Published November 4, 2026 · Updated November 9, 2026
First published November 4, 2026 (results as of 10:31 AM ET): our estimates came within an average of 2.4 points of the 11 competitive races then counted.
In the 15 competitive Senate and governor races counted far enough to check, PharosGraph’s pre-election estimates came within an average of 2.3 points of the unofficial count as of 3:15 PM ET on November 9, 2026, and within four points in 12 of the 15. PharosGraph’s market-to-vote conversion turned Kalshi win prices, recorded on November 1, 2026, at 10:07 PM ET, before any polls closed, into vote-share estimates for 67 Senate and governor races. We published them on November 2 and locked them with a digital fingerprint (SHA-256 6d6e61cdccc0f30639dc826dcd7d40250c1946aaf8e106477c7637f2f1755648).
Our estimates against the 2026 midterm results so far
We estimated a shift toward Democrats, and the count so far shows one, larger than we estimated: a median 8.2 points of the two-party vote from the 2024 presidential vote, against our 5.7. Where we missed, we mostly missed in the same direction. Democrats ran an average of 2.1 points ahead of our estimates, and our three largest misses (Texas governor, 4.9 points; South Carolina Senate, 4.5; Louisiana Senate, 4.1) all underestimated the Democrat.
The market favorite leads in 12 of the 15. The three exceptions, the South Carolina Senate, Kansas governor and Texas governor races, were priced at 68%, 69% and 61%. Favorites at those prices lose about a third of the time or more, and across these 15 races the prices implied about four upsets.
Across all 24 competitive races, our November 1 estimate was a median shift of 5.8 points of the two-party vote toward Democrats. The races counted so far are in states that count quickly, and 13 of the 15 are in states that voted Republican for president in 2024. Still being counted: Alaska Senate, Alaska governor, Iowa Senate, Iowa governor, Maine Senate, Nevada governor, Oregon governor and Wisconsin governor. We haven’t read the count for Mississippi Senate yet. Read these figures as early.
Our estimates against Kalshi’s margin markets
Kalshi also runs margin-of-victory markets. Where they traded, the table shows Kalshi’s margin next to ours as a cross-check. Theirs is measured on the total vote and ours on the two-party vote, so the two can differ where third-party candidates drew a real share. They differed by more than five points of margin in the Nebraska governor race. Before the vote, Kalshi’s margin markets priced most of these races closer than our conversion did, about two thirds of our margin at the median, with the same leader in every race.
Against the count so far, Kalshi’s margin market came closer to the actual margin in eight of the 15 compared races where it traded, and our conversion came closer in seven. Across those races, the median shift from 2024 in the two-party vote, read on Kalshi’s margins, was 6.3 points toward Democrats; on our conversion, 5.7 points toward Democrats; in the count so far, 8.2 points toward Democrats. Kalshi’s margin is scored against the margin in the total vote, and ours against the margin in the two-party vote.
Races where one side was priced above 90%
The average covers competitive races only. Where one side traded above 90%, our conversion tops out well short of the real margin, so those races appear in the second table without an error figure. Every favorite priced above 90% that’s counted far enough to compare is leading.
In the Rhode Island governor race, an independent is in the top two, so we don’t set our estimate beside the result.
What changed since the first version
- We now compare 66 of the 124 Senate, governor and House races in our results, up from 44 on November 4.
- Florida (eight races) came off our list of states still counting mail ballots on November 5, by judgment rather than a full count of every county: every precinct was in, and the county pages we could read showed no mail ballots left, but most county pages weren’t readable. Its figures are our November 4 reading of the state’s results page. Overseas ballots are accepted until November 13, so Florida’s figures can still change.
- Some states’ figures are from earlier reads, the latest we could take: Kansas on November 4, New Hampshire on November 5 and Ohio on November 6.
- Not read yet: Kentucky Senate, Massachusetts Senate, Massachusetts governor, Mississippi Senate, West Virginia Senate, Wyoming Senate and Wyoming governor. We’ll add them when we have a count.
How we check our estimates against the count
Before the election, we tested the conversion on 43 races from the 2024 and 2025 elections that Kalshi priced between 10% and 90% three days before the vote. It landed within four points of the certified two-party result in 42 of them, with an average miss of 1.6 points. This year’s set is smaller, so it’s a check on the conversion, not a new accuracy figure for it.
Every figure here is unofficial, from counts read between November 4 at 10:19 AM and November 9 at 3:11 PM ET. We compare a race only once at least 95% of its precincts have reported, with three exceptions where states report differently:
- In Alabama, Iowa, Michigan and Tennessee, a race is compared once every county has finished.
- In New Hampshire, a race is compared once every town has finished.
- In Florida, whose precinct figures don’t measure the count, a race is compared once the count has stopped moving and no large batch of mail ballots is left to count.
Two more conditions apply everywhere. The count has to be stable, moving no more than 0.1 point against a reading at least an hour older, and the state can’t be one that counts a large share of its votes, mostly mail ballots, after Election Day. The other nine competitive races are marked “count ongoing” and stay out of the error figures until their counts finish. None of these races is certified yet. We’ll update the tables as counting continues and compute the final accuracy figure once states certify.
How close our estimates came, race by race
| Race | Price source | PharosGraph estimated Democratic share | Result | Miss | PharosGraph estimated swing | Actual swing | PharosGraph implied margin / Kalshi margin | Closer to actual margin | Market favorite and price | Status | Reporting |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Texas Governor | Kalshi | 48.7% | 53.6% | +4.9 | +5.7 | +10.6 | −2.6 / −2.4 | Kalshi | Abbott (R), 61% | leading: D | 100% |
| South Carolina Senate | Kalshi | 47.8% | 52.3% | +4.5 | +6.9 | +11.4 | −4.4 / −2.7 | Kalshi | Graham (R), 68% | leading: D | 100% |
| Louisiana Senate | Kalshi | 45.1% | 49.1% | +4.1 | +6.2 | +10.3 | −9.9 / −5.9 | Kalshi | Letlow (R), 86% | leading: R | 100% |
| Texas Senate | Kalshi | 52.6% | 56.2% | +3.6 | +9.5 | +13.1 | +5.2 / +3.5 | ours | Talarico (D), 71% | leading: D | 100% |
| Florida Senate | Kalshi | 45.0% | 48.5% | +3.5 | +1.6 | +5.2 | −10.0 / −5.4 | Kalshi | Moody (R), 86% | leading: R | no figure |
| New Hampshire Governor | Kalshi | 44.6% | 47.7% | +3.1 | −6.8 | −3.7 | −10.8 / −6.3 | Kalshi | Ayotte (R), 88% | leading: R | every town in |
| Kansas Governor | Kalshi | 47.7% | 50.3% | +2.6 | +5.9 | +8.5 | −4.6 / −3.8 | Kalshi | Masterson (R), 69% | leading: D | 100% |
| Georgia Governor | Kalshi | 52.2% | 54.3% | +2.1 | +3.3 | +5.4 | +4.4 / +2.5 | ours | Bottoms (D), 68% | leading: D | 100% |
| Nebraska Governor | Kalshi | 44.1% | 45.7% | +1.6 | +4.5 | +6.1 | −11.8 / −6.7 | Kalshi | Pillen (R), 90.0% | leading: R | 100% |
| Vermont Governor | Kalshi | 48.1% | 46.6% | −1.5 | −18.3 | −19.7 | −3.8 / −3.1 | ours | Scott (R), 66% | leading: R | 100% |
| Kansas Senate | Kalshi | 51.1% | 52.5% | +1.4 | +9.3 | +10.7 | +2.3 / +2.0 | ours | Hamilton (D), 60% | leading: D | 100% |
| Florida Governor | Kalshi | 48.7% | 49.3% | +0.7 | +5.3 | +6.0 | −2.7 / −2.3 | Kalshi | Donalds (R), 61% | leading: R | no figure |
| Michigan Senate | Kalshi | 53.2% | 53.6% | +0.4 | +4.0 | +4.3 | +6.5 / +3.3 | ours | El-Sayed (D), 76% | leading: D | every county in |
| Ohio Governor | Kalshi | 53.9% | 54.1% | +0.2 | +9.6 | +9.8 | +7.8 / +4.8 | ours | Acton (D), 80% | leading: D | 100% |
| Ohio Senate | Kalshi | 52.4% | 52.5% | +0.1 | +8.1 | +8.2 | +4.8 / +3.9 | ours | Brown (D), 70% | leading: D | 100% |
| Alaska Governor | Kalshi | 55.0% | +11.9 | +10.1 / +8.0 | Kreiss-Tomkins (D), 86% | count ongoing | no results yet | ||||
| Alaska Senate | Kalshi | 52.5% | +9.4 | +5.1 / +2.6 | Peltola (D), 71% | count ongoing | no results yet | ||||
| Iowa Governor | Kalshi | 54.8% | +11.5 | +9.6 / +4.9 | Sand (D), 85% | count ongoing | counties still counting | ||||
| Iowa Senate | Kalshi | 49.7% | +6.4 | −0.6 / −0.6 | Hinson (R), 52% | count ongoing | counties still counting | ||||
| Maine Senate | Kalshi | 52.7% | −0.9 | +5.4 / +3.3 | Jackson (D), 72% | count ongoing, official county results expected after the canvass | no results yet | ||||
| Mississippi Senate | Kalshi | 45.0% | +6.6 | −10.0 / −7.3 | Hyde-Smith (R), 86% | count ongoing | results not yet read | ||||
| Nevada Governor | Kalshi | 53.0% | +4.6 | +6.0 / +3.6 | Ford (D), 74% | count ongoing | no results yet | ||||
| Oregon Governor | Kalshi | 53.9% | −3.5 | +7.8 / +4.9 | Kotek (D), 80% | count ongoing | no results yet | ||||
| Wisconsin Governor | Kalshi | 54.2% | +4.7 | +8.5 / +5.0 | Crowley (D), 82% | count ongoing, official county results expected after the canvass | no results yet |
Compared races first, by size of miss; “count ongoing” races follow. Shares and swings are in points of the two-party vote, and margins in points of margin; positive is toward Democrats. Kalshi’s margin is on the total vote and appears only where its margin market traded; it’s scored against the total-vote margin and ours against the two-party margin.
Races not in contention
| Race | Price source | Favorite and price | Status | Democratic share | Reporting |
|---|---|---|---|---|---|
| Alabama Governor | Kalshi | Tuberville (R), 94% | leading: R | 48.0% | every county in |
| Arkansas Governor | Kalshi | Sanders (R), 97% | leading: R | 41.7% | 100% |
| Arizona Governor | Kalshi | Hobbs (D), 91% | count ongoing | 99% | |
| California Governor | Kalshi | Becerra (D), 97% | count ongoing | 100% | |
| Colorado Governor | Kalshi | Weiser (D), 95% | count ongoing | 17% | |
| Connecticut Governor | Kalshi | Lamont (D), 96% | leading: D | 59.6% | 100% |
| Hawaii Governor | Kalshi | Green (D), 98% | count ongoing | 50% | |
| Idaho Governor | Kalshi | Little (R), 98% | leading: R | 36.2% | 100% |
| Illinois Governor | Kalshi | Pritzker (D), 97% | count ongoing | no results yet | |
| Massachusetts Governor | Kalshi | Healey (D), 97% | count ongoing | results not yet read | |
| Maryland Governor | Kalshi | Moore (D), 99% | count ongoing | 100% | |
| Maine Governor | Kalshi | Pingree (D), 96% | count ongoing, official county results expected after the canvass | no results yet | |
| Michigan Governor | Kalshi | Benson (D), 93% | leading: D | 55.5% | every county in |
| Minnesota Governor | Kalshi | Klobuchar (D), 97% | leading: D | 56.5% | 100% |
| New Mexico Governor | Kalshi | Haaland (D), 94% | leading: D | 57.4% | 100% |
| New York Governor | Kalshi | Hochul (D), 96% | count ongoing | 93% | |
| Oklahoma Governor | Kalshi | Mazzei (R), 91% | leading: R | 46.9% | 100% |
| Pennsylvania Governor | Kalshi | Shapiro (D), 98% | count ongoing | 99% | |
| Rhode Island Governor | Kalshi | Foulkes (D), 99% | independent in top two | 100% | |
| South Carolina Governor | Kalshi | Wilson (R), 96% | leading: R | 47.7% | 100% |
| South Dakota Governor | Kalshi | Rhoden (R), 97% | leading: R | 36.5% | 100% |
| Tennessee Governor | Kalshi | Blackburn (R), 94% | leading: R | 48.5% | every county in |
| Wyoming Governor | Kalshi | Barlow (R), 99% | count ongoing | results not yet read | |
| Alabama Senate | Kalshi | Moore (R), 99% | leading: R | 45.2% | every county in |
| Arkansas Senate | Kalshi | Cotton (R), 96% | leading: R | 43.9% | 100% |
| Colorado Senate | Kalshi | Hickenlooper (D), 99% | count ongoing | 17% | |
| Delaware Senate | Kalshi | Coons (D), over 99% | leading: D | 63.6% | 100% |
| Georgia Senate | Kalshi | Ossoff (D), 97% | leading: D | 59.0% | 100% |
| Illinois Senate | Kalshi | Stratton (D), 98% | count ongoing | no results yet | |
| Kentucky Senate | Kalshi | Barr (R), 96% | count ongoing | results not yet read | |
| Massachusetts Senate | Kalshi | Markey (D), 98% | count ongoing | results not yet read | |
| Minnesota Senate | Kalshi | Flanagan (D), 91% | leading: D | 55.2% | 100% |
| North Carolina Senate | Kalshi | Cooper (D), 97% | leading: D | 60.0% | 100% |
| New Hampshire Senate | Kalshi | Pappas (D), 92% | leading: D | 57.0% | every town in |
| New Jersey Senate | Kalshi | Booker (D), 98% | count ongoing | no results yet | |
| New Mexico Senate | Kalshi | Luján (D), 98% | leading: D | 63.3% | 100% |
| Oklahoma Senate | Kalshi | Hern (R), 97% | leading: R | 40.7% | 100% |
| Oregon Senate | Kalshi | Merkley (D), 98% | count ongoing | no results yet | |
| Rhode Island Senate | Kalshi | Reed (D), over 99% | leading: D | 69.5% | 100% |
| Tennessee Senate | Kalshi | Hagerty (R), 97% | leading: R | 43.8% | every county in |
| Virginia Senate | Kalshi | Warner (D), 98% | leading: D | 61.5% | 100% |
| West Virginia Senate | Kalshi | Capito (R), 98% | count ongoing | results not yet read | |
| Wyoming Senate | Kalshi | Hageman (R), 99% | count ongoing | results not yet read |
No error figure for these races.
House races in the target states
| Race | Price source | PharosGraph estimated Democratic share | Result | Miss | Market favorite and price | Status | Reporting |
|---|---|---|---|---|---|---|---|
| AZ-01 | Kalshi | 54.0% | Shah (D), 81% | count ongoing | 98% | ||
| AZ-04 | Kalshi | 57.9% | Stanton (D), 96% | count ongoing | 99% | ||
| AZ-06 | Kalshi | 55.6% | Mendoza (D), 89% | count ongoing | 100% | ||
| CA-01 | Kalshi | 57.6% | McGuire (D), 95% | count ongoing | 100% | ||
| CA-09 | Kalshi | 60.7% | Harder (D), 99% | count ongoing | 100% | ||
| CA-13 | Kalshi | 57.4% | Gray (D), 95% | count ongoing | 100% | ||
| CA-21 | Kalshi | 57.5% | Costa (D), 95% | count ongoing | 100% | ||
| CA-22 | Kalshi | 56.2% | Villegas (D), 91% | count ongoing | 100% | ||
| CA-26 | Polymarket | 58.2% | Irwin (D), 96% | count ongoing | 100% | ||
| CA-27 | Kalshi | 60.5% | Whitesides (D), 99% | count ongoing | 100% | ||
| CA-41 | Kalshi | 60.2% | Sanchez (D), 99% | count ongoing | 100% | ||
| CA-45 | Kalshi | 56.5% | Tran (D), 92% | count ongoing | 100% | ||
| CA-47 | Kalshi | 58.1% | Min (D), 96% | count ongoing | 100% | ||
| CA-48 | Polymarket | 57.9% | Wilpert (D), 96% | count ongoing | 100% | ||
| CA-49 | Kalshi | 58.4% | Levin (D), 97% | count ongoing | 100% | ||
| FL-02 | Kalshi | 43.6% | 47.9% | Rogers (R), 92% | leading: R | no figure | |
| FL-13 | Kalshi | 48.3% | 49.6% | +1.3 | Luna (R), 64% | leading: R | no figure |
| FL-14 | Kalshi | 53.0% | 52.7% | −0.3 | Castor (D), 74% | leading: D | no figure |
| FL-22 | Kalshi | 52.5% | 51.9% | −0.7 | Dandiya (D), 71% | leading: D | no figure |
| FL-23 | Kalshi | 60.5% | 65.6% | Frankel (D), 99% | leading: D | no figure | |
| FL-25 | Kalshi | 53.9% | 52.4% | −1.5 | Moskowitz (D), 80% | leading: D | no figure |
| GA-04 | Polymarket | 60.2% | 82.6% | Johnson (D), 99% | leading: D | 100% | |
| GA-05 | Polymarket | 63.6% | 90.2% | Williams (D), over 99% | leading: D | 100% | |
| GA-06 | Polymarket | 60.0% | 80.9% | McBath (D), 99% | leading: D | 100% | |
| GA-07 | Polymarket | 43.0% | 43.8% | McCormick (R), 94% | leading: R | 100% | |
| GA-11 | Polymarket | 42.1% | 44.6% | Cowan (R), 96% | leading: R | 100% | |
| GA-13 | Kalshi | 59.0% | 79.8% | Clark (D), 97% | leading: D | 100% | |
| ME-02 | Kalshi | 49.3% | LePage (R), 56% | count ongoing, official county results expected after the canvass | no results yet | ||
| MI-01 | Kalshi | 44.8% | 45.1% | +0.4 | Bergman (R), 87% | leading: R | every county in |
| MI-04 | Kalshi | 51.3% | 50.9% | −0.4 | McCann (D), 61% | leading: D | every county in |
| MI-07 | Kalshi | 53.5% | 51.7% | −1.8 | Lawrence (D), 78% | leading: D | every county in |
| MI-08 | Kalshi | 58.4% | 57.0% | Rivet (D), 97% | leading: D | every county in | |
| MI-10 | Kalshi | 53.1% | 53.7% | +0.6 | Hines (D), 75% | leading: D | every county in |
| MI-11 | Kalshi | 59.0% | 62.3% | Moss (D), 97% | leading: D | every county in | |
| NC-01 | Kalshi | 55.2% | 51.7% | −3.5 | Davis (D), 87% | leading: D | 100% |
| NC-11 | Kalshi | 54.8% | 58.0% | +3.2 | Ager (D), 85% | leading: D | 100% |
| NE-02 | Kalshi | 56.4% | Powell (D), 92% | count ongoing | 100% | ||
| NY-01 | Polymarket | 46.0% | 48.9% | +2.8 | LaLota (R), 80% | leading: R | 100% |
| NY-03 | Kalshi | 56.5% | 55.2% | Suozzi (D), 92% | leading: D | 96% | |
| NY-04 | Kalshi | 56.5% | Gillen (D), 92% | count ongoing | 91% | ||
| NY-08 | Kalshi | 59.4% | 80.0% | Jeffries (D), 98% | leading: D | 100% | |
| NY-12 | Polymarket | 59.1% | 79.4% | Lasher (D), 98% | leading: D | 100% | |
| NY-15 | Polymarket | 59.0% | 87.8% | Torres (D), 97% | leading: D | 100% | |
| NY-17 | Kalshi | 51.4% | Conley (D), 62% | count ongoing | 60% | ||
| NY-19 | Kalshi | 57.9% | 57.3% | Riley (D), 96% | leading: D | 100% | |
| NY-21 | Kalshi | 48.1% | 49.1% | +0.9 | Constantino (R), 66% | leading: R | 100% |
| PA-01 | Kalshi | 50.5% | Harvie (D), 54% | count ongoing | no results yet | ||
| PA-07 | Kalshi | 55.0% | Brooks (D), 86% | count ongoing | no results yet | ||
| PA-08 | Kalshi | 55.0% | Cognetti (D), 86% | count ongoing | no results yet | ||
| PA-10 | Kalshi | 54.1% | Stelson (D), 81% | count ongoing | no results yet | ||
| PA-17 | Kalshi | 58.1% | Deluzio (D), 96% | count ongoing | no results yet |
51 races. 10 are priced on Polymarket only. These prices weren’t back-tested. Races priced above 90% show the result without a miss, because our conversion tops out well short of a lopsided margin. House markets trade less than the Senate and presidential markets the back-test is built on, so these races aren’t in the average. The rest of the House follows in a later update as results firm up.
Footnotes
- Results are unofficial, from state reporting read between November 4 at 10:19 AM and November 9 at 3:11 PM ET.
- Prices are Kalshi’s where Kalshi had a party-level market and Polymarket’s otherwise; the source column shows which. The back-test uses Kalshi prices; Polymarket prices weren’t back-tested. The back-test’s one Alaska race was scored on first-round counts.
- CA-06 and CA-14 are in our results but had no frozen market price, so they aren’t shown.
- New England states report by town; shown here as counties.
- The 2024 baseline is the presidential Democratic share of the two-party vote in each state, statewide for Alaska.
- Not shown: the Idaho, Montana, Nebraska and South Dakota Senate races, where the markets don’t price a Democrat-against-Republican matchup. In Nebraska, the challenger, Dan Osborn, is an independent.
- Alaska’s governor and Senate races use ranked-choice voting. Our estimate is for the two-candidate final round. Races without a majority are tabulated on November 18, so Alaska stays “count ongoing” until then.
- Competitive means neither side above 90% on the two-party price. Prices shown are each market’s price on a two-party basis, not the raw market price.
Where the 2026 swing from 2024 happened, county by county: which counties swung most in the 2026 midterms.
Follow the count race by race on the PharosGraph Atlas: maps.pharosgraph.com/atlas/2026