WSJ: Unveiling the Secret Jury That Controls Disputes on Polymarket

marsbitPublished on 2026-05-18Last updated on 2026-05-18

Abstract

Last month, Garrick Wilhelm lost a $567 bet on the Polymarket prediction platform about whether a ceasefire would be reached with Hezbollah. When a truce was announced, some traders argued it counted, but Wilhelm disagreed. The dispute was settled not by Polymarket, but by a decentralized group of UMA token holders who vote on such disagreements. As trading surges, resolving ambiguous outcomes is a growing challenge for prediction markets. Unlike competitors like Kalshi that decide internally, Polymarket outsources dispute resolution to UMA. Its token holders, mostly anonymous and with voting power weighted by holdings, arbitrate cases. Critics argue this system is prone to manipulation, as voters can also bet on the same markets they judge. A Wall Street Journal analysis found that over the past year, at least 60% of active UMA voters had corresponding Polymarket accounts and held positions in disputes they voted on. Voting power is also concentrated among a few large holders. Polymarket says only 0.2% of bets go to UMA and that the system disperses authority. Its founder has acknowledged flaws and promised fixes. UMA's backers deny any proven manipulation, dismissing critics as sore losers. The platform penalizes voters in the minority to incentivize "correct" outcomes. Disputes are rising, covering topics from a streamer's pregnancy announcement to Iran. This model also helps Polymarket argue it's an offshore platform outside U.S. regulation, a shift made after a 2022 s...

Last month, Garrick Wilhelm joined the rush into prediction market trading, but he soon began to regret it. This user based in British Columbia, Canada, registered on the Polymarket platform and started placing bets on events related to the Middle East situation, with one betting option being whether Israel and Hezbollah could reach a ceasefire agreement.

Wilhelm put in $567 on a losing bet, convinced that the armed group would absolutely not sign a ceasefire agreement, calling it a sure win.

Subsequently, Israel and the Lebanese government reached a truce agreement. Some traders deemed this equivalent to reaching a ceasefire with Hezbollah. After carefully studying the platform rules, Wilhelm firmly disagreed with this determination.

The outcome of this bet involving millions of dollars was not ultimately decided by the Polymarket platform. Wilhelm then learned that the resolution of his trade was to be arbitrated by a loosely organized group holding cryptocurrency tokens.

With a massive influx of new users and explosive growth in trading volume, trading disputes have become an increasingly thorny problem for prediction market platforms like Polymarket. Platforms originally strive to frame betting topics as clear-cut, black-and-white questions, but real-world events are complex, and the line between win and loss is often blurry.

Most other similar prediction market platforms, like Kalshi, resolve controversies and determine final outcomes themselves; but Polymarket chooses to outsource its dispute arbitration business to a third-party service called UMA. When parties to a trade disagree on the payout result, the UMA voting arbitration mechanism is triggered. Voting power is controlled by UMA token holders, with more tokens conferring greater voting weight, and the vast majority of voters remain anonymous.

Polymarket explicitly states in its user agreement that the platform assumes no responsibility for arbitrating disputes related to trading contracts.

Many traders and crypto industry veterans point out that this UMA voting system is highly susceptible to manipulation. Token holders can freely vote on contentious bets in which they have a financial stake, with no institutional constraints.

Analysis by The Wall Street Journal, combining Polymarket trading data and on-chain information, found: Over the past year, at least 60% of active UMA voters could be matched to Polymarket trading accounts; in over 300 dispute cases during the same period, UMA voters were found to hold financial positions in the very bets under dispute.

UMA bills itself as decentralized, but on-chain data shows voting power is highly concentrated among a few "whales." Statistics indicate that in the vast majority of dispute votes, the top ten wallet addresses account for over half the votes.

Nic Carter, founding partner of venture capital firm Castle Island Ventures, bluntly stated that Polymarket shouldn't shirk its dispute resolution responsibilities. "Arbitrating disputes is inherently part of Polymarket's job. It shouldn't be outsourced to this anonymous, unidentified third-party group of token holders."

A Polymarket spokesperson responded that only 0.2% of betting contracts on the platform trigger UMA voting arbitration, and added that UMA disperses adjudication authority into an open market system rather than leaving it to a single entity.

In March of this year, Polymarket founder Shayne Coplan admitted at a Harvard Business School forum that the platform's current dispute resolution mechanism indeed has many flaws. "Related optimization solutions are about to be implemented," he said, but did not reveal specific reform details. It is reported that Polymarket has entered into a data partnership with Dow Jones, the parent company of The Wall Street Journal.

UMA was co-founded by two former Goldman Sachs traders and is managed by the Risk Labs Foundation, registered in the Cayman Islands. Foundation spokesperson James Fry stated that no conclusive evidence has been found to suggest UMA manipulates trades. "Much of the external skepticism is just losing traders looking for excuses."

When disputes arise, UMA token holders debate on the Discord social platform, citing various pieces of evidence to support their views. UMA also has a penalty mechanism that financially penalizes users who vote in the minority, a move the platform says aims to guide voters toward factually correct determinations.

According to statistics from Polymarket's dedicated trading terminal Betmoar, from the beginning of 2026 to now, the platform has seen over 1,150 betting disputes, a number that has already surpassed the total for all of 2025.

A recent popular dispute involved a bet on whether streamer Clavicular would officially announce a pregnancy: the streamer did publicly announce a partner's pregnancy, but many traders argued this announcement didn't meet the contract's requirement of a "formal and valid declaration." Ultimately, UMA voting ruled the announcement was compliant. Additionally, multiple arbitration disagreements have erupted over bets related to the situation in Iran.

Public regulatory filings show that Polymarket initially handled various disputes internally; in early 2022, the platform reached a settlement with the U.S. Commodity Futures Trading Commission over alleged violations of U.S. financial regulations and subsequently handed over all dispute arbitration work to UMA.

This model of adjudication by a dispersed group of token holders also serves as a key basis for Polymarket to argue it is an offshore platform, not subject to U.S. domestic regulation.

However, Polymarket occasionally overturns UMA's final rulings and also proactively adds supplementary clarifications to betting contract details to avoid potential disputes.

The novice trader Wilhelm mentioned earlier ultimately lost his bet related to the ceasefire agreement; 87% of UMA token voters determined that the Israel-Lebanon truce agreement applied to the Hezbollah-related bet. Despite reasoned arguments from Wilhelm and others, they were powerless to change the ruling.

A group of losing traders formed a Discord community called "Whale Hunters," collectively accusing top UMA voting users of suspected backroom dealings.

Traders have pointed fingers at the startup project UMA.rocks, a platform that allows UMA token holders to pool their voting tokens and delegate voting authority to a specialized decision committee. Its votes recently accounted for about 8% in various dispute votes and are also seen by the market as an important indicator of UMA's overall ruling trends.

UMA.rocks founder Lancelot Chardonnet responded: "Many traders' losses are essentially because they didn't seriously read the betting rules, but afterward they shift all the blame to UMA and our platform. We are the easiest targets."

At the end of April, UMA.rocks formally expelled voting committee member Scout, citing his past suspected market manipulation behavior.

Contacted via Discord, Scout denied manipulating the market or deliberately guiding incorrect voting outcomes but readily admitted to participating in Polymarket bets on disputed markets while exercising UMA dispute voting power.

Scout believes that voters with such conflicting interests can actually make rulings more aligned with the facts. "Voters with no skin in the game might spend at most 5 minutes casually looking into the matter; but we, as stakeholders, driven by our own financial interests, will verify the full picture of the event to make an accurate judgment."

He acknowledged the industry faces a dilemma: "Either use traders with conflicting interests in the adjudication, or let complete outsiders with no expertise drive the voting. There's currently no perfect answer."

Related Questions

QAccording to the article, what is the main difference between how Polymarket and other prediction markets like Kalshi handle trade disputes?

AMost prediction markets like Kalshi resolve disputes internally. In contrast, Polymarket outsources the arbitration of disputes to a third-party service called UMA, where voting power is held by anonymous holders of UMA tokens.

QWhat finding did the Wall Street Journal's analysis reveal about UMA voters and their activity on Polymarket?

AThe Wall Street Journal's analysis found that over the past year, at least 60% of active UMA voters could be linked to trading accounts on Polymarket, and in more than 300 dispute cases, UMA voters themselves held financial positions in the contested markets.

QWhat did Polymarket's founder, Shayne Coplan, admit about the platform's current dispute resolution mechanism?

AShayne Coplan admitted at a Harvard Business School forum that the platform's current dispute resolution system has many flaws and stated that improvements are on the way, though he did not provide specific details.

QHow does UMA claim its system encourages accurate voting outcomes?

AUMA has a penalty mechanism that financially punishes users who vote with the minority opinion. The platform claims this is designed to incentivize voters to make factually correct decisions.

QWhat dilemma does the expelled UMA.rocks committee member Scout highlight regarding the dispute resolution system?

AScout highlights a dilemma: the system must either use arbitrators who have a conflict of interest (because they are also traders in the market) or rely on outsiders with no professional stake or deep understanding of the event, suggesting there is currently no perfect solution.

Related Reads

From South Korea to the United States: Blue-Collar Jobs Are Becoming Increasingly Popular, Thanks to AI

AI is reshaping the labor market's value proposition. The traditional four-year college degree is losing its appeal as a guaranteed career path, while skilled blue-collar trades like electricians, welders, and plumbers are experiencing historic demand and wage premiums. This shift is driven by dual pressures: AI's displacement of certain white-collar roles and a booming need for physical infrastructure and data center construction. Data confirms the trend. In the U.S., vocational school revenue surged, and a significant portion of recent layoffs are AI-related. Surveys show a majority of Gen Z adults plan to pursue blue-collar work, citing better job security against AI automation. Vocational education interest has exploded recently. Experts cite a psychological shift as younger generations seek tangible, AI-resistant careers and avoid high student debt. In many cases, salaries for skilled trades now match or exceed those requiring a bachelor's degree. In South Korea, semiconductor vocational high schools boast near-total employment, with graduates securing high-paying roles at companies like Samsung. The shortage is structural, exacerbated by a retiring baby boomer workforce and massive infrastructure projects. Companies like JPMorgan Chase, Meta, and Lowe's are investing heavily in training programs. However, overcoming historical stigma and a "perception gap" around trade careers remains a key challenge to closing the talent gap.

marsbit33m ago

From South Korea to the United States: Blue-Collar Jobs Are Becoming Increasingly Popular, Thanks to AI

marsbit33m ago

Qualcomm: AI Hype Subsides, When Will Smartphones Emerge from the Gloom?

Qualcomm reported its Q3 FY2026 results (ending June 2026), with revenue of $9.95B, down 4% YoY but above expectations. Gross margin declined to 53.1%, pressured by rising costs across manufacturing and memory. Key business segments showed mixed performance: Handset revenue fell 19.6% YoY to $5.09B, dragged by an 11% decline in non-Apple Android shipments and weaker high-end mix. Conversely, Automotive revenue surged 61% to $1.59B, and IoT grew 9% to $1.83B. Core operating profit dropped 41% YoY due to margin compression and higher expenses. Management's Q4 FY2026 guidance projects revenue of $9.7B-$10.5B, in line with consensus, but Non-GAAP EPS guidance of $2.05-$2.25 fell short of expectations. Amidst persistent weakness in its core handset market, Qualcomm is pursuing growth in AI, focusing on Edge AI (smartphones, PCs, automotive) and Data Center AI. Its data center strategy includes four pillars: AI accelerators (e.g., AI200), commercial CPUs (Dragonfly C1000), custom silicon, and connectivity solutions. While these initiatives initially boosted its stock, concerns over AI capital expenditure sustainability have since erased those gains. The company targets $5B in data center revenue for FY2027 and $15B for FY2029. The report concludes that with the traditional handset business still under pressure, the data center opportunity is currently viewed as a longer-term option, and a more conservative valuation based on core operations may be warranted until AI contributions materialize.

marsbit37m ago

Qualcomm: AI Hype Subsides, When Will Smartphones Emerge from the Gloom?

marsbit37m ago

From TPU to Self-Evolving Agents: How Jeff Dean Predicts the Next Step in AI

At the 2026 YC Startup School, Jeff Dean outlined his vision for AI's next phase, shifting focus from simply scaling models to building intelligent, autonomous systems. He believes AI's progress is no longer just about creating smarter models, but about integrating them into systems capable of long-term, iterative work, automated experimentation, and continuous learning. This evolution moves the competition from "who has the bigger model" to "who can best organize intelligence." Dean suggests AI capabilities are now comparable to a junior engineer, enabling the automation of complex workflows. However, the true challenge and opportunity lie in managing these AI "workers" at scale. He emphasizes the importance of **context engineering**—structuring tools, memory, and feedback loops—over raw model power. For startups, this means building deep expertise in niche domains where general models currently fail (near 0-1% success rates), leveraging proprietary data, specialized tools, and domain-specific evaluators. A recurring theme is re-examining fundamental constraints. Dean's past work, like moving Google's search index to memory or creating the TPU, stemmed from questioning outdated assumptions about hardware and cost. He sees similar inflection points today, particularly in **specialized inference hardware** to drastically reduce latency and energy consumption for real-time Agent operation. Notably, he points out that in modern AI systems, the dominant cost is often not computation but **data movement**. Reliable, long-running Agents require robust system design, borrowing concepts from distributed computing like checkpointing, state management, and parallel exploration to handle failures and maintain progress over days or weeks. As AI automates execution, the scarcest human skills will shift to **defining clear specifications**, **judging what problems are worth solving** (taste), and designing effective feedback loops. Ultimately, Dean's framework prioritizes understanding the problem deeply, identifying the true bottlenecks, and systematically building closed-loop systems where AI can not only perform tasks but also improve AI itself.

marsbit37m ago

From TPU to Self-Evolving Agents: How Jeff Dean Predicts the Next Step in AI

marsbit37m ago

Trading

Spot
活动图片