Polymarket’s methods on how to grade is one of the biggest problems

When I first started covering the gambling industry in the late 1990s, I had a discussion with bet managers at some of the largest offshore sportsbooks at the time along with a bet manager at Westgate Sportsbook and one of the questions I asked is what criteria they used on deciding whether to offer a line for an event or a futures market. Each manager had the same criteria and I guess you could call it bookmaking 101.

 -- The bet has to be measurable
 -- The bet has to be unambiguous
 -- The rules for settling the bet has to be clear
 -- The bet has to have a set time limit
 -- The bet cannot be in bad taste

One of the largest sportsbooks in the world at the time had a novelty bet on whether the world would end by a certain date and while it was clearly put up in jest (I believe they offered 1,000 to 1 odds that it would happen by a certain date), they eventually took it down due to backlash about the bet itself. Obviously, the bet was only designed to generate some PR for the company, but many gambling writers and other media called it bad taste since it would be impossible for a winner to collect -- and more importantly winning the bet meant that everyone actually lost. Other examples of bets in bad taste that were sometimes offered on dark web sites included assassination bets, whether celebrities would die before the end of the year (with higher odds for younger celebrities vs. older celebrities and better odds if they didn’t die naturally) and bets on wars breaking out. Not surprisingly the owners of many of these sites avoided showing their face in public and never used their real name for fear of being arrested by the FBI.

No Bets Off-Limits for Predictions Markets 

These controversial bets seemed to die out until prediction markets were created by companies like PredictIt, Kalshi and Polymarket. It seems that no bet is off limits for these sites and the owners say they are only put up because there is demand by their users and they have no vested interest in the outcome. They are quick to point out that all bets are player-to-player and only involve markets where there is interest. One such  market that Polymarket had up was whether Venezuela would be militarily invaded by the United States by the end of January. Based on the criteria listed above under bookmaking 101, the owners of Polymarket should have known that the bet was very ambiguous and would cause confusion and frustration when trying to grade it. Yes, the bet had a clear date, but what is an invasion? Polymarket decided that what happened on January 3rd was not an invasion because under their rules an invasion required the U.S. military to "commence a military offensive intended to establish control over any portion of Venezuela." But Donald Trump and others in government said that wasn’t their goal and it was simply a law enforcement action aimed to arrest Maduro for drug trafficking. Not surprisingly, those who had market positions on "yes" for there to be a military invasion started crying foul. Polymarket said it settled the market not only because of Trump and the DOJ’s comments but also after consulting other credible sources.

Many traders are still outraged by the final outcome and say that Polymarket is rewriting its own rules and is not grading based on the intent of the market.  Unfortunately, that argument is likely to go nowhere and it is not something a court or the Commodity Futures Trading Commission (CFTC), which now regulates Polymarket, would likely be willing to take up. One Polymarket player I spoke to personally said it reminds him of the early days of offshore betting where the Gaming Commissions in many small islands did nothing to resolve disputed bets and the decision of the company was final. The only action a player had was to  find organizations like OSGA, badmouth the company on posting forums and stop playing there.

This isn’t the first instance where Polymarket had this same issue. In 2024 Polymarket put up a market on whether Israel would invade Lebanon in September of that year? Israel indeed launched a ground operation in into Lebanon on September 30th to drive Hezbollah away from the border. The rules of that market were worded as follows:

This market will resolve to "Yes" if Israel commences a military offensive intended to establish control over any portion of Lebanon between September 17, 2024 and September 30, 2024, 11:59 PM ET. Otherwise, this market will resolve to "No".

The resolution source for this market will be official confirmation by Lebanon, Israel, the United Nations, or any permanent member of the UN Security Council, however a consensus of credible reporting will also be used.

In this case there was a disagreement among Lebanon, Israel and the UN whether it was military offensive, so Polymarket went to the consensus of credible reporting, i.e. their UMA. And to say the UMA itself is controversial is an understatement. I spoke to someone very familiar with the UMA and this is how he explained it to me.

If a  market resolution is not clear and requires the UMA then a Polymarket user puts up money to propose an outcome. Other Polymarket users can dispute that proposed outcome by putting up money. Again, users can then dispute the disputed outcome by putting up money to try and get the original proposed outcome graded as the correct outcome. But again, that can be disputed by someone putting up money (or cryptocurrency).  If the same outcome gets disputed twice then it goes to the UMA to resolve before a final outcome is released.  

Clear as mud, right? Well, it gets better.

According to my UMA expert, disputes are debated on UMA’s Discord server, where users present their reasoning and evidence for their positions, but voting is only done by people on UMA that hold tokens and their vote is weighted by the amount of tokens they are willing to stake on an outcome. So, if 10 people who want it resolved as “no” put up 100 tokens each and 2 people who want it resolved as “yes” and put up 1,000 tokens then the final outcome will be "no", because more tokens are staked for that position. The fact that the vote is 5 to 1 is irrelevant, it’s not majority rules.  

So very large token holders known as UMA whales generally settle the markets that go to UMA, and the holder of one Ethereum wallet known as Borntoolate.Eth just acquired 1.3 million tokens in 2025 and while it is unclear what the total number of tokens are or how many the next highest token holder has, by all accounts the person with the 1.3 million tokens has the capacity to resolve all markets if they preferred. There is also speculation that Polymarket works with the whales to resolve markets how they want so all the discussion and evidence is pointless. Thus far, the Venezuela invasion market has not gone to the UMA.

Regardless of whether the UMA is a good resolution source or not, it violates the exact rule of bookmaking that I started bookmaking 101 with. These markets are subjective and ambiguous and hence cannot be graded with absolute certainty. With elections, the Oscars and sports bets, the outcome is clear when the final results are in. But any bet that requires an outside panel to resolve a question should never be offered. In many ways it reminds me of the  Olympic ice dancing scandals where judges from different countries worked with each other to determine who would win and who would get a medal before any dancers took to the ice. The actual performance of the dancers was completely irrelevant, which of course goes against the spirit of the Olympics. It’s why many books will only offer bets on Olympic events that have a clear winner (fastest time, highest jump etc.) and will not offer bets on sports that are judged like ice dancing, breaking, or rhythmic gymnastics.

Should these bets even be offered?

So, the question remains whether these ambiguous and even bad taste markets should be offered? Opponents say that aside from being in bad taste, there is concern that if there’s enough money (or cryptocurrency) riding on an event it could incentivize real-life actions aimed at manipulating outcomes. This is particularly concerning for assassination markets or other bets that could lead to tragic outcomes. Proponents give a different opinion. Vitalik Buterin, the co-founder of Ethereum gave this answer when asked it these invasion or similar markets should be offered:

"It's not about ‘[making] money from bad stuff happening,' it's about creating an environment where speech has consequences (so both unjustified fearmongering and unjustified complacency are punished), without relying on governmental or corporate censors."  In better words these markets are a form of free speech.

In any case proponents say that knowing whether an event has a 10% or 70% chance of happening provides valuable insight and information. They point to the 2024 election as an example where Polymarket and Kalshi both had over 60% of their market predicting  Trump would win the election even when polls were showing he would lose badly. Hence, they say when money is on the line it is far more meaningful than polls, where people lie.

Worry about insider trading 

One other concern with these types of markets is that there is inevitably insider trading. One person on Polymarket won $400,000 betting that Maduro would be removed from office before the end of January. There were other smaller bets made much earlier on the same result but this particular person made only 13 bets, totalling $33,934.34. All of them were placed between Dec. 27 and Jan. 3 and were related to the probability that the U.S. would go into Venezuela and remove Maduro from office. The obvious conclusion is that this bettor probably had inside information of when the U.S. would go into Venezuela, but there is no way to prove that and unlike the invasion market, this market only required Maduro to be removed to the U.S., so there no ambiguity over the settlement and there is no reason to go to the UMA. Was this person someone in the government (or related to someone in government) that that had inside information? Who knows. More importantly will the CFTC, which now regulates Polymarket as of November 2025, or the SEC be willing to look into it? Probably not. The person has already cashed out their winnings with Solana and since Polymarket only transacts with cryptocurrency it would be way too timely and costly to do an investigation. And no doubt if asked, the CFTC would probably say they have bigger fish to fry then chasing down $400,000 one by a degenerate gambler. But it’s also a reason we will likely never see anything like a UMA or ambiguous or bad bet markets on traditional bookmaking sites like FanDuel, DraftKings or BetFanatics on their new prediction sites.

Read insights from Hartley Henderson every week here at OSGA and check out Hartley's RUMOR MILL!