United Russia at 74% and 99.1% on the same day, one word apart

On September 9, 2026, Polymarket priced United Russia at 74% to gain the most seats in Russia's parliamentary election and at 99.1% to win the most seats. Same party, same election, same settlement date, 25 points apart. Nobody doubted the result. The gap was the price of one verb, and the market spent two weeks arguing about what it meant.

What the market said

The September 1 issue had United Russia at 71% (+7) to take the most Duma seats, on $235K traded, with the market settling September 29 and no verified catalyst behind the move. By the September 8 issue it was 73% (+4) on $191K of daily volume against $398K of liquidity, 21 days from resolution.

That issue's Futures column made the structural case for a higher price: a 27% chance that Putin's own party fails to finish first sat next to a Putin-out-by-December contract at 7%, and the two did not fit. The column put fair at 88% to 93%. It also flagged where the missing points might live: "the contract wording decides what 'most seats' means," and "watch the market page's resolution notes."

The next morning answered it. The September 9 issue reported the contract asking whether United Russia "gains" the most seats at 74% (+6), and a sibling asking whether it "wins" the most seats at 99.1%. Its read: "That 25-point gap comes from the wording, not from any doubt about the result. Nobody expects the ruling party to finish anywhere but first; the money is arguing over what 'gain' means in the rules."

The gain contract then drifted rather than converged: 77% (−2) on September 15, 75% (−1) on September 16 with the New People party at 21.2% (+1.5), and 75% again on September 17, when United Russia's leg alone turned over $897K in a day. The full event that morning read United Russia 75%, New People 22% (+1.2), the Communist Party 1.2%, and A Just Russia 0.2%, resolving September 29.

What happened

As of September 17 the vote has not been held and neither contract has settled. What happened is the market's own behaviour. A 99.1% contract on "wins most seats" says traders treat a United Russia plurality as a formality. A 22% contract on New People "gaining" the most seats cannot mean traders think New People will finish first; the sibling market rules that out. It has to mean that a meaningful share of the money reads "gain" as a net change in seats, where a small party adding a few dozen can outgain a ruling party that already holds most of the chamber and has more room to lose seats than to add them.

Under that reading, 75% against 22% is a real question about seat arithmetic. Under the plurality reading, the gain contract is a 99% event trading at 75%. The market did not resolve which reading applies; it priced both and let the gap sit. How settlement works calls this the space between the title and the rule, and here it was worth 25 points.

A worked example

Take 100 YES contracts on each sibling at the September 17 prices. Polymarket has historically charged no trading fee on most markets, as of September 2026, so no fee line appears; check the venue's own page before relying on it.

Contract	Price	Cost of 100 YES	Pays if plurality reading holds	Pays if net-gain reading holds and New People outgains
United Russia wins most seats	99.1%	$99.10	$100.00, profit $0.90	$100.00, profit $0.90
United Russia gains most seats	75%	$75.00	$100.00, profit $25.00	$0, loss $75.00
New People gains most seats	22%	$22.00	$0, loss $22.00	$100.00, profit $78.00

The event's four legs sum to 75 + 22 + 1.2 + 0.2 = 98.4%, so the market is pricing a near-complete set of outcomes under one consistent reading, not two markets talking past each other. The question is which reading. A trader who buys the 75% leg thinking it is the 99.1% leg at a discount is paying $75 to collect $25 on the assumption that "gain" means "win." If the rule says net gain, that same $75 is a bet on seat counts that the sibling market says are a genuine contest. The how to read prediction market odds guide covers why sibling legs should add up; this family adds up, which is the tell that the wording is doing the work.

The fine print that mattered

The editions do not carry the contract's rule text, and this page will not invent it. What the editions carry is enough to know where to look. Polymarket contracts settle through a proposed-then-finalized resolution process, and a disputed word like "gain" is exactly what that process exists to argue over. The Polymarket guide explains the mechanics, and the September 8 issue's instruction stands: read the market page's resolution notes before the price.

Election contracts on our own trackers, such as House 2026 and Senate 2026, show the rule text verbatim for this reason. The election markets guide walks through how "control," "most seats," and "vote share" each settle on different numbers from the same night.

The lesson

When two contracts on the same event disagree by 25 points and both are liquid, the first suspect is the wording, not the event. The September 8 column reached for a structural argument about Putin's grip on the Duma. The September 9 issue found a simpler answer in a verb.

Second, a sibling family that sums to roughly 100% is evidence about the rule. If "gain" meant "win," the New People leg would trade near zero and the United Russia leg near its sibling's 99.1%. The fact that New People holds 22% tells you how the money reads the question, before any resolution note does.

What it settled on

Neither contract had settled as of the September 17 issue. The wins-most-seats contract at 99.1% will settle on the official seat count, and the gains-most-seats contract will settle on whatever its rule text says "gain" means, read against the same count. The catalyst calendar carries both contracts' resolution date. The venue's settled record is the only place the answer will appear; a 99.1% price is a forecast of it, not a copy.

Nothing on this page is investment advice; the venue's current rule text and fee page are the only things that pay.