This is a fundamental problem in how the platform is built. In this article, I\u2019ll explain why Polymarket, with $26 billion in quarterly volume, outsourced market resolution to a small protocol called UMA with a market cap of about $95 million. How whales make money from this, why you can\u2019t really do anything about it, and why Ukraine blocked Polymarket back in December 2025.<\/p>\n\n\n\n
In short: I do not recommend treating Polymarket as a serious tool. Details below.<\/p>\n\n\n\n
Polymarket: scale and why it matters<\/h2>\n\n\n\n![\"Polymarket](\"https:\/\/lwallet.com.ua\/wp-content\/uploads\/2026\/05\/blog-why-polymarket-is-a-scam-29-05-2026-content-pic2-1024x573.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nPolymarket is now the largest decentralized prediction market in the world. A few numbers to show the scale:<\/p>\n\n\n\n
\n- $21.5 billion in volume in 2025, and $26.2 billion in the first quarter of 2026 alone.<\/li>\n\n\n\n
- $10.57 billion in March 2026 \u2014 the first time monthly volume passed $10 billion.<\/li>\n\n\n\n
- Single-day record: $425 million in February 2026, more than on the day of Trump\u2019s 2024 election.<\/li>\n\n\n\n
- An $8 billion valuation after a $2 billion investment from ICE\/NYSE in October 2025.<\/li>\n\n\n\n
- 95 million transactions in 2025 and around 600,000 active traders.<\/li>\n<\/ul>\n\n\n\n
Polymarket runs on Polygon, and all bets are settled in USDC. The key point for this discussion: Polymarket is a non-custodial service. The platform does not hold user funds and does not make the final decision on event outcomes. This architecture was a deliberate choice after the CFTC issued Polymarket a cease-and-desist order in January 2022 and forced it to pay a $1.4 million fine for operating as an illegal derivatives exchange. The team redesigned the system so that, formally, it controls nothing: trading goes through smart contracts, while outcome resolution goes through the external UMA oracle. Legally, this means: \u201cwe are not an exchange, we are just an interface.\u201d<\/p>\n\n\n\n
For the user, it means something else: when you buy \u00abYes\u00bb, the decision on whether you won is not made by Polymarket. It is made by another project. Now let\u2019s look at how that project works.<\/p>\n\n\n\n
What UMA is and where this oracle came from<\/h2>\n\n\n\n
Before getting into the mechanics, we need some context. UMA (Universal Market Access) is a separate project founded in 2018 by Hart Lambur, a former Goldman Sachs trader. UMA was originally built as a platform for synthetic assets and derivatives, where users could create tokenized versions of Apple stock or the S&P 500 index. The synthetic-assets idea did not take off \u2014 Synthetix did it better and earlier \u2014 so UMA shifted toward the product that survived: an oracle for real-world data.<\/p>\n\n\n\n
Today, UMA Optimistic Oracle serves more than Polymarket. Its clients include Across, a cross-chain bridge, Sherlock, a smart-contract insurance project, and other DeFi protocols. But Polymarket is its largest and most controversial client. If you look at the statistics, most UMA disputes come from Polymarket markets, because this is the one use case where outcomes are subjective and politically charged.<\/p>\n\n\n\n
Polymarket and UMA are legally two different companies, but they are connected. In 2024, Polymarket and UMA announced a partnership with EigenLayer to develop a \u201cnext-generation\u201d oracle based on restaking mechanics. Polymarket officially calls UMA its trusted oracle, while UMA officially calls Polymarket its largest client. In other words, they are formally independent partners, but without Polymarket, UMA would lose a large part of its reason to exist. That creates another hidden asymmetry of interests, which I\u2019ll return to below.<\/p>\n\n\n\n
How bet resolution works through UMA Optimistic Oracle<\/h2>\n\n\n\n![\"UMA](\"https:\/\/lwallet.com.ua\/wp-content\/uploads\/2026\/05\/blog-why-polymarket-is-a-scam-29-05-2026-content-pic3-1024x573.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nUMA\u2019s architecture is called the Optimistic Oracle (OO), and it has two levels.<\/p>\n\n\n\n
Level 1: optimistic assertion.<\/strong> When a market ends, anyone can propose an outcome \u2014 for example, \u201cYes\u201d on the question of whether there was a ceasefire. To do this, they need to post a bond; on Polymarket this is usually $750 USDC. Then a 2-hour challenge period begins. If nobody disputes the proposal, the outcome is treated as correct, and the proposer gets their bond back plus a small reward. According to UMA\u2019s statistics, only about 1.5% of proposals are disputed. This is the fast path, and it covers the overwhelming majority of markets.<\/p>\n\n\n\nLevel 2: Data Verification Mechanism (DVM).<\/strong> If someone disputes the proposal, also by posting a $750 USDC bond, a vote begins. This is where the interesting part starts.<\/p>\n\n\n\nOnly holders of UMA tokens who have staked them can vote. Voting happens in two stages:<\/p>\n\n\n\n
1. Commit phase (24 hours): voters publish a hash of their vote. The real vote is hidden so nobody can simply copy someone else\u2019s decision.<\/p>\n\n\n\n
2. Reveal phase (24 hours): voters reveal their votes together with the salt used to create the hash.<\/p>\n\n\n\n
Voting power is proportional to the number of staked UMA tokens. A decision requires a supermajority of more than 65%. Those who vote with the majority receive rewards, while those in the minority are slashed, roughly 0.1% of their stake. The slashing is small, but it creates a strong psychological effect: smaller voters are afraid to be wrong, so they vote \u201cwith everyone else\u201d rather than according to reality.<\/p>\n\n\n\n
After the vote ends, the result is returned to the Polymarket smart contract, the market closes, and payouts go to the winners. That\u2019s it. Polymarket\u2019s own help center states it directly: \u00abOnce finalized by UMA, outcomes are immutable\u00bb. After UMA finalizes the resolution, the outcome cannot be changed. Polymarket cannot change it even if it wants to.<\/p>\n\n\n\n
Now the key question: who are the people voting?<\/p>\n\n\n\n
UMA\u2019s structural problem: the numbers do not add up<\/h2>\n\n\n\n
This is where the circus begins. Let\u2019s look at the numbers.<\/p>\n\n\n\nMetric<\/strong><\/th>Value<\/strong><\/th><\/tr><\/thead>Monthly Polymarket volume<\/td> $10.5 billion (March 2026)<\/td><\/tr> UMA market cap<\/td> around $95 million<\/td><\/tr> Total staked UMA<\/td> around 20 million tokens<\/td><\/tr> Top 10 voters hold<\/td> around 30% of average voting participation<\/td><\/tr> Largest known staker<\/td> up to 7.5 million UMA (37% of total stake)<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\nThink about that. The platform\u2019s monthly volume is more than 100 times larger than the market cap of the project that decides all disputes on it. When the Zelenskyy suit story exploded in July 2025, one market with $242 million in volume was decided by people holding tokens worth roughly $95 million. In literal terms, it is cheaper to buy influence over voters than to lose a large market.<\/p>\n\n\n\n
In March 2025, analyst Folke Hermansen looked into vote concentration and found that two whales controlled more than 50% of active voting power in an average vote. One individual held up to 7.5 million out of 20 million staked UMA, about 37% of the entire stake. This is not decentralization. This is an oligopoly covered with nice words about \u201ctruth-seeking\u201d.<\/p>\n\n\n\n
Separately: these whales often trade on Polymarket. In other words, they take a $1 million position and then vote in UMA in favor of their own outcome. The conflict of interest is built directly into the architecture. UMA says honesty is economically more profitable because if the system is discredited, the token will fall. That is a neat theory, but it breaks in specific situations where a one-off gain on a $7 million market is larger than the expected loss from a token drop.<\/p>\n\n\n\n
And the final point: only UMA holders can vote. If you entered Polymarket, put $50,000 on a market, and are clearly right on the facts, you still have zero influence over resolution. The decision is made by completely different people, who may not trade the market at all or, more often, may be trading the opposite side. This is the core asymmetry from which the whole problem grows.<\/p>\n\n\n
[vc_message color=”warning” message_box_style=”classic” message_box_color=”alert-warning” style=”rounded”]<\/p>\n
Only holders of staked UMA tokens can vote in the UMA Data Verification Mechanism. The holder of a USDC position on Polymarket, even if that position is worth a million dollars, cannot influence the resolution. This is a fundamental asymmetry of the system.<\/p>\n
[\/vc_message]<\/p>\n\n\n
Fresh case: Russia x Ukraine ceasefire by end of 2026? ($220 million, May 2026)<\/h2>\n\n\n\n![\"Polymarket](\"https:\/\/lwallet.com.ua\/wp-content\/uploads\/2026\/05\/blog-why-polymarket-is-a-scam-29-05-2026-content-pic4-1024x573.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nThe freshest and most revealing case. Let\u2019s break it down.<\/p>\n\n\n\n
Background.<\/strong> A series of \u00abRussia x Ukraine ceasefire by 2026\u00bb markets had been trading on Polymarket for a long time. The largest ones had deadlines of May 31, 2026 and June 30, 2026. The rules were extremely clear, and I am not paraphrasing here. This is the direct wording:<\/p>\n\n\n\n\u00abThis market will resolve to “Yes” if there is an official ceasefire agreement, defined as a publicly announced and mutually agreed halt in military engagement, between Russia and Ukraine. Only ceasefires which constitute a general pause in the conflict will qualify. Ceasefires which only apply to energy infrastructure, the Black Sea, or other similar agreements will not qualify. Any form of informal agreement will not be considered an official ceasefire. Humanitarian pauses will not count toward the resolution of this market.\u00bb<\/p>\n\n\n\n
The key part: humanitarian pauses do not count, informal agreements do not count, and the ceasefire must be a \u201cgeneral pause in the conflict\u201d \u2014 a real halt to hostilities across the front.<\/p>\n\n\n\n
What happened.<\/strong> On May 8, 2026, Trump announced a three-day ceasefire for May 9\u201311, timed to the May 9 parade in moscow. Both sides described it as a short-term pause in air strikes. A 1,000-for-1,000 prisoner exchange was also announced. Meanwhile, active assaults by russian occupying forces continued along the line of contact, so it was not even close to a ceasefire.<\/p>\n\n\n\nWas this an \u201cinformal agreement\u201d without an official deal? Yes. Was it a \u201cgeneral pause in the conflict\u201d? Obviously not: on May 12, one day after the \u201cceasefire\u201d ended, russia launched more than 200 drones at Ukrainian cities. The war continued without any real pause.<\/p>\n\n\n\n
<\/figure>\n<\/div>\n\n\nPolymarket is now the largest decentralized prediction market in the world. A few numbers to show the scale:<\/p>\n\n\n\n
- \n
- $21.5 billion in volume in 2025, and $26.2 billion in the first quarter of 2026 alone.<\/li>\n\n\n\n
- $10.57 billion in March 2026 \u2014 the first time monthly volume passed $10 billion.<\/li>\n\n\n\n
- Single-day record: $425 million in February 2026, more than on the day of Trump\u2019s 2024 election.<\/li>\n\n\n\n
- An $8 billion valuation after a $2 billion investment from ICE\/NYSE in October 2025.<\/li>\n\n\n\n
- 95 million transactions in 2025 and around 600,000 active traders.<\/li>\n<\/ul>\n\n\n\n
Polymarket runs on Polygon, and all bets are settled in USDC. The key point for this discussion: Polymarket is a non-custodial service. The platform does not hold user funds and does not make the final decision on event outcomes. This architecture was a deliberate choice after the CFTC issued Polymarket a cease-and-desist order in January 2022 and forced it to pay a $1.4 million fine for operating as an illegal derivatives exchange. The team redesigned the system so that, formally, it controls nothing: trading goes through smart contracts, while outcome resolution goes through the external UMA oracle. Legally, this means: \u201cwe are not an exchange, we are just an interface.\u201d<\/p>\n\n\n\n
For the user, it means something else: when you buy \u00abYes\u00bb, the decision on whether you won is not made by Polymarket. It is made by another project. Now let\u2019s look at how that project works.<\/p>\n\n\n\n
What UMA is and where this oracle came from<\/h2>\n\n\n\n
Before getting into the mechanics, we need some context. UMA (Universal Market Access) is a separate project founded in 2018 by Hart Lambur, a former Goldman Sachs trader. UMA was originally built as a platform for synthetic assets and derivatives, where users could create tokenized versions of Apple stock or the S&P 500 index. The synthetic-assets idea did not take off \u2014 Synthetix did it better and earlier \u2014 so UMA shifted toward the product that survived: an oracle for real-world data.<\/p>\n\n\n\n
Today, UMA Optimistic Oracle serves more than Polymarket. Its clients include Across, a cross-chain bridge, Sherlock, a smart-contract insurance project, and other DeFi protocols. But Polymarket is its largest and most controversial client. If you look at the statistics, most UMA disputes come from Polymarket markets, because this is the one use case where outcomes are subjective and politically charged.<\/p>\n\n\n\n
Polymarket and UMA are legally two different companies, but they are connected. In 2024, Polymarket and UMA announced a partnership with EigenLayer to develop a \u201cnext-generation\u201d oracle based on restaking mechanics. Polymarket officially calls UMA its trusted oracle, while UMA officially calls Polymarket its largest client. In other words, they are formally independent partners, but without Polymarket, UMA would lose a large part of its reason to exist. That creates another hidden asymmetry of interests, which I\u2019ll return to below.<\/p>\n\n\n\n
How bet resolution works through UMA Optimistic Oracle<\/h2>\n\n\n
\n<\/figure>\n<\/div>\n\n\nUMA\u2019s architecture is called the Optimistic Oracle (OO), and it has two levels.<\/p>\n\n\n\n
Level 1: optimistic assertion.<\/strong> When a market ends, anyone can propose an outcome \u2014 for example, \u201cYes\u201d on the question of whether there was a ceasefire. To do this, they need to post a bond; on Polymarket this is usually $750 USDC. Then a 2-hour challenge period begins. If nobody disputes the proposal, the outcome is treated as correct, and the proposer gets their bond back plus a small reward. According to UMA\u2019s statistics, only about 1.5% of proposals are disputed. This is the fast path, and it covers the overwhelming majority of markets.<\/p>\n\n\n\n
Level 2: Data Verification Mechanism (DVM).<\/strong> If someone disputes the proposal, also by posting a $750 USDC bond, a vote begins. This is where the interesting part starts.<\/p>\n\n\n\n
Only holders of UMA tokens who have staked them can vote. Voting happens in two stages:<\/p>\n\n\n\n
1. Commit phase (24 hours): voters publish a hash of their vote. The real vote is hidden so nobody can simply copy someone else\u2019s decision.<\/p>\n\n\n\n
2. Reveal phase (24 hours): voters reveal their votes together with the salt used to create the hash.<\/p>\n\n\n\n
Voting power is proportional to the number of staked UMA tokens. A decision requires a supermajority of more than 65%. Those who vote with the majority receive rewards, while those in the minority are slashed, roughly 0.1% of their stake. The slashing is small, but it creates a strong psychological effect: smaller voters are afraid to be wrong, so they vote \u201cwith everyone else\u201d rather than according to reality.<\/p>\n\n\n\n
After the vote ends, the result is returned to the Polymarket smart contract, the market closes, and payouts go to the winners. That\u2019s it. Polymarket\u2019s own help center states it directly: \u00abOnce finalized by UMA, outcomes are immutable\u00bb. After UMA finalizes the resolution, the outcome cannot be changed. Polymarket cannot change it even if it wants to.<\/p>\n\n\n\n
Now the key question: who are the people voting?<\/p>\n\n\n\n
UMA\u2019s structural problem: the numbers do not add up<\/h2>\n\n\n\n
This is where the circus begins. Let\u2019s look at the numbers.<\/p>\n\n\n\n
Metric<\/strong><\/th> Value<\/strong><\/th><\/tr><\/thead> Monthly Polymarket volume<\/td> $10.5 billion (March 2026)<\/td><\/tr> UMA market cap<\/td> around $95 million<\/td><\/tr> Total staked UMA<\/td> around 20 million tokens<\/td><\/tr> Top 10 voters hold<\/td> around 30% of average voting participation<\/td><\/tr> Largest known staker<\/td> up to 7.5 million UMA (37% of total stake)<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Think about that. The platform\u2019s monthly volume is more than 100 times larger than the market cap of the project that decides all disputes on it. When the Zelenskyy suit story exploded in July 2025, one market with $242 million in volume was decided by people holding tokens worth roughly $95 million. In literal terms, it is cheaper to buy influence over voters than to lose a large market.<\/p>\n\n\n\n
In March 2025, analyst Folke Hermansen looked into vote concentration and found that two whales controlled more than 50% of active voting power in an average vote. One individual held up to 7.5 million out of 20 million staked UMA, about 37% of the entire stake. This is not decentralization. This is an oligopoly covered with nice words about \u201ctruth-seeking\u201d.<\/p>\n\n\n\n
Separately: these whales often trade on Polymarket. In other words, they take a $1 million position and then vote in UMA in favor of their own outcome. The conflict of interest is built directly into the architecture. UMA says honesty is economically more profitable because if the system is discredited, the token will fall. That is a neat theory, but it breaks in specific situations where a one-off gain on a $7 million market is larger than the expected loss from a token drop.<\/p>\n\n\n\n
And the final point: only UMA holders can vote. If you entered Polymarket, put $50,000 on a market, and are clearly right on the facts, you still have zero influence over resolution. The decision is made by completely different people, who may not trade the market at all or, more often, may be trading the opposite side. This is the core asymmetry from which the whole problem grows.<\/p>\n\n\n
[vc_message color=”warning” message_box_style=”classic” message_box_color=”alert-warning” style=”rounded”]<\/p>\n
Only holders of staked UMA tokens can vote in the UMA Data Verification Mechanism. The holder of a USDC position on Polymarket, even if that position is worth a million dollars, cannot influence the resolution. This is a fundamental asymmetry of the system.<\/p>\n
[\/vc_message]<\/p>\n\n\n
Fresh case: Russia x Ukraine ceasefire by end of 2026? ($220 million, May 2026)<\/h2>\n\n\n
\n<\/figure>\n<\/div>\n\n\nThe freshest and most revealing case. Let\u2019s break it down.<\/p>\n\n\n\n
Background.<\/strong> A series of \u00abRussia x Ukraine ceasefire by 2026\u00bb markets had been trading on Polymarket for a long time. The largest ones had deadlines of May 31, 2026 and June 30, 2026. The rules were extremely clear, and I am not paraphrasing here. This is the direct wording:<\/p>\n\n\n\n
\u00abThis market will resolve to “Yes” if there is an official ceasefire agreement, defined as a publicly announced and mutually agreed halt in military engagement, between Russia and Ukraine. Only ceasefires which constitute a general pause in the conflict will qualify. Ceasefires which only apply to energy infrastructure, the Black Sea, or other similar agreements will not qualify. Any form of informal agreement will not be considered an official ceasefire. Humanitarian pauses will not count toward the resolution of this market.\u00bb<\/p>\n\n\n\n
The key part: humanitarian pauses do not count, informal agreements do not count, and the ceasefire must be a \u201cgeneral pause in the conflict\u201d \u2014 a real halt to hostilities across the front.<\/p>\n\n\n\n
What happened.<\/strong> On May 8, 2026, Trump announced a three-day ceasefire for May 9\u201311, timed to the May 9 parade in moscow. Both sides described it as a short-term pause in air strikes. A 1,000-for-1,000 prisoner exchange was also announced. Meanwhile, active assaults by russian occupying forces continued along the line of contact, so it was not even close to a ceasefire.<\/p>\n\n\n\n
Was this an \u201cinformal agreement\u201d without an official deal? Yes. Was it a \u201cgeneral pause in the conflict\u201d? Obviously not: on May 12, one day after the \u201cceasefire\u201d ended, russia launched more than 200 drones at Ukrainian cities. The war continued without any real pause.<\/p>\n\n\n\n
![\"Volodymyr](\"https:\/\/lwallet.com.ua\/wp-content\/uploads\/2026\/05\/blog-why-polymarket-is-a-scam-29-05-2026-content-pic5-1-1024x573.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\n![\"Polymarket](\"https:\/\/lwallet.com.ua\/wp-content\/uploads\/2026\/05\/blog-why-polymarket-is-a-scam-29-05-2026-content-pic6-1024x573.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\n