PLATNER MULTIVERSE

The hidden math inside every Polymarket contract

Polymarket and Werner K. Zagrebbi

Jul 23, 2026

🇺🇸 Midterms are LIVE on Polymarket USA.

New users get a $20 bonus with $10 deposit.

Download now with code “ORACLE” to claim your bonus.

Werner K. Zagrebbi is a board-certified Superforecaster™ who writes about politics and economics at Right Rationalism.

Graham Platner dropped out of the Maine Senate race on July 10th.

This probably isn’t news to you. You’ve surely tired of the panoply of takes that have been bandied about after his withdrawal: Was that symbol on his pectoral associated with Imperial Germany? Is it cringe to be on Kik after the age of 25?

But there is an important question left unanswered by all these takes: was this a good decision for his party?And after this kind of scandal hit their nominee, wouldn’t you expect the Democrats to be less likely to win in Maine? I certainly would. And yet, on the same news, the Democrats’ chance of winning that Senate seat went up, from 54% to 63% in a couple hours - days before Platner bowed out of the race.

THIS is the real Platner story!

Platner vs. Arrow

GMU’s Tyler Cowen went so far as to rank Kenneth Arrow as the fourth greatest modern economist, after Hayek, Keynes, and Milton Friedman. Arrow’s most important contribution was to “build up economics from a theory of ‘state prices,’” according to Cowen.

What does this mean? We can think of every possible world as a separate “state.” A state price is what you would pay today for a dollar that arrives in exactly one of those states and no other. Naturally, the more likely you consider that state, the more you will pay for a dollar that “lives there.” Thus each possible future carries its own price, and those prices behave like probabilities.

This kind of thinking may come naturally to Polymarketeers, but its implications are not necessarily intuitive.

Consider Sergeant Platner. The price of “Democrats win the seat” is essentially a weighted average of two possible worlds: Democrats’ odds of winning the Maine Senate seat if Platner goes, and their odds of doing so if he stays, each weighted conditional on the likelihood of it occurring. The Platner accusations, in this reading, is an “exogenous shock” that moved the weights. Said Alex Tabarrok of Marginal Revolution:

The market is signaling that Platner reduces the Democrats’ chances of victory. We can be more precise. If an 87-point increase in the probability of dropping out gets you 9 points of winning, then a 100% chance of dropping out implies a gain of 9/0.87 ≈ 10.3 percentage points. Thus the market’s best estimate is that Platner is reducing the Democrats’ chance of winning by about 10 percentage points (compared to an unknown replacement). That’s a pretty big number!

It was a pretty big number! And thus Platner was probably right to drop out: That decision, at the time, made the universe where a Democrat is Maine’s senator 10% more likely. The scandal hurt Democrats, but Platner’s withdrawal helped them more.

Share The Oracle by Polymarket

Watch The Unlisted Market

The lesson for traders here is to keep an eye out for the unlisted market. A visible contract is often but one leg of a hidden state-price bundle, and when news changes a price, adjacent contracts don’t always update in the same order. Perhaps write out the synthetic before you trade:

In this case, that would be Democrats’ overall chance of winning = (Their chance of winning if Platner leaves) × (The chance that he leaves) + (Their chance of winning if Platner stays) × (The chance that he stays).

Endeavor to ask, “Have the displayed prices caught up with the implied conditional prices?”

Every Dem name in the race to succeed Platner as Maine Senate nominee has a live probability of being nominated on Polymarket. Their sum is a synthetic “Democratic nominee is X” bundle. If it prices at less than the “Dems win” market divided by a plausible general-election win rate, someone is mispriced.

You can use this trick elsewhere. JD Vance’s chance of being out as VP by December 31, 2026 is about 7%, for instance, and his chance of winning the Republican nomination in 2028 is 41%, so his implied nomination probability conditional on remaining VP is 0.41 / 0.93 ≈ 44%. Rubio is at 28% in the same market, and if the VP-exit market rises by 20 points on some news, but Rubio doesn’t rise 8 points, there is an arbitrage.

The total-cuts-in-2026 polymarket prices the odds of no interest rate cuts at 81%, 1 cut at 15%, and 2 or more at 4%, for expected cuts of about 0.25; the Jul–Oct decision-path market prices Pause-Pause-Pause at 47%, and the two describe the same underlying object, so summing the path probabilities should recover the total-cuts distribution.

Zohran Mamdani out as mayor before 2027 is 5.5%, and his rent-freeze market is 92%, so the rent-freeze conditional on him remaining in office is about 97%. Every one of his policy contracts (millionaire tax, $30 wage, city grocery store) is a compound bet of P(Mamdani still in office) times P(policy passes given he is), so a shock to the tenure market should reduce every policy market by the same factor.

Or consider Israel. The Likud party is priced at 63.5% to win the next Knesset election, while Netanyahu himself is at 34.5% to be the next Prime Minister. But if Likud wins, its leader is who forms a coalition, and the two markets describe the same underlying object. Divide one by the other and you get the market’s implied probability that Netanyahu is the one still leading Likud when the votes are counted: 0.345 / 0.635 ≈ 54%. No Polymarket contract asks that question directly, but the two that do trade can together get you to a price. On any news that resolves it (a corruption verdict, a coalition fight, a health scare), both markets will move together.

You should also take advantage of spanned information in other ways when you can. If you’re a Democrat who wants to know how to donate to races in which it will be maximally effective, Polymarket is what you should be watching.

It’s easy to underestimate the beauty and promise of this idea: Conditional markets offer rigorous evaluation before policies are even enacted. Robin Hanson might go as far as suggesting we use them to decide what the government does, but there are a whole host of other use cases short of that: Central bankers already closely watch market-implied expectations through things like CME FedWatch, and people have proposed going further with NGDP prediction markets or decision markets for monetary policy. And research funders could look to markets for scientific replicability, or to more exotic experiments like Scott Alexander’s impact markets.

Certainly once we know what we want, we should more often ask the Arrow question: What is the market price of the world where such a thing happens, compared to the world where it doesn’t?

Discussion about this post

Ready for more?