid phone, which would be as anonymous as a call placed from a public pay phone. The fact that her identity remains unknown unhinges the risk of consequences from the fact of the calls. • Knowledge of the caller’s- number — without more — has not altered the factors that underlie J.L.’s demand that anonymous calls be supported by additional indicia of reliability. Moreover, reasonable suspicion must exist at the time a Terry stop is made. Terry, 392 U.S. at 20, 88 S.Ct. 1868. At the time this stop was made — and continuing to this day — the police did not know if they would be able to track down the caller, and thus, had and have no way of knowing whether the consequences for false reporting at all influenced this caller to tell the truth. Thus, even though the call was recorded, the two factors that distinguish tips from known and unknown sources are both still operative in this case — the caller’s credibility cannot be assessed and there is no risk of consequences for a false report in this instance.
[99]*99In his concurrence in J.L., Justice Kennedy supposed that advances in technology may provide reliability to tips that in earlier years would have been considered unreliable anonymous tips. J.L., 529 U.S. at 276, 120 S.Ct. 1375 (Kennedy, J., concurring). However, not every advance in technology in the ensuing decade has advanced the government’s ability to identify callers and hold them to account for false tips. For example, mobile phones have proliferated. While a landline is necessarily registered to a particular person and particular place, some mobile phones can be prepaid with cash, stripping them of all identifying information. Given this uncertain and constantly shifting landscape, the government cannot claim the benefit of a general trend toward reliability driven by technology. In each case, the government must show its relevant technological capacities and how.those capacities enhanced reliability in that particular instance. Here, the record does not suggest a material increase in the reliability of the tip.
The dissent suggests that we are imposing on the government an extraordinary burden in each case to “hunt down the citizen tipster” to disprove anonymity in each case of 911 cell phone calls. That is not accurate. We are simply applying the J.L. anonymity analysis to the specific nature of the calls here and concluding there was not enough in the record before the district court to find sufficient indicia of reliability. It may very well be that similar calls with more evidence of identification of the caller could satisfy J.L., or that different calls with more information would have the necessary reliability. And, it may very well be that the specific nature of the cell phone technology or contact may contribute to an identification of the caller. But it does not here.
That the caller contacted 911 twice simply means that the content of both calls could not be assessed based on the caller’s reputation for honesty and that the caller could face the consequences for untruthful reporting for neither one of the calls. The fact that the anonymous caller made both of these calls; refused to identify herself, and could not be reached back via phone by the 911 operator merely indicates that there were two anonymous calls from the same individual, instead of just one. It does nothing to bolster her credibility. Furthermore, that the caller was an eyewitness makes the instant case no different than J.L.; there, in order to observe that J.L. was waiting at the bus stop clad in a plaid shirt, the caller would have been an eyewitness as well. Here, as the government stresses, the caller gave more information about Freeman’s ’ appearance and location than did the caller in J.L., but that does not make the information provided anything but “[a]n accurate description of [Freeman’s] readily observable location and appearance.” J.L., 529 U.S. at 272, 120 S.Ct. 1375. The fact that the anonymous call here was more detailed as to physical description and location does not alter the fact that such details merely serve to “correctly identify the person whom the tipstér means to accuse.” Id. Increased specificity on these dimensions does nothing to “show that the tipster has knowledge of concealed criminal activity.” Id. Just as in J.L., “[t]he reasonable suspicion here at issue requires that a tip be reliable in its assertion of illegality, not just in its tendency to identify a determinate person.” Id. Likewise, the district court’s reliance on the fact that Freeman was the only individual in the area matching the description, as compared to three individuals who matched the description in J.L., does nothing to alter the reliability of the tip or the overall analysis of whether there was reasonable suspicion. That Freeman was the only individual in the area matching the description simply [100]*100means that it was that much easier to “correctly identify the person whom the tipster means to accuse,” id. Identifying “a determinate person,” a task made easier by Freeman being the only individual matching the description, does not bolster the tip’s reliability “in its assertion of illegality,” and thus this case remains governed by the rule laid out in J.L. Id.
The district court seems to indicate that the caller’s description of the individual’s location — namely, walking east on Burke Avenue — constituted “predictive” information that distinguishes this call from thát in J.L. In doing so, the lower court misapprehends what constitutes predictive information. The caller did riot predict that the individual would begin walking in a certain direction, such that it demonstrated that the “tipster had knowledge of concealed criminal activity.” Id. at 272, 120 S.Ct. 1375. Rather, the caller simply described the fact that the individual was indeed walking on Burke Avenue. Such contemporaneous description does not constitute prediction of future behavior that imbues a tip with greater reliability. “Anyone could have ‘predicted’ that fact because it was a condition presumably existing at the time of the call.” White, 496 U.S. at 332, 110 S.Ct. 2412.
2.
In J.L., the Court expressly rejected a firearms exception to Terry’s demand that a stop be supported by reasonable suspicion, specific and particularized. J.L., 529 U.S. at 272-73, 120 S.Ct. 1375. The Court acknowledged the danger posed by firearms and reaffirmed that this danger is already accounted for by the very rule of Terry — allowing the police to stop and frisk based upon reasonable suspicion, instead of demanding “that officers meet the higher standard of probable cause.” Id. at 272, 120 S.Ct. 1375. The reasonable suspicion standard of Terry and its progeny is already a downward deviation from the probable cause that is otherwise required in order for a search or seizure to be1 reasonable and within the bounds of the Constitution.
We recognized a narrow exception to the rule of J.L. in United States v. Simmons, 560 F.3d 98. In Simmons, an anonymous 911 caller reported an “assault in progress” that possibly involved a firearm. Id. at 101. Based upon the police’s need to “to act on reports of an emergency situation without delay,” we held “that an anonymous 911 call reporting an ongoing emergency is entitled to a higher degree of reliability and requires a lesser showing of corroboration than a tip that alleges general criminality.” Id. at 105. Notably, however, in Simmons, we' recognized that it was a ‘close’ case. Id. at 107. Recognizing this fact, we now expressly decline to expand the reach of that limited exception, because to expand Simmons to other anonymous reports made to the police, without more, would serve to ignore the rule laid out by the Court in J.L. Any exigency in this case was weaker than the exigency in Simmons, as this case contained no report of an ongoing assault (or violence of any kind). Consequently, more particularized evidence was required in this case. Insofar as the police had the same level of reasonable suspicion as they did in Simmons, this suspicion was insufficient to justify a Terry stop in this case.
Nearly every anonymous call made to 911 implicates the need for the police in some way; thus, this jurisprudence must recognize the difference between a latent crime, even “simple possession of a firearm,” Simmons, 560 F.3d at 104, and an ongoing emergency, or it fails to heed the rule of J.L. To the extent the district court made a factual finding that a “gun run” was more serious than mere “possession of [101]*101a firearm,” id., we conclude the district court clearly erred. The officers’ own testimony indicates that such a¡ radio call is indicative of an individual possibly having a gun. To accept the government’s contention, and the district court’s apparent conclusion, to the contrary would be— through the operation of the emergency exception this Court laid out in Simmons — to create the very firearm exception that the Court rejected in J.L. J.L., 529 U.S. at 272-73, 120 S.Ct. 1375.
C.
In assessing whether a stop is supported by reasonable suspicion we “look at the totality of the circumstances” in order to determine “whether the detaining officer has a particularized and objective basis for suspecting legal wrongdoing.” Arvizu, 534 U.S. at 273, 122 S.Ct. 744 (internal quotation marks omitted). In the instant case, in addition to the 911 .call we have already discussed, the district court also relied upon the facts that the stop occurred at night in a ‘high crime’ area and that Freeman continued walking, at the same pace, when initially approached by the plain clothes officers.1 We conclude that these additional factors do not nudge the officers’ concerns over the line to reasonable suspicion.
To begin, the neighborhood and time of the stop are not individualized facts specific to Freeman. Nor do they enhance the reliability of the phone call by confirming in it some individualized detail. Of course, these factors could still have relevance. “[OJfficers are not required to ignore the relevant characteristics of a location in determining whether the circumstances are sufficiently suspicious to warrant further investigation.” Illinois v. Wardlow, 528 U.S. 119, 124, 120 S.Ct. 673, 145 L.Ed.2d 570 (2000). On the facts of this ease, however, these factors do not contribute meaningfully to a finding of reasonable suspicion. For instance, the fact that the encounter occurred late at night is a relatively weak and generic factor in this case,2 and the general label “high crime area” is not a substitute for analysis of the underlying testimony. • On direct examination, Officer Walsh recited some of the crimes that took place in the area, but gave no sense of the length of time over which those incidents occurred or whether the number of incidents was atypical. Officer Walsh also suggested that the police treated the area as being relatively high in crime, given the duties and attire of officers in the area, but offered little elaboration on the basis for this treatment. ' Indeed, on cross-examination, Officer Walsh [102]*102confirmed that he had told the prosecutor only that the area at which the stop took place was ‘one of the higher crime areas in [the] precinct,’ meaning that ‘there are areas with more crime and there are areas with less crime’ within the precinct. It is difficult to assign significant probative value to these observations.3
The government also cites the fact that Freeman did not stop and shrugged off the officers who touched his arms. Up until this point, there was not reasonable suspicion for. the officers to seize Freeman. “[W]hen an officer, without reasonable suspicion or probable cause, approaches an individual, the individual has a right to ignore the police and go about his business. And any refusal to cooperate, without more, does not furnish the minimal level of objective justification needed for a detention or seizure.” Wardlow, 528 U.S. at 125, 120 S.Ct. 673 (internal quotation marks and citation omitted). It is true that we have previously noted that such refusals “when viewed in light of the circumstances,” can serve to “reinforce[ ]” the police’s reasonable suspicion to stop the individual on the grounds he may be “engaged in criminal activity.” Simmons, 560 F.3d at 108. This observation, however, is not a mandate that converts every such refusal into support for a finding of reasonable suspicion. It still demands an assessment of the totality of the circumstances and serves only to “reinforce” such reasonable suspicion, not to create such suspicion where it does not otherwise validly exist. In Simmons, the case wherein we made this point, we already determined that the call was imbued with reliability because of the emergency nature of the 911 call, which, as discussed above, is not the case here.
In the instant case, we conclude that, at the time the officers initially approached Freeman, they did not have reasonable suspicion to stop him — in fact, the government has not even at this stage of the litigation identified the “specific and artic-ulable facts” that would have justified a stop of Freeman at that point, Bayless, 201 F.3d at 132; instead, they have attempted to delay the point of the seizure in order to include incidents that occurred after the seizure in the reasonable suspicion analysis. As the police lacked reasonable suspicion, Freeman certainly had the right to ignore the officers and continue on his way. Freeman merely continued walking in the same direction, and ostensibly at the same pace, as the police never contended otherwise. If we accepted the government’s argument that such a simple refusal to comply could create reasonable suspicion where none existed before, we would create a truly paradoxical class of individuals: individuals who cannot be stopped by officers, but who can be stopped if they refuse to stop. Such a conclusion would gut the Court’s repeated determination that an individual approached by the police “need not answer any question put to him; indeed, he may decline to listen to the questions at all and may go on his way. He may not be detained even momentarily without reasonable, objective grounds for doing so; and his refusal to listen or answer does not, without more, furnish those grounds.” Florida v. Royer, 460 U.S. 491, 497-98, 103 S.Ct. 1319, 75 L.Ed.2d 229 (1983) (plurality opinion) (emphasis added) (internal citations omitted); see also Wardlow, 528 U.S. at 125, 120 S.Ct. 673. The “more” eontem-[103]*103plated by Royer, sufficient to give rise to reasonable suspicion, does not exist in this case.
III.
“Reasonable suspicion” does not mean simply accepting whatever circumstances are offered by the government as necessarily demonstrating sufficient grounds to suspect “legal wrongdoing.” Arvizu, 534 U.S. at 273, 122 S.Ct. 744. Judicial review admittedly considers the facts through the eyes of a reasonable officer, id., but this is not a rubber stamp. We will not and “district court[s] must not merely defer to [a] police officer’s judgment.” Bayless, 201 F.3d at 133. Just because the police, and subsequently the prosecution, have offered circumstances that they contend constitute reasonable suspicion, this Court will not read out the importance of using the perspective of a .“reasonable” officer in assessing such factors. A reasonable officer is one who has been trained in and has knowledge of the governing law, including the constitutional protections guaranteed by the Fourth Amendment. Cf. Simms v. Vill. of Albion, 115 F.3d 1098, 1106 (2d Cir.1997) (qualified immunity analysis considers “reasonably well-trained officer”). Taken together, the circumstances must provide grounds for suspecting actual legal wrongdoing, Arvizu, 534 U.S. at 273, 122 S.Ct. 744.
CONCLUSION
For the reasons stated above, the judgment of the district court hereby is REVERSED, and the case is REMANDED to the district court for further proceedings consistent with this opinion.