2026 Election Night Odds: 8 p.m. Decides 75% of It

Published 2 November 2026 · Prices frozen 2 November 2026, 17:30 UTC

You can find a poll-closing table tonight and you can find 2026 election night odds, and not one page does both. That is a problem, because the two halves answer different questions and the one you actually want — which close decides the Senate — needs them joined.

Here is what joining them gets you. Five of the seven Senate toss-ups close at 8 p.m. ET, and in 75.08% of the market’s own probability mass those five settle the chamber by themselves. The 7:30 p.m. group settles nothing at all. And 12.04% of the Senate sits on Alaska, which closes at midnight Eastern and takes days to count.

We pulled all 128 legs of Polymarket’s seven-state board, both chamber-control markets, the Balance of Power board, the seat ladder, eight single-state boards and thirteen Kalshi contracts in one pass, normalised each book against its own sum, took the seat arithmetic from the published control rule rather than a headline seat count, and ran the threshold through the joint distribution against the Eastern clock. Prices are frozen at 2 November 2026, 17:30 UTC; the live boards are linked throughout.

Where the 2026 election night odds stand

The House stopped being a question some time ago. Polymarket pays 94.5¢ on a Democratic House and Kalshi 93.9/94.0¢, on a board that has traded $38.39m. The Senate is the night’s only live chamber, and it has been moving: 66.5¢ on 25 October, 73.5¢ tonight.

The Balance of Power board is the biggest market of the midterms at $53.57m and it prices the two chambers jointly. Democrats taking both pays 72.5¢. A split with Republicans holding the Senate pays 21.5¢. A Republican sweep pays 4.5¢. And the fourth cell — Democrats holding the Senate while losing the House — pays 0.65¢.

That 0.65¢ is the most informative number on the board. It says the Senate is strictly harder than the House in every world traders can imagine — no path where Democrats win the hard chamber and lose the easy one. Which is why the rest of this piece is about seven Senate races.

The election night clock, race by race

Eight races matter: the seven that Polymarket’s own seven-state board names, plus North Carolina. Here they are in the order their polls close, with what the two exchanges pay on the Republican.

Two things jump out. The clock is front-loaded: six of the eight races are shut by 8 p.m. ET, and the most important remaining unknown is in Alaska, four hours later. And Kalshi is dearer on the Republican in six of the eight, by one to two points. The exceptions are Kansas and Iowa — the only two races where Polymarket is the more Republican venue.

Note which seat belongs to whom. Six of the seven toss-ups are Republican-held. Michigan, where Gary Peters is retiring, is the one Democratic seat, so a Rogers win there is a Republican gain, not a hold. Getting that backwards inverts the whole seat model.

Three of seven: the only number you need

Republicans hold 53 seats against 45 Democrats and two independents who caucus with them. Polymarket’s control rule is explicit that a party needs a majority of voting seats, and that a 50-50 chamber goes to the party holding the vice presidency — which is the Republicans. So Democrats need 51, not 50, and that single clause is what makes tonight’s threshold three rather than four.

North Carolina is an open Republican seat and Cooper trades at 97.9¢. Grant that, and Republicans start the seven on 52. Each of the six Republican-held toss-ups they lose costs a seat; Michigan, if Rogers takes it, hands one back.

Add the top four rows and the board pays 77.76% on a Democratic Senate. The control board itself pays 73.5¢, and the eleven-rung seat ladder reads 76.74% once normalised. Three routes, four and a quarter points apart, which is a tighter cluster than the fourteen points these same books were showing on Thursday. The board has grown up: it has traded $34.68m in a week, and all 128 legs now carry a two-sided quote where 58 of them had no bid at all three days ago.

What each poll close actually settles

Now put the threshold against the clock. At each point in the evening, ask a narrow question: of the races that have closed, do their results alone fix whether Republicans end up above or below four, whatever the rest do? Where the answer is yes, the Senate is decided. Where it is no, you are still waiting.

The 0.00% at 7:30 is not a rounding artefact. With one of the seven in and six still out, Republicans’ total can land anywhere in a seven-wide range, which always straddles the line. No result in Ohio, of any size, can settle the Senate at 7:30 p.m. That is worth knowing before you spend the first half hour of the night watching Franklin County.

Then 8 p.m. does three quarters of the work in a single minute. The structure underneath it is clean: if Republicans win none or one of the five that have closed, they cannot reach four and the Democrats have the chamber — that is 65.06%. If they win four or five of the five, they are there already — 10.02%. Two or three of five leaves it open, and those two cases are very different from each other: at two of five the Senate is still 77.3% Democratic, at three of five it is 17.2%.

Iowa at 9 p.m. then clears another 12.9 points, and what is left is the one branch nobody wants: the first six split three-three, the chamber hangs on Alaska, and Alaska closes at midnight Eastern. The board puts that at 12.04%, with the Democrat 61.7% in it.

The closest race is the least decisive one

There is a neat identity hiding in the threshold. A race decides the Senate exactly when flipping it flips the chamber — and because the line sits at three versus four, that happens if and only if the other six split three-three. Nothing about the race’s own price enters it. Pivotality is a statement about the company a race keeps.

So Iowa — the one genuine coin flip on the board at 50.5¢ — is the least decisive of the seven, and Texas at 24.5¢ is the most. The intuition that the closest race is the one to watch is simply wrong here, and the reason is correlation. Iowa’s price is near 50 because it sits in the middle of the national distribution, which means the worlds where Iowa is close are disproportionately worlds where the other six have already broken one way together.

Texas and Ohio are the opposite case: both are priced as likely Democratic pickups, so the worlds in which they are live are the worlds in which everything is live. It is also why the spread across the seven is narrow, 10.60% to 14.05%. No single race here is the tipping point; seven races are each about an eighth of one.

What one result does to the Senate price

Pivotality says how often a race is the margin. It does not say how much a result moves the price, which is the number you want if you are watching a board rather than a map. For that, condition the joint distribution on each race and re-read the chamber.

The ordering is the same as pivotality, which is reassuring, and the magnitudes are the practical point: a called Texas moves the Senate price by nearly 60 points in whichever direction it lands. An early Paxton win at 8 p.m. takes the chamber from 77.76% Democratic to 32.95%. The same call for the Democrat takes it to 91.32%.

Iowa is at the other end for an understandable reason: at 50.5¢ the market has priced both branches in, so neither outcome is news — a Hinson win still leaves the Senate 61.70% Democratic. If you have one screen tonight, put Texas on it. It closes at 8 p.m. and carries the biggest swing of the seven.

North Carolina at 7:30, and the 2¢ that resets it

Everything above holds North Carolina at Cooper, because the market does — 97.9¢, against 2.05¢ for Whatley. But North Carolina closes at 7:30 p.m. ET alongside Ohio, and it is the one 7:30 result that would change the shape of the night rather than a line in it.

If Whatley wins, Republicans start the seven on 53 rather than 52 and the threshold moves with them: Democrats would then need Republicans held to two or fewer of the seven, not three. Run the same 128 legs against the tighter line and Democratic control falls from 77.76% to 64.52%.

So the honest way to read a Whatley win is not “minus one seat” but “minus thirteen points, instantly, at 7:31 p.m.” It is a 2.05¢ event. It is also the only thing in the first half-hour of the night with that much information in it.

The market will not settle when the networks call it

This is the part that catches people every cycle, and it is in the published rules of both venues rather than anyone’s reading of them.

Polymarket’s seven-state board resolves only once all three of the Associated Press, Fox News and NBC have conclusively called the races. One network calling Texas does nothing. The same text counts a state as not won by the Republican if its result “cannot be determined by March 31, 2027”, and excludes independents from the party legs whatever their affiliation.

Kalshi resolves on a different event altogether. Its state contracts pay if “a representative of the Republican party is sworn in as a Senator” for the term beginning in 2027, and they close on 3 November 2027 — a year after the election. Its chamber contracts run to 1 February 2027, because the fallback is the party of the President pro tempore.

Eight days of prices, and what would surprise us

The last week has been one-directional: the chamber boards moved together while the seven moved a long way further.

Kansas is the story of the week: Roger Marshall has shed 24 points in eight days, in a state that has not sent a Democrat to the Senate since 1932. Alaska is the only one of the seven to have moved the Republicans’ way, by three points — a small thing to be carrying 12.04% of the chamber.

What would surprise us tonight? A Whatley win in North Carolina, for the thirteen points it takes off the Senate at 7:31 p.m. A Husted hold in Ohio, which the board has at 22.37% and which drops the chamber to 32.80%. And most of all, four of the five 8 p.m. states going Republican: the board gives that 10.02%, and it ends the night at one minute past eight.

Related articles

• 2026 Senate swing state odds — the week Kansas broke
• Polymarket vs Kalshi 2026 midterms — two rulebooks, one price
• 2026 House odds: markets 90¢, models 76-93%
• 2026 governor race odds: 36 boards vs one ladder
• 2026 Senate odds and oil — five conditional boards
• 2026 midterms hub — every Senate and governor race we track

Sources: the Polymarket Gamma and CLOB public APIs (events 32224, 32225, 32228, 106193, 1262243 and the eight state boards 59233, 57647, 57648, 57650, 57658, 57682, 57672 and 57664) and the Kalshi public markets API (CONTROLS-2026, CONTROLH-2026, SENATEAK/IA/KS/ME/MI/TX/NC-26 and SENATEOHS-26), all read at 2 November 2026, 17:30 UTC. Poll-closing times, including the split-time-zone caveats, are Election Central’s 2026 table; for a specific polling place the state’s own election office is the authority. The resolution terms are quoted from the seven-state board itself and from the SENATETX-26-R contract, in both cases the operative text the venue publishes rather than a summary. The 53-45-2 going-in split, the open seats and the nominees are as described here, and the seat map is 270toWin’s. Kalshi volume and open interest are contract counts; Polymarket volume and liquidity are dollars, and the two are never added on this page. The settlement percentages are arithmetic on the joint board, not a prediction of when any race will be called. Prices here are a frozen snapshot; the live numbers are on the market pages, which refresh every 10 minutes. Nothing here is financial or betting advice.