Maine Senate 2026: BUY NO at 30 Cents
· By Tyler James Webber
Kalshi prices Democrats to win the Maine Senate race at 71% YES. A five-factor model — polling baseline, Collins overperformance discount, midterm wave, Platner biographical risk, and ranked-choice redistribution — sets fair value at 58% YES. BUY NO at 30 cents, edge +13pp, sized small (25%).
Kalshi "Maine Senate winner?" contract, Dec 2024 – May 2026. Democrats 70%, Republicans 30%, \$554k volume.
Position. BUY NO on Democrats win Maine Senate race at 30 cents.
Fair Value. 58% YES (range 24% to 94%).
Edge. +13 percentage points on the NO side.
Conviction. 5/10. Data quality 7/10, model confidence 6/10, completeness 7/10, edge robustness 2/10.
Key Date. November 3, 2026 (election day).
Platform. Kalshi. Volume \$553,028, open interest 126,886 contracts.
Sizing. Small (25% of full size). An edge robustness of 2/10 fails the first standard adverse perturbation on the polling factor, so the position is sized down.
Market Question. Will Democrats win the Maine Senate race?
Event Overview. Five-term Republican incumbent Susan Collins, seeking a sixth term, faces Democratic challenger Graham Platner, a Marine and Army veteran turned oyster farmer. The Democratic field cleared in April when Governor Janet Mills suspended her primary campaign, citing fundraising shortfalls, leaving Platner as the presumptive nominee.
Maine uses ranked-choice voting for federal contests, so the race resolves on the eventual winner regardless of the round in which they clear 50% of votes. Maine is one of three Republican-held Senate seats in states Kamala Harris carried in 2024 (Harris +7) and remains the single most quoted seat in any Democratic path to a Senate majority in the 2026 midterms. Collins is also the Senate Appropriations Committee chair, which her campaign is using as a defensive argument about the continuation of federal dollars flowing to Maine.
Current Odds. Kalshi prices the binary winner contract at 71 percent YES (Democrats win), 30 percent NO, with \$553,028 in volume and 126,886 contracts in open interest. The margin curve trades in line: D wins by one or more points at 63%, three or more at 57%, five or more at 39%, seven or more at 28%, and nine or more at 26%. The contract has held in a narrow 70-71% band for roughly two months, surviving the April 30 withdrawal of Janet Mills, the staggered release of additional Reddit material from Platner's acknowledged account, and the May 20 release of the Pan Atlantic Research poll. Price stability through these events suggests the market has settled on a polling lead with a modest incumbency discount, and is not actively repricing on new information.
Expert Predictions and Crowd Narrative. As of May 2026, Cook Political Report and Sabato's Crystal Ball both rate the Maine Senate race as Tossup. Inside Elections rated it Tilt Republican as of March 2026, and has since moved toward a Tossup after Mills' withdrawal. The analyst consensus centers near 50% rather than above it, which means the model's 58% fair value sits modestly above forecaster consensus. CNBC reporting from May 15 framed Collins's appropriations-chair leverage as the central defensive argument, citing roughly \$1.5 billion in directed Maine spending since 2021. Senator Chris Van Hollen flagged the ongoing Platner disclosures as a continuing concern for Senate Democrats. Bernie Sanders has endorsed Platner publicly and continues to defend him through the disclosure cycle. The market is pricing the Pan Atlantic +7 point lead, Platner's Q1 2026 fundraising lead (\$4.1 million to Collins's \$3.1 million), and the broader anti-incumbent midterm environment, then applying a modest discount for incumbency and uncertainty to land at 71%.
The narrative gives limited independent weight to two factors: Collins's documented Maine-specific overperformance, and her 3.7x cash-on-hand advantage at end-Q1 (\$10 million versus \$2.7 million), which positions her to outspend Platner directly even before the \$80 million in booked outside money is deployed.
Position. The market is overpricing the Democratic win probability. Fair value is 58% YES against a market of 71%, an edge of 13pp on the NO side. The model's 58% sits between the Cook and Sabato Toss-up rating (~50%) and the market at 71%. The forecasters see no favorite, the model has Democrats modestly favored, and the market has them clearly favored. The position bets that the market is the most overconfident of the three.
The mispricing is driven by two factors the market is underweighting: Susan Collins's measurable history of outperforming her Maine polling, and a biographical controversy stack around Graham Platner that is substantively larger than its Q1 polling footprint and still in an active disclosure cycle. The recommended position is BUY NO at 30 cents, sized small (25% of full size) given a 2/10 edge robustness score that flags the position as fragile to a single positive polling cycle for Platner.
Fair Value Methodology
The analysis adjusts the polling lead for four additional factors and converts the resulting margin into a win probability using a normal distribution. The polling-error standard deviation is anchored to FiveThirtyEight's Senate accuracy series (5.2 points weighted-average final-stretch error since 1998), widened to 7 points to account for the 165 days remaining until the election. Maine 2020 (the 14-point Collins overperformance) is captured in factor 2 rather than used to set the error band, to avoid double-counting.
Factor 1. Polling Baseline (+6.5 points)
The Pan Atlantic Research poll, May 8-18, 2026, with 827 likely voters and a 3.7 point margin of error, has Platner at 48%, Collins at 41%, and undecided at 11%. Directional reads from Emerson College in March (Platner 48, Collins 41, out of 1,075 likely voters) and the Maine People's Resource Center in March (Platner 48.3, Collins 39.3, out of 1,167 likely voters) align. The slight discount from a raw +7 to +6.5 reflects a single-pollster discount applied to the most recent read, with cross-pollster directional agreement preserved. The 11% undecided residual matters under ranked-choice voting and is handled separately in factor 5.
Factor 2. Collins Overperformance Discount (–3.5 points)
Susan Collins has a documented Maine polling miss. The 2020 RealClearPolitics final average showed Sara Gideon leading by 5.5 points; Collins won by 8.6 points, a net miss of roughly 14 points. The Bangor Daily News retrospective from November 5, 2020 attributed the miss principally to late undecided movement toward the incumbent and an underestimate of Maine ticket-splitting, alongside Collins's incumbency brand. The analysis applies a discounted 3.5 points, roughly one quarter of the 2020 miss. The discount reflects several offsetting considerations: 2026 is a midterm with no presidential coattail, and the out-party historically gains in Senate races when a first-term president is unpopular; Collins's favorability has eroded, with Emerson's March 2026 read at 38% favorable against 57% unfavorable (–19 net, –30 net among independents); and Platner's base skews heavily grassroots, with the Maine Wire's FEC review showing roughly 64% of his \$11.8 million in individual contributions came from unitemized small-dollar donors under \$200, a proxy for grassroots enthusiasm Collins does not match. Collins does retain 29 years of incumbency and appropriations-chair leverage, which is why the discount is not zeroed out.
Factor 3. Midterm Environmental Wave (+2.0 points)
Midterm elections with a first-term president have favored the out-party in Senate races. Trump's national approval sits below 50% in May 2026, and the generic congressional ballot leans Democratic by 2 to 3 points. Maine carries a structural Democratic federal-level tilt, though it has been narrowing (Biden +9 in 2020, Harris +7 in 2024). Maine has consistently elected the independent Angus King to the Senate, suggesting receptiveness to candidates positioned outside the MAGA wing of the Republican party. The 2-point assumption captures the wave without overweighting it, and sits below the raw national environment because the trend in the Maine presidential margin is downward.
Factor 4. Biographical Risk (–3.0 points)
This factor carries the largest revision relative to a polling-led model. The original disclosure was a chest tattoo resembling the Totenkopf, a Nazi SS symbol. Platner first acknowledged the tattoo on Pod Save America in October of 2025, stating he got it drunk in Croatia in 2007 with fellow Marines and was unaware of its Nazi association. CNN's KFile follow-up later that same month surfaced Reddit posts from his acknowledged account discussing the symbol's military adoption in detail, with an October 2025 Jewish Insider report covering separate posts in which Platner defended a man with an SS lightning-bolt tattoo who had impersonated a federal officer at a 2020 Black Lives Matter protest in Las Vegas. A named former acquaintance recalled to both outlets that Platner had referred to the marking as his Totenkopf years earlier, and his own former political director Genevieve McDonald (who resigned in October 2025) publicly stated Platner himself had acknowledged the tattoo could be problematic. The campaign has dismissed both accounts. USA Today subsequently reported the tattoo has been covered and quoted Platner describing himself as a lifelong opponent of Nazism, antisemitism, and racism.
But the disclosure cycle did not end with the tattoo. From October 2025 through May 2026, additional Reddit posts attributed to Platner have surfaced, spanning multiple Democratic-coalition vulnerabilities at once, including self-description as a socialist-turned-communist, all-cops-are-bastards (ACAB) framing, victim-blaming language toward sexual assault survivors, homophobic jokes, derogatory characterizations of rural white Americans, and a May 2026 post mocking a Purple Heart recipient. The pattern is not a single-domain story that can be neutralized with one apology. Republican-aligned outside groups have booked more than \$80 million in advertising through Election Day, and the closing argument will be built around this material. Intra-party criticism has been visible but not decisive, with primary challenger Jordan Wood calling for Platner to withdraw in October 2025. Sanders's endorsement and the candidate's response (framing the posts as products of post-deployment trauma and pre-therapy anger) have held the campaign together. Net contribution is –3.0 points: +1.0 for fundraising and grassroots enthusiasm, against –4.0 for the biographical narrative running across \$80 million of paid media. A left tail of 10-20% probability of a candidacy-breaking disclosure shifts the factor to –6 to –10 points and is reflected in the fair value range.
Factor 5. Ranked Choice Voting Redistribution (–0.5 points)
Maine uses ranked-choice voting for federal elections. With 11% of voters undecided or allocated to neither major candidate in the May Pan Atlantic poll, the second-choice math matters if no candidate clears 50% in round one. The 2020 Colby College ranked-choice exit-poll work showed asymmetric second preferences (roughly 51% of Green-aligned Lisa Savage voters ranked Sara Gideon second versus 13% for Collins), but the race never reached round two because Collins cleared 50% outright. With no declared major third-party candidate in 2026, the 11% undecided pool is dominated by low-engagement moderates who recognize Collins. The net assumption is a mild Collins lean of 0.5 points. The factor would re-rate larger if a recognizable independent entered the race, but that scenario is not currently in play.
Roll-Up and Cross-Market Consistency
Net adjusted margin equals +6.5 plus (–3.5) plus (+2.0) plus (–3.0) plus (–0.5), which sums to +1.5 points. Translating that through the normal distribution with a 7-point standard deviation gives a win probability of 0.5848, or 58% YES, 42% NO. The implied two-party split (Platner 51.5, Collins 48.5) sits within one standard deviation of the Pan Atlantic read, with the undecided pool splitting roughly evenly after ranked-choice redistribution. The Cook and Sabato Toss-up ratings sit roughly 8 points below this point estimate, meaning the model is more optimistic about Democrats than the analyst consensus, even as it is meaningfully less optimistic than the market. The model is also internally consistent across the Democratic margin curve. The implied probability of Dems win by 3 or more is 42% (market 57), by 5 or more is 31% (market 39), and by 9 or more is 14% (market 26). The largest absolute mispricing is on the Dems-win-by-3 line at 15 points, with the Dems-win-by-9 contract being the cleanest expression of the thesis that the market is concentrating too much probability mass in the upper margin tail.
NMT Sensitivity Table
The table tests how fair value moves under standard adverse perturbations to each factor. Sensitivities use the normal probability density at the base point (Φ of 0.214 = 0.39).
| Factor | Margin Contribution | Stress (pp) | FV Sensitivity | |---|---|---|---| | 1. Polling baseline | +6.5pp | ±3 | ~17pp per 3pp shift (HIGH) | | 2. Collins overperformance discount | –3.5pp | ±2 | ~11pp per 2pp shift (HIGH) | | 3. Midterm environmental wave | +2.0pp | ±2 | ~11pp per 2pp shift (HIGH) | | 4. Biographical risk (Platner) | –3.0pp | ±2 | ~11pp per 2pp shift (HIGH) | | 5. RCV redistribution | –0.5pp | ±0.5 | ~3pp per 0.5pp shift (LOW) |
Four of the five factors are high-sensitivity. The standard adverse perturbation (1x = the typical one-cycle move in each input) tightens the edge sharply in every case. On Factor 1 (polling), a 3 point positive shift for Platner takes fair value to 74% and flips the edge to the YES side by 3pp. On Factors 2, 3, and 4 a 1x adverse shift takes fair value to roughly 69%, compressing the NO edge from 13 points to under 2 points, which is below a reasonable threshold for holding a position. Worst-case edge robustness is therefore 2/10, set by the polling factor, and the broader picture is that the headline edge does not survive standard stress on any single high-sensitivity input. This is the central reason the position is sized small.
Risk Factors and Catalysts to Watch
The dominant risk is that the biographical disclosure cycle stabilizes. If the May 19, 2026 release proves to be the last meaningful drop and voters habituate over the 165 days remaining, the negative contribution compresses and fair value moves toward market. The Sanders endorsement and the structural inability to replace Platner on the ballot lock the campaign in regardless of further damage, so voter response is the only variable. The Collins overperformance assumption itself carries judgment risk, as the 2020 miss may not repeat in a midterm with a different turnout composition, and Pan Atlantic is just one pollster. A consolidation of the 11% undecided pool toward Platner expands his margin past the level where NO remains tenable. A Trump approval recovery would compress the wave factor.
Catalysts to watch: a Maine poll showing Platner +9 or more closes the edge, while a read inside +4 confirms the thesis. A clean two-week period without new Reddit material works against the position, and a fresh disclosure breaking above CNN or major-network television supports it. If the contract on a 5+ point Democratic win rallies from its current 39% to above 50%, close the position.
Democrats winning the Senate seat in Maine at 71 cents YES appears overpriced. Fair value sits at 58% (range 24-94%) on a +6.5 polling baseline, a 3.5 point Collins overperformance discount anchored to her 14 point 2020 miss, a 2.0 point favorable midterm wave, a 3.0 point biographical drag, and a 0.5 point Collins lean in ranked-choice redistribution. Recommended position is BUY NO at 30 cents, sized small (25%).