claims about game provider's bot use in "fair" and "skill-based" games must go to trial

Skillz Platform Inc. v. Papaya Gaming, Ltd., 2025 WL 3012836, 24cv1646(DLC) (S.D.N.Y. Oct. 27, 2025)

Most recent previous opinion discussed here. Skillz sued its competitor in the “real-money skill-based mobile gaming” market for violating the Lanham Act and NY GBL § 349 by stating or implying that its games pit human players against each other when in fact Papaya employed bots against human players. Here, the court denies summary judgment to Papaya.

“In RMSB games, players are matched by the platform with other users on games created by third parties and compete to win cash prizes or for game rewards.” Skillz most commonly offers head-to-head competition between two players, while Papaya offers multi-player tournaments with larger cash prizes. In various ads, Papaya has described its games as “fair” and “skill-based,” represented that Papaya has “no vested interest” in and does not “profit” from who wins or loses its tournaments, and refers to “players” “over the age of 18,” “individuals,” and “winners” as the users on its platform. Its video ads showed images of humans playing games, and claimed that its games “are directly determined by your level of skill” and that Papaya will “match [you] against players with similar skills.”

It was undisputed that Papaya used bots in its games from 2019 until at least November 2023. Its liquidity bots filled in user slots to ensure that a tournament “won’t stay open for too long” as it awaits players. Its tailored bots were used to create “a predetermined outcome” where Papaya “want[s] the player to finish at some predefined rank,” which requires “control[ling] the scores of the rest of the players.” All players in a tailored-instance outside the ‘receiving’ player” are bots. “Papaya customers inquired between 2021 and 2023 whether they were playing against other human beings or against a computer or a bot. Papaya’s responses generally emphasized that it matched players with similar skill levels and assured the complaining customers over 200 times that ‘we do not use bots.’” Papaya even closed the accounts of two players who expressed suspicion about Papaya’s use of bots.

Papaya regularly identified Skillz as its largest direct competitor and benchmark for performance assessments. During Papaya’s bot era, Skillz saw a decrease in market share while Papaya saw an increase.

Papaya argued that Skillz lacked sufficient evidence that any of its advertising statements communicated a false message to consumers, were material, or caused Skillz injury. The court disagreed.

A reasonable jury could find literal falsity or falsity by necessary implication, leading to a presumption of deception. Papaya’s representations that its tournaments are “fair,” “skill-based,” and only between “players” and “individuals” who are “over the age of 18” could be false based on the undisputed evidence of bot use. Similarly, statements that Papaya has “no vested interest” in who wins or loses, and that it does not “profit on the outcome of a Tournament,” could be literally false because there is evidence that Papaya used bots to control the outcomes of tournaments (and speed them up so players could use their winnings on new games). [Frankly, I’d be inclined to grant summary judgment to Skillz on falsity!]

Papaya argued that it never expressly stated in its advertisements that there were no bots in its games. “The issue is what a reasonable consumer would understand Papaya’s statements to mean. Should a jury find, for instance, that the advertising necessarily and unambiguously implied that only human players competed in Papaya’s games, it would be entitled to find that the advertising was literally false.”

Papaya argued that “skill-based,” “fair,” “players,” and “individuals” were not literally false because those terms had multiple meanings. [Insert goose meme: what meanings? What ones???]. The jury could find that, in context, they were literally false or false by necessary implication. Papaya’s expert concluded that skill played a greater role than chance in the outcome of Papaya’s tournaments, regardless of bot usage. So what? It was undisputed that Papaya used tailored bots to control the outcomes of tournaments. Papaya “could prevent players from winning–or allow them to win–no matter how they performed in the game.”

Papaya also argued that its statements that it had “no vested interest” in and did not “profit” on tournament outcomes were accurate, because Papaya made less money in sessions with bots because bots do not pay entry fees. Its experts opined that human players received more in total cash prize awards due to Papaya’s use of bots. That was a jury question. [Seems pretty inconsistent with the profit motive—did human players walk away with more money total because of bots, or did they play more games/stay longer?]

Implied falsity: This can be shown through either “extrinsic evidence of consumer confusion” or “evidence of the defendant’s deliberate deception,” the latter of which must be “egregious” to “create[] a rebuttable presumption of consumer confusion.”

There was sufficient evidence for a reasonable jury to find deliberate deception, given Papaya’s repeated denials of bots’ existence. E.g., when one player asked, “Is every single player I play against in a tournament a real player or are there computers playing at times?”, Papaya responded, to “clarify that we do not use bots or computer players” and that “if it were possible to control [tournaments] manually, it would not be fair to other players.” Papaya CSRs were instructed to escalate any complaints that referenced bots to management, and one employee stated that a customer service representative “displayed poor judgment” by keeping open an account of a consumer who had complained about bots and demanded a refund. Papaya had an internal policy of removing posts in its Facebook group that mentioned bots.

Papaya argued that there should be no presumption of deception because Skillz failed to provide evidence that any intent to deceive was linked to a specific challenged statement. But internal statements by executives and the company’s awareness of consumer confusion, among other things, are relevant to intent. Here, there was evidence that Papaya and its executives recognized that its use of bots “did not align” with its public statements and that this “misalignment” confused customers.

Indeed, executives decided to respond to a consumer who asked for a direct answer on whether Papaya used bots by informing the user that the platform has “real players” and placing the blame on “rare cases where players use automated systems.” Recognizing that consumers noticing bots on the platform was a growing issue and impacted “trust in our fairness,” Papaya executives modified its bots’ performance to make bot profiles appear more human so that fewer users would detect their usage going forward. Another employee stated: “While understanding the need for bots, the margin of their winnings should be smaller and less obvious to our players.” A reasonable jury could find intentional deception, despite the testimony of Papaya’s Rule 30(b)(6) witness, who stated that instances in which Papaya told customers that it did not use bots were “an unfortunate mistake” committed by the customer support vendor and “not Papaya’s position.” The court noted that “Papaya’s executives have not made similar representations. They invoked their Fifth Amendment privilege to remain silent during their depositions and have not offered affidavits in support of this motion.” [Yikes.] Regardless, there was enough to go to a jury.

Skillz also offered extrinsic evidence of consumer confusion created by Papaya’s use of the terms “skill-based,” “fair,” “players,” “winners,” “individuals,” and “no vested interest,” in its advertising. Along with individual consumer responses, Skillz provided survey evidence.

Respondents had played at least one of four games that Skillz and Papaya offer on their platforms. Upon watching a Papaya ad, 45.8% of respondents indicated that they believed that the game used solely human players. The survey next presented respondents with a mock app store description constructed from Papaya’s public statements:

Solitaire cash makes sure to match players against other opponents with a similar skill level to ensure a fun and fair experience for everyone. You’ll be matched with other players within the same skill level, and you will get the same deck – so the game is totally fair and skill-based. The outcome of Solitaire cash is based on the skill of the players, rather than luck or chance. Solitaire cash has no vested interest in who wins or loses, nor does it profit on the outcome of a tournament that we provide.

After reading the description, 60% of respondents indicated that they thought the game included only human players.

Papaya argued that the survey was inadmissible because it did not use a control, stripped the statements to which respondents reacted of their “real world” context, and did not survey the appropriate populations, excluding prospective game players and thereby including too many men. These were issues for cross-examination, not demanding exclusion.

Papaya didn’t explain what kind of control could have improved the survey’s reliability. [This is a missed opportunity! A control ad could have just described the game mechanics and not the opponents and tested whether consumers thought they’d received any message about whether they were playing against other humans. I wouldn’t be at all surprised if the net confusion levels were still pretty high, given the explicit claims. Still, the court didn’t appear to understand the purpose of a control: “The survey is designed to test the impression created by the words Papaya chose to describe its games. Testing responses to statements that Papaya did not use in its advertising would be neither relevant nor illuminating.”]

Papaya also argued that the survey wrongly screened for past players of RMSB games, not prospective consumers to whom the ads were directed towards, who differed demographically. Skillz’ expert filed a supplemental report looking at what happened when he removed randomly selected male respondents to match the ratio of male to female respondents in Skillz’s own survey of Skillz’s actual customers. The differences were, he reported, statistically insignificant.

Materiality: While the materiality of the falsity and the injury to the plaintiff resulting from the defendant’s falsity are “separate essential elements ... where the defendant and plaintiff are competitors in the same market and the falsity of the defendant’s advertising is likely to lead consumers to prefer the defendant’s product over the plaintiff’s,” proof of injury to the plaintiff may also demonstrate the materiality of the falsehoods.

Skillz relied on evidence that consumers complained to Papaya and closed their accounts with Papaya when they came to believe that Papaya was using bots in its games, as well as on Papaya’s efforts to deceive consumers about its use of bots. It also offered a second survey that sought to measure the likelihood of consumers continuing to play if they learned that some of their opponents in real-money games whom they thought were human were actually bots.

Half the respondents were randomly asked if they would “continue” to play if they were informed that some of their opponents were bots, and the other half were asked if they would “stop” playing. Of the former group, 51.2% of respondents indicated they would be unlikely or very unlikely to continue playing. Of the latter, 66.8% indicated they would be likely or very likely to stop playing.

[It seems to me that with less evidence of deception, that amount of a swing (over 15% of respondents) from a small wording change would be concerning—but frankly, even if you take 51.2% as a midpoint and say that it might reflect only a range from 35%-66% materiality, those are big enough numbers that it shouldn’t matter. Which is also to say that I’m not sure that you should draw anything but qualitative conclusions from surveys like this. Also:]

The survey had a control: half the respondents were also asked if they would continue playing – with the other half being asked if they would stop – if they were informed that they had to wait one hour to learn the results of the game. Here, the numbers were 40.8% (unlikely or very unlikely to continue playing) and 54.8% (likely or very likely to stop playing), suggesting that the wording affected about 15% of respondents, which with these numbers didn’t matter much. “At bottom, regardless of how the question was phrased, a significant number of respondents showed a propensity to cease playing when they learned there were bots in the games.”

Papaya argued that the survey didn’t measure the materiality of bots, but rather the materiality of learning of intentional deception. That went to the weight of the survey, not its admissibility. [Especially if there was intentional deception!]

Papaya also argued that bots are so ubiquitous in the industry that consumers “fully expect” them to be deployed, and that its failure to disclose was in line with industry practice. “These arguments fail to engage with the evidence of deception and materiality offered by Skillz.”

Injury: Injury “can be established when defendant and plaintiff are competitors in a relevant market and plaintiff demonstrates a logical causal connection between the alleged false advertising and its own sales position.” Papaya regularly identified Skillz as its largest competitor and benchmark for performance assessments, and Papaya’s growth in market share in 2021 coincided with Skillz’s decline in market share. “Even if Skillz were unable to quantify its damages with sufficient certainty to recover those damages, it could seek injunctive relief and disgorgement of the defendant’s ill-gotten profits.” And Skillz offered an expert to calculate both damages to Skillz and Papaya’s unjust enrichment, the latter of which was between $650 and $719 million, depending on whether that profit was measured by Papaya’s cost savings from its use of bots or by how much less it would have earned without deception. That report was admissible and relevant.

Papaya argued that “the but-for world in a false advertising case must be one in which the false statements were never made and cannot be one in which the accused business practice did not occur.” But the expert’s choice was not unreasonable. “There is evidence that Papaya may not have been able to operate in the RMSB marketplace at all if it publicly and accurately disclosed its use of bots.” If it did so, it might have been classified as a gambling platform by app stores—unlike RMSB platforms, gambling platforms hold a stake in the results. RMSB games don’t involve “playing against the house” and are supposed to collect the same revenue regardless of who wins, whether from entry fees, advertising revenue, or in-app purchases. “[H]ad Papaya disclosed its use of bots, Papaya may have been faced with the regulatory and tax requirements with which gambling platforms must comply. In addition, it may have lost access to app stores and payment processors and have been excluded from the RMSB market.”

GBL claims survived for the same reasons.