concurring.

After a considerable amount of time spent reviewing the pertinent cases and in applying their teachings to the facts of the instant case, I have decided that I am able in good conscience to concur in the result reached by the majority. I consider this to be an extraordinarily close case — one in which the “correct” answer was by no means apparent.at the outset. What in the end convinced me that probable cause was lacking is the fact that, after considering the other factors in a totality of the circumstances review, I was struck by the reality that there is no indication in the record before us that the firstvtime informant advised the police of the basis for his or her assertion that the defendant would be at the Royal Buffet in possession of crack cocaine.

OPINION

The defendant, Ray Burgess, stands convicted of one count of possession of cocaine. On appeal, he argues that the trial justice erred in denying his motion to suppress physical evidence seized and statements he made following an unlawful arrest. We agree and, therefore, vacate his conviction.

Facts and Travel

On October 26, 2011, during the course of a traffic stop, Rhode Island State Police Officer Brian Macera (Det. Macera)1 discovered that the passenger of the stopped vehicle was the subject of an outstanding warrant.2 We shall refer to this passenger *197as the confidential informant. Before the traffic stop, the confidential informant was unknown to Det. Macera. In the hopes of avoiding arrest, the confidential informant told Det. Macera that a clean-shaven African American male with short-cropped hair and a thin build was distributing crack cocaine in the vicinity of Providence and Cranston. The confidential informant identified this drug dealer only by his nickname, “CJ.” Detective Macera also was informed that CJ drove a blue Mercury vehicle with unknown Rhode Island registration plates. Finally, the confidential informant told Det. Macera that CJ either was at or would soon arrive at the Royal Buffet restaurant in Cranston, where he would be found in possession of crack cocaine.3 Armed with this information, Det. Macera contacted his supervisor, who, in turn, contacted the Rhode Island State Police High Intensity Drug Trafficking Area Task Force (task force).

Later that day, Det. Macera, the confidential informant, and several members of the task force proceeded to a Cranston shopping plaza in which the Royal Buffet is located. Detective Macera testified at the suppression hearing that “surveillance was set up at the Royal Buffet,” but it is evident from his testimony that the “surveillance” consisted of nothing more than establishing a police perimeter around the shopping plaza and waiting for CJ to emerge. No attempt was made to locate a blue Mercury vehicle in the shopping piar za’s parking lot because of the “large amount of vehicles” in the plaza.

When defendant — who Det. Macera testified matched the physical description of CJ provided - by the confidential informant — exited the Royal Buffet, the confidential informant identified him as CJ. The defendant immediately was taken into custody. According to Det. Macera, he and his colleagues “surrounded” defendant,4 had him place his hands on a vehicle,5 and conducted a pat-down search. During the course of the pat-down of defendant’s outer clothing for weapons, Det. Macera felt a “bulge” in defendant’s front pocket; Det. Macera reached into defendant’s pocket to remove the material, which turned out to *198be $62 in cash. Detective Macera read defendant his Miranda rights and explained that^ he was suspected of selling drugs. At first,- defendant told the officers that he was not in possession of any drugs.

During this encounter, after he had been “surrounded” by police officers, defendant “appeared [to Det. Macera] to be very nervous”;' “his body was trembling,” and he “would not make eye contact” with the officer. When’Det. Macera asked defendant about the odor of burnt marijuana that he detected, defendant admitted that he had smoked that substance earlier that day. Detective Macera informed defendant that a State Police K-9 unit was en route to the scene. Faced with this prospect, defendant lowered his head and said, “It’s in my pocket.” A search of the waistband of defendant’s jacket revealed two bags of suspected crack cocaine. After the discovery of the contraband, defendant identified his vehicle: a blue Chevrolet Impala with Rhode Island license plates. Detective Macera testified that defendant’s vehicle is similar to'the vehicle described by the confidential informant. In consideration of the assistance provided, to law enforcement, the confidential informant was not charged with any crime,- and the record is silent ■ about what," if anything, happened with respect to the outstanding warrant or why it'was issued.6

The defendant, on the other hand, was charged with one count of possession of cocaine. He moved to. suppress the two bags of crack cocaine and his statements to Det. Macera, arguing that he was arrested without probable cause and that the subsequent search of his jacket violated the state and federal constitutions. After a suppression hearing, the trial justice denied the motion, concluding that police had probable cause to arrest defendant. In reaching his probable-cause determination, the trial justice relied on the following evidence from the suppression hearing: Det, Macera’s testimony that he smelled the odor of burnt marijuana when questioning defendant; defendant’s nervous behavior and incriminating statement that “[fit’s in my pocket”; the “surveillance” that occurred at the Royal Buffet before defendant was apprehended; and the match between defendant’s appearance and the description given by the confidential informant.

After trial,.a jury convicted defendant of the charged offense, and the trial justice sentenced him to two years of probation. The defendant timely appealed.

Standard of Review

Our review of the -grant or denial of a motion to suppress is multifaceted. We defer to the trial justice’s findings of historical fact unless clearly erroneous, but we review de novo the ultimate question of whethér probable cause exists. See State v. Chum, 54 A.3d 455, 460 (R.I. 2012); *199State v. Casas, 900 A.2d 1120, 1129 (R.I. 2006); see also Ornelas v. United States, 517 U.S. 690, 691, 699, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996).

Analysis

Before embarking on our analysis, we reiterate that the state has conceded on appeal that, from the moment he was surrounded by the officers in the parking lot of the Royal Buffet, defendant was under arrest. This circumstance alters the factual landscape upon which the probable-cause determination so heavily depends.7 A valid arrest must be supported by probable cause at its inception; information— however incriminating it might be — that is discovered after a suspect is arrested cannot be used to establish the existence of probable cause. See State v. Ortiz, 824 A.2d 473, 480 (R.I. 2003) (“Rhode Island case law indicates that the existence of probable cause at the time of arrest determines its legality. * * * ‘[P]robable cause * * * must not be derived from evidence that a subsequent search might disclose.’ ” (quoting State v. Kryla, 742 A.2d 1178, 1182 (R.I. 1999))); In re John N., 463 A.2d 174, 178 (R.I. 1983) (“It is axiomatic that the legality of an arrest is determined, by the existence of probable cause at the time of the arrest and not by what a subsequent search may disclose.”); see also United States v. Di Re, 332 U.S. 581, 595, 68 S.Ct, 222, 92 L.Ed. 210 (1948) (“[A] search is not to be made legal by what it turns up. In law it is good or bad when it starts and does not change character from its success.”).

In this case, several of the facts relied upon by the trial justice came to light after the point in time at which the state concedes defendant was arrested. The trial justice’s conclusion that probable cause existed was grounded, in part, on Det. Mac-era’s testimony that he smelled the odor of burnt marijuana while questioning defendant after providing him with Miranda warnings and defendant’s nervous behavior and incriminating-statement, that “[i]t’s in my pocket.” Detective Macera discovered this evidence, however, only after he and several other officers “surrounded” defendant and, as the state concedes, arrested him in the parking lot of the Royal Buffet. Similarly, police did not learn that defendant drove a blue Chevrolet Impala with Rhode Island license plates or that defendant had $62 in cash and two bags of crack cocaine on his person until- after they arrested him. Thus,- the only pre-arrest evidence that might support probable cause to arrest defendant consists of the confidential informant’s- tip, the “surveillance” conducted by the officers in the shopping plaza-before defendant exited the Royal Buffet, and the consistency between defendant’s physical appearance and the confidential informant’s description of CJ. To its credit, the state does not rely on any of the post-arrest evidence,- but instead argues that' probable cause existed at the time of defendant’s arrest. We are not convinced.

Probable cause to arrest a suspect exists when “the facts and circumstances within * * * [the officers’] knowledge and of which they had reasonably trustworthy information [are] sufficient in themselves to warrant a man of reasonable caution in the belief that an offense has been or is being committed” and that the suspect is the perpetrator. Chum, 54 A.3d *200at 462 (quoting State v. Flores, 996 A.2d 156, 161 (R.I. 2010)); see also Ortiz, 824 A.2d at 480. Because “probable cause is a fluid concept — turning on the assessment of probabilities in particular factual contexts — not readily, or even usefully, reduced to a neat set of legal rules,” we must carefully examine the totality of the circumstances in determining whether probable cause exists. Illinois v. Gates, 462 U.S. 213, 232, 103 S.Ct. 2317, 76 L.Ed.2d 527 (1983); see also id. at 230-31, 238, 103 S.Ct. 2317.

This fluidity extends to probable-cause determinations involving informants’ tips. “Informants’ tips, like all other clues and evidence coming to a policeman on the scene, may vary greatly in their value and reliability.” Gates, 462 U.S. at 232, 103 S.Ct. 2317 (quoting Adams v. Williams, 407 U.S. 143, 147, 92 S.Ct. 1921, 32 L.Ed.2d 612 (1972)). Thus, courts no longer analyze an informant’s tip by artificially compartmentalizing the informant’s veracity or reliability, on the one hand, and his or her basis of knowledge, on the other. See id. at 230, 233, 103 S.Ct. 2317. Instead, we examine all relevant considerations surrounding the tip and any other relevant factors of the case in the familiar totality-of-the-cireumstances analysis. See id. at 230-31, 233, 103 S.Ct. 2317. In this analysis, “an informant’s veracity, reliability, and basis of knowledge remain highly relevant.” State v. Cosme, 57 A.3d 295, 303 (R.I. 2012) (quoting State v. King, 693 A.2d 658, 661 (R.I. 1997)); see also Gates, 462 U.S. at 230, 103 S.Ct. 2317. “[A] deficiency in one [of these three aspects of an informant’s tip] may be compensated for, in determining the overall reliability of a tip, by a strong showing as to [another aspect], or by some other indicia of reliability.” Gates, 462 U.S. at 233, 103 S.Ct. 2317. Additionally, “corroboration of details of an informant’s tip by independent police work” can strengthen the argument for probable cause in a particular case. Id., at 241, 103 S.Ct. 2317; see also Cosme, 57 A.3d at 303. Indeed, independent police investigation can carry the day in a probable-cause determination. Cf. State v. Keohane, 814 A.2d 327, 328, 330, 330-31 (R.I. 2003) (concluding that police had reasonable suspicion — and explaining that “any suspicions * * * approached the level of probable cause” — to stop a van where police were informed by an anonymous tip that the defendant “would be traveling to Providence in a van to purchase heroin” that would be sold in Woonsocket, and police corroborated the tip by observing the defendant, accompanied by “a known heroin user,” drive to Providence in a van, “approach a group of males,” and then “proceed[ ] back to Woonsocket, following the pattern predicted by the tipster,” driving in an erratic manner).

In examining the totality of the circumstances in this case, we conclude that, because of the combined effect of several aspects of the tip and the accompanying failure of the police to corroborate any information provided by the confidential informant, there was no probable cause to arrest this defendant. We begin by examining the tip itself. For starters, the confidential informant had not previously provided information to Det. Macera. Although past experience between an informant and police is by no means required, see State v. Ricci, 472 A.2d 291, 297 (R.I. 1984) (explaining that “[f]irst-time informants can be found to be reliable”), it is a factor that, when present, increases the likelihood that the informant’s information is reliable and truthful on the occasion under review. See, e.g., Draper v. United States, 358 U.S. 307, 309, 312-13, 313, 79 S.Ct. 329, 3 L.Ed.2d 327 (1959) (tip from informant who was a “ ‘special employee’ of the Bureau of Narcotics” for six months *201when tip was made and who had given “accurate and reliable” information in the past; court concluded that, in these circumstances, the officer “would have been derelict in his duties had he not pursued [the tip]”).

To be sure, the confidential informant hardly was anonymous to police; he or she8 was in State Police custody as a result of an outstanding warrant at the time the tip was given and when defendant was arrested. The state argues that this circumstance bolsters the confidential informant’s reliability because an untruthful tip would have resulted in negative consequences for the confidential informant. It is true that an informant’s statement against his or her penal interest may, in some circumstances, be deemed credible. See United States v. Harris, 403 U.S. 573, 583-84, 91 S.Ct. 2075, 29 L.Ed,2d 723 (1971) (plurality op.); State v. Read, 416 A.2d 684, 689 (R.I. 1980). Such a statement generally arises after the informant has been caught in the act. In this case, however, the confidential informant did not make a statement against his penal interest, as was the case in Read, upon which the state relies. See Read, 416 A.2d at 686-87, 689 (informant admitted to purchasing marijuana from the address that police later searched pursuant to a search warrant); see also Harris, 403 U.S. at 575-76, 583, 91 S.Ct. 2075 (informant admitted to purchasing illicit whiskey from address controlled by the defendant); United States v. Tyler, 238 F.3d 1036, 1038-39 (8th Cir. 2001) (informant identified the defendant as his supplier of drugs); State v. Grossi, 588 A.2d 607, 608 (R.I. 1991) (informant implicated himself in a drug-smuggling ring involving the defendant); State v. Germano, 559 A.2d 1031, 1035 (R.I. 1989) (informant admitted to purchasing cocaine from the defendant). Unlike the informants in the above-cited cases, the confidential informant did not admit to purchasing crack cocaine from defendant or that he was involved in defendant’s drug-dealing operations.9

The statement-against-penal-interest concept is closely tied to the informant’s basis of knowledge, another relevant consideration in the totality-of-the-circumstances analysis. See Gates, 462 U.S. at 230-31, 103 S.Ct. 2317; Cosme, 57 A.3d at 303. In all of the above-cited cases, each informant’s statement against penal interest established his basis of knowledge; having admitted to committing a crime, the informant necessarily had personal knowledge of the place, instrumentalities, or persons involved in that criminal activity. See, e.g., Harris, 403 U.S. at 575-76, 578-79, 91 S.Ct. 2075 (informant’s sworn tip that the defendant sold illicit whiskey at a particular address was based on informant’s purchase of illicit whiskey at that address); Grossi, 588 A.2d at 608 (informant’s tip that the defendant was involved *202in a drug-smuggling ring was based on informant’s own involvement in that operation); Germano, 559 A.2d at 1033, 1035 (informant’s tip that the defendant was selling cocaine out of his home was based, in part, on informant’s own purchase of cocaine from the defendant; informant also had visited the defendant’s home on numerous occasions and observed the defendant sell cocaine); Read, 416 A.2d at 686 (informant’s tip that marijuana was located at a particular address was based on informant’s purchase of marijuana at that location earlier that evening); see also Tyler, 238 F.3d at 1038-39 (informant’s tip that the defendant was a drug dealer was based, in part, on informant’s own purchase of drugs from the defendant); cf. Ricci, 472 A.2d at 293, 296-97 (informant’s tip that the defendant was in possession of stolen earrings was based on the fact that the defendant gave informant a sample of the stolen earrings, and informant confirmed with the victim of the theft that the earrings were stolen). In stark contrast to these cases, there is no evidence in this record to suggest that the confidential informant had personal knowledge that defendant was engaged in the trafficking of crack cocaine.10 All we know about the confidential informant is that he was the subject of an outstanding warrant issued for reasons not disclosed by the record.

Finally, the type of detail.provided by the confidential informant further undermines the value of this tip in the totality-of-the-circumstances appraisal. Apart from the assertion that defendant was in possession of crack cocaine (which may or may not have been based on the confidential informant’s personal knowledge), the tip boils down to an accurate characterization of defendant’s physical appearance, a technically incorrect, albeit close, description of defendant’s vehicle, arid a correct statement of the restaurant at which defendant either was or would soon be dining, see supra note 3. Although verifiable, the first two components of the tip — the description of defendant and his car — have no bearing on the reliability of the informant’s assertion that defendant was engaged in criminal activity. As the United States Supreme Court explained in the related context of a stop-and-frisk based on a tip from an anonymous informant:

“An accurate description of a subject’s readily observable location and appearance is of course reliable in this limited sense: It will help the police correctly .identify the person whom the tipster means to accuse. Such a tip, however, does not show that the tipster has knowledge of concealed criminal activity. The 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.” Florida v. J.L., 529 U.S. 266, 272, 120 S.Ct. 1375, 146 L.Ed.2d 254 (2000).

Cf. Alabama v. White, 496 U.S. 325, 332, 110 S.Ct. 2412, 110 L.Ed.2d 301 (1990) (“The fact that, the officers found a car precisely matching the [anonymous] caller’s description in front of the 235 building is an example of [an ‘easily obtained fact[ or] condition[ ] existing at the time of the tip’]. Anyone could have ‘predicted’ that fact because it was a condition presumably existing at the time of the call.” *203(quoting Gates, 462 U.S. at 245, 103 S.Ct. 2317)).

The tip in this ease failed to provide predictive detail from which police could assess the informant’s claims that he had knowledge of defendant’s criminal conduct. The state disputes this conclusion, arguing that the tip contained predictive information because the confidential informant “called [defendant] to confirm that he would be dining at the Royal Buffet sometime around 4:00 pm.” We are unpersuaded. At the outset, we note that the record is unclear on whether defendant was already at the Royal Buffet or instead on his way there at the time of the tip. See supra note 3. This uncertainty is caused by the state’s failure to elicit testimony concerning the telephone call during the suppression hearing. Instead, this information made its debut during defense counsel’s cross-examination of Det. Macera at trial, and the state successfully objected to several attempts by defense counsel to explore the details of this call.

In any event, even had the tip contained information regarding defendant’s future plans to visit the Royal Buffet, this lone detail pales in comparison to the predictive information provided by the confidential informants in the cases cited by the state. See, e.g., Draper, 358 U.S. at 309-10, 79 S.Ct. 329 (tip — given by an informant who was a paid employee of the Bureau of Narcotics and had provided reliable information in the past — indicated that the defendant had gone to Chicago the day before and would return to Denver by train either the next day or the day after, accurately described the precise clothing the defendant would be wearing and the tan zipper bag that he would be carrying, and correctly stated that the defendant “habitually” walked very quickly); United States v. Miller, 925 F.2d 695, 697 (4th Cir. 1991) (Powell, J.) (informant’s tip indicated that the defendant — a picture of whom the informant identified — would be traveling by bus and arriving on one of two days later that week; informant also accurately described the precise clothing that the defendant would be wearing and a tote bag that she would be carrying); cf. Keohane, 814 A.2d at 330 (concluding that anonymous tip was “sufficiently detailed, and thereafter corroborated, to warrant an experienced detective to become reasonably suspicious of [the defendant’s] behavior” where “[t]he tip provided details of where [the defendant] lived, the type of vehicle he would be driving, and the itinerary and alleged purpose of his travel to and from Providence”).

For these reasons, we conclude that the tip in this case, standing alone, failed to provide probable cause to arrest defendant. This conclusion, however, does not end the analysis because an informant’s tip that alone does not furnish probable cause can be shored up by independent police investigation. See Gates, 462 U.S. at 241, 103 S.Ct. 2317; Cosme, 57 A.3d at 303. The state contends that police corroborated the details of the confidential informant’s tip. We disagree with this assertion.

The “surveillance” referred to by Det. Macera consisted of establishing a police perimeter at the shopping plaza and confirming that the confidential informant recognized defendant. The police did not even attempt to corroborate the confidential informant’s description of defendant’s vehicle, having summarily deemed such an endeavor impractical because of the large number of vehicles in the parking lot; nor is there any evidence that they performed a records check. We conclude, on the basis of this record, that the efforts of the police in this case, even when combined with the tip, fell short of what was necessary to establish probable cause to arrest. *204Cf. Casas, 900 A.2d at 1132 (remarking, in the course of expressing doubt as to whether reasonable suspicion existed to justify vehicle stop, that “little, if any, informant information was confirmed before the stop,” notwithstanding four or five surveillance efforts by police).

In the final analysis, the state has not presented this Court with — and our independent research has not unearthed — any case in which probable cause to arrest a suspect was found based on a tip from a first-time informant who has merely been detained on an outstanding warrant, where the tip is devoid of predictive detail and fails to indicate the informant’s basis of knowledge of the alleged criminal activity, and police undertake no effort to corroborate or independently investigate such a bare-bones tip. Instead, the cases cited by the state are readily distinguishable. For example, in Draper, 358 U.S. at 309, 79 S.Ct. 329, the tip came from an informant who “had been engaged as a ‘special employee’ of the Bureau of Narcotics * * * for about six months, and from time to time [had given] information to Marsh,” the federal narcotics agent who arrested the defendant, and “Marsh had always found the information given by [the informant] to be accurate and reliable.” See also id. at 310, 79 S.Ct. 329. The Court explained that, because the tip “[came] from one employed for that purpose and whose information had always been found accurate and reliable, it is clear that Marsh would have been derelict in his duties had he not pursued it.” Id. at 313, 79 S.Ct. 329. The established reliability of the informant in Draper stands in stark contrast to the circumstances of this case, which involve a first-time informant who neither made a statement against his penal interest nor revealed his basis of knowledge.

Moreover, the tip in Draper contained predictions relating to specific details that the officer corroborated. The informant predicted that the defendant would arrive in Denver from Chicago by train on one of two consecutive mornings; the informant also told Marsh that the defendant would be “wearing a light[-]colored raincoat, brown slacks[,] and black shoes,” and that he would be walking at a rapid pace while carrying a tan zipper bag. Draper, 358 U.S. at 309 & n. 2, 79 S.Ct. 329. Two days later, Marsh observed the defendant, “wearing the precise clothing described by [the informant]” and “carrying a tan zipper bag,” disembark from an incoming Chicago train and quickly walk towards the exit of the train station. Id. at 309, 310, 79 S.Ct. 329. As the Supreme Court explained:

“Marsh had personally verified every facet of the information given him by [the informant] except whether [the defendant] had accomplished his mission and had the three ounces of heroin on his person or in his bag. And surely, with every other bit of [the informant’s] information being thus personally verified, Marsh had ‘reasonable grounds’ to believe that the remaining unverified bit of [the informant’s] information — that [the defendant] would have the heroin with him — was likewise true.” Id. at 313, 79 S.Ct. 329.

Another ease cited by the state, Miller, 925 F.2d at 697, involved an informant who similarly predicted specific details. The informant told Patrick, the investigator who ultimately made the arrest, that the defendant, whose identity the informant confirmed by identifying her picture, “would be bringing a shipment of drugs [by bus] from New York City” on one of two consecutive days. Id. Two days later, the informant confirmed that the defendant’s bus would be arriving later that day and that the defendant “would be wearing blue jeans and a blouse” and “carrying a brown tote bag with a shoulder strap.” *205Id. Later that day, Patrick observed the defendant, wearing the clothing the informant had predicted and carrying the brown shoulder bag, disembark from a bus arriving from New York City. Id. In this case, by contrast, the tip did not contain a prediction of the kind of specific detail related by the tips in Draper and Miller, and police did not even attempt to corroborate the readily verifiable detail concerning defendant’s vehicle or learn to whom it was registered.

Additionally, this case is distinguishable from Miller for another significant reason: police in this case had no prior knowledge that defendant was a drug dealer. In Miller, 925 F.2d at 697, Patrick had arrested the defendant for a drug offense about a year earlier. The court explained that prior knowledge of the defendant’s involvement in the narcotics trade, when coupled with Patrick’s substantial corroboration of the predictive detail related in the tip, “pointed towards the tip’s reliability.” Id. at 700; see also id. at 699-700. The factor of independent knowledge of a defendant’s previous involvement in drug trafficking was also present in- Tyler, another case relied upon by the state. In that case, the police “found that [the informant’s] account matched [the defendant’s] known reputation as a drug dealer.” Tyler, 238 F.3d at 1038. Moreover, the informant initially made contact with police when they arrested him for possession of drugs. Id. at 1039. Caught red-handed, the informant identified his source of the drugs — the defendant — and added that “he had purchased drugs from [the defendant] on ‘numerous occasions’ over the previous year.” Id. The court concluded that “[the informant’s] disclosures were presumptively credible because they were made against his penal interest.” Id. Unlike in Tyler, the confidential informant in this case neither made a statement against his penal interest nor revealed how he knew about defendant’s criminal activity.

In sum, we hold that the defendant’s arrest was not supported by probable cause. We emphasize that no one feature 'of this case compels this conclusion. Instead, after conducting the requisite totality-of-the-circumstances analysis, we are of the opinion that the facts and circumstances were not sufficient to “warrant a man of reasonable caution in the belief that an offense ha[d] been or [was] being committed” and that the defendant was the perpetrator. Chum, 54 A.3d at 462 (quoting Flores, 996 A.2d at 161); see Ortiz, 824 A.2d at 480. The defendant’s motion to suppress the evidence seized and his statements to police should have been granted.

Conclusion

For these reasons, we vacate the judgment of conviction. The papers may be remanded to the Superior Court.