Tools chosen: polymarket_direct kalshi_related polymarket_related claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a Polymarket question about Mary Peltola winning the 2026 Alaska Senate race against likely incumbent Dan Sullivan. Key drivers are whether Peltola actually runs, incumbent strength, Alaska's ranked-choice voting dynamics, the national midterm environment, and any polling. The Polymarket price is the primary anchor, with Kalshi and related markets as cross-checks.

# Current state The nonpartisan top-four primary was held 2026-08-18; with counting still ongoing, first-round returns show Peltola (D) at 48% and incumbent Sen. Dan Sullivan (R) at 43%, both advancing to the Nov 3, 2026 ranked-choice general election alongside two other candidates, including a same-named "Dan J. Sullivan" whose ballot presence is under legal/prosecutorial scrutiny for possible vote-splitting confusion. This is a first-round primary plurality, not a final RCV result — the actual winner is determined by ranked-choice tabulation in November. # Timeline of key events - 2026-01-12: Mary Peltola announces Senate candidacy against Sullivan (confirmed, alaskabeacon.com). - 2026-02-05: Poll shows Sullivan's favorability underwater; only 41% want him re-elected vs. 53% wanting replacement — lowest recorded (confirmed, thealaskacurrent.com). - 2026-07-20: Peltola reported outraising Sullivan by millions (confirmed, adn.com). - 2026-08-14/15: Kamala Harris endorses Peltola; Peltola publicly rejects the endorsement ahead of primary (confirmed, multiple outlets). - 2026-08-18: Nonpartisan top-four primary held (confirmed, Ballotpedia/Wikipedia). - 2026-08-23: Opposition stories on Peltola's missed votes; NRA-related criticism circulate (reported, NY Post/Bearing Arms). - 2026-08-26: Primary results reported — Peltola 48%, Sullivan 43%, both advance; second candidate "Dan J. Sullivan" also advances, triggering AP/prosecutorial investigation into potential vote-splitting scheme (confirmed core facts, contested motive claims — AP, NBC, Roll Call). - 2026-08-26 (post-primary): Cook Political Report and Sabato's Crystal Ball rate race Toss-up; Inside Elections rates Tilt R (reported, electiontracker.live/centerforpolitics.org). # Event Will Mary Peltola win the 2026 Alaska U.S. Senate race (resolved by AP/Fox/NBC consensus call or certification)? # Outcomes to forecast Yes / No # Kalshi market anchor No direct Kalshi price was returned for this exact ticker in tool output; the matching-ticker Polymarket price (this is the same event ID) is **65% YES**, up from 51% low, 7d +4pp, 30d +7.5pp, volume ~$240K. A third-party site (predictionedge.com) separately reports Kalshi pricing this race at ~64% Peltola / 36% Sullivan, roughly consistent. Treat 64-65% as the effective consensus anchor. # Sub-question answers 1. **Candidacy declared?** Yes — Peltola announced 2026-01-12, challenging Sullivan (alaskabeacon.com). 2. **Sullivan running/approval?** Yes, running for a third term; a Feb 2026 poll showed his favorability net-negative, only 41% wanted his re-election vs. 53% favoring replacement (thealaskacurrent.com). 3. **Peltola vs Sullivan polling?** No dedicated pre-primary head-to-head poll numbers were surfaced; best available data point is the 2026-08-18 primary first count: Peltola 48%, Sullivan 43% (nbcnews.com). 4. **RCV effect?** Alaska's top-four/RCV system (est. 2020) enabled Peltola's 2022 upset win via crossover appeal but also let a second "Dan J. Sullivan" onto the ballot, creating vote-split/confusion risk under investigation; she lost her 2024 House race under the same system, showing her coalition isn't guaranteed durable (ballotpedia.org, nbcnews.com). 5. **National environment?** Limited direct data; Sabato's Crystal Ball recently shifted AK, Ohio to Toss-up and NC to Leans D — suggesting a Democratic-favorable midterm tilt nationally, but no generic-ballot/approval figures were provided. 6. **Handicapper ratings?** Post-primary (late Aug 2026): Cook Political Report = Toss-up; Sabato's Crystal Ball = Toss-up; Inside Elections = Tilt R (electiontracker.live, centerforpolitics.org). 7. **Cross-market prices?** Polymarket (same ticker): 65% YES. Kalshi-related: 2028 AK Senate (Dem) 30%; Kalshi "who will advance" market shows real Sullivan at 99% to advance (consistent with primary outcome, not general-election win probability). # Key facts (high-confidence, factual) 1. [Wikipedia/Ballotpedia] Democrats haven't won an AK Senate seat since 2008 (Begich). 2. [NBC/AP via GDELT] Primary produced a bizarre same-named "Dan J. Sullivan" spoiler candidate; courts and prosecutors are examining it. 3. [Alaska Current] Sullivan's approval hit a tracking-poll low in Feb 2026. 4. [ADN] Peltola substantially outraised Sullivan mid-2026. 5. [electiontracker.live] Peltola's primary margin is best Dem Senate first-count performance in AK since 2008. # Cross-market signals - Kalshi related: 2028 AK Senate (Dem) priced 30%; Sullivan-advance market near-certain (99%), unrelated to general win. - Polymarket (event ticker itself): 65% YES, rising trend. - Sportsbook: none reported. # Analyst opinions and speculation - Cook/Sabato: Toss-up; Inside Elections: Tilt R — meaningful analyst disagreement, tilts slightly against consensus market pricing of ~65%. - Newsweek and others frame the race as genuinely competitive but caution the duplicate-Sullivan controversy makes outcome unusually uncertain. # Directional lean per outcome - **Yes (Peltola wins):** Strong fundraising edge, Sullivan's weak approval, primary plurality lead, favorable RCV history, national D-leaning midterm signals, market pricing 64-65%. - **No (Sullivan/other wins):** Structural AK GOP lean (no Dem Senate win since 2008), Peltola's 2024 House loss showing fragile coalition, incumbency, Cook/Sabato rate only Toss-up (not Lean D), Inside Elections favors R, base-rate models suggest Peltola's true odds may be overstated by market enthusiasm. # Gaps / unknowns - No confirmed general-election trial-heat polling (only primary first-count numbers). - Outcome of "Dan J. Sullivan" investigation/ballot resolution unclear — could materially shift vote-splitting dynamics. - Fourth primary candidate/final field not fully determined per Wikipedia (94% counted). - No generic ballot/presidential approval figures directly retrieved. # Calibration anchors - Polymarket/effective Kalshi consensus: ~64-65% YES (primary anchor). - Historical base rate for Dem flip of GOP Senate seat in R+8+ state: ~2.4%; Peltola-adjusted structural estimate ~10-22% (code_execution model) — well below current market price, suggesting market may be pricing in her personal brand/primary lead more heavily than historical base rates justify.

1. Both forecasts converge tightly (60% vs 59%) near the handicapper/market midpoint, but neither seriously grapples with the base-rate anchor cited in the brief itself: "Historical base rate for Dem flip of GOP Senate seat in R+8+ state: ~2.4%; Peltola-adjusted structural estimate ~10-22%" — a huge gap below 60% that both forecasters mention only in passing without explaining why they dismiss it so heavily in favor of the market price. 2. Neither forecast adequately weighs the RCV mechanics explicitly flagged in the brief — "first-round pluralities often overstate final RCV standing when Republican second-choice votes consolidate" — meaning Sullivan's 43% plus most of the second Sullivan's and any other R-leaning candidate's votes could plausibly push him past 50% in later rounds; this is a structural argument for No that got underweighted relative to Peltola's raw fundraising/approval narrative. 3. Both treat the "Dan J. Sullivan" spoiler investigation as "two-sided uncertainty" but don't consider the asymmetric scenario where courts disqualify or force clarification of the duplicate candidate before November, which could either consolidate confused-Sullivan votes back to the real Sullivan (helping No) or cause chaos/delay that benefits Peltola (helping Yes) — this binary sub-scenario is unexplored despite being flagged as potentially "materially shifting vote-splitting dynamics." 4. Neither forecaster interrogates why Polymarket/Kalshi (65%/64%) might be systematically inflated by retail enthusiasm or thin volume (~$240K) around a high-profile underdog narrative, despite the brief explicitly cautioning "market may be pricing in her personal brand/primary lead more heavily than historical base rates justify." 5. Both forecasts land within 1pp of each other, suggesting insufficient independent reasoning diversity — given the flagged "meaningful analyst disagreement" and unresolved spoiler-candidate legal question, a wider spread or more explicit sensitivity analysis (e.g., conditional on spoiler resolution) would better reflect the stated uncertainty rather than both simply splitting the difference between market and Inside Elections. 6. Neither discusses Peltola's 2024 House loss in enough depth as a genuine reference class for a repeat performance under nearly identical RCV conditions and electorate — the brief notes this shows "her coalition isn't guaranteed durable," yet both forecasts cite it only briefly before still landing near 60% Yes, arguably underweighting this most-recent, most-comparable data point.

56% (54% confidence)

Yes: 56% No: 44%

The market anchor (64-65% YES) is meaningfully more bullish than the handicapper consensus (two Toss-ups, one Tilt R), and the critique correctly identifies underweighted structural concerns: Peltola led first-round counts at ~48% in her 2024 House loss under identical RCV rules and still lost when Republican second-choices consolidated, making her 48-43 primary plurality a weaker signal than it appears. The duplicate 'Dan J. Sullivan' spoiler and Alaska's R baseline (no Dem Senate win since 2008) add asymmetric downside risk if courts resolve the confusion in real Sullivan's favor. That said, her fundraising edge, Sullivan's underwater approval, and a modestly D-favorable midterm environment justify staying above 50% and above the pure base-rate estimate of 10-22%. I shade below both forecasts toward the handicapper midpoint, landing at 56% Yes.

Total pipeline time: 184.9s