Tools chosen: kalshi_direct kalshi_related claude_news gdelt_news wikipedia polymarket_related code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a Kalshi-sourced long-tail candidate market in a multi-candidate top-four primary field, so the direct Kalshi price and the prices of the other candidates in the same event (which should roughly sum to 4 slots) are the dominant anchors. Secondary research must establish which office the primary is for, whether John Brendan Williams is a real declared candidate with any name recognition/funding, and the base rate for obscure filers advancing in Alaska's top-four system.

# Current state The 2026 Alaska top-four primary (for the U.S. House at-large seat) was held 2026-08-18; ballot counting continued through late August with certification expected 2026-08-31. As of the Aug 26 update, John Brendan Williams (D-Fairbanks) placed 5th (2.6%) but is positioned to move into the 4th and final advancing slot after 3rd-place finisher Matt Schultz withdrew — per Alaska Public Media, "the next two challengers will advance... Hafner in third and Williams in fourth." However, the Kalshi JWIL contract is currently priced at just 2¢, a sharp divergence from what the news narrative implies, and down 61 points from a 30-day high of 70¢ — this gap is unresolved in the research and should be treated as a major open flag. # Timeline of key events - 2026 (pre-primary): John Brendan Williams (Fairbanks Democrat) files with FEC, becomes AK Democratic Party's first declared House candidate — confirmed (mustreadalaska.com). - 2026-08-18: Alaska top-four nonpartisan primary held; preliminary House results: Begich (R, incumbent) 46.2%, Hill (I) 31%, Schultz (D) 8.1%, Hafner (D, federal inmate) 3.8%, Williams (D) 5th at 2.6% — confirmed (alaskabeacon.com, alaskasnewssource.com). - 2026-08-21: Reports that Schultz has suspended his campaign and is expected to withdraw, which would elevate Williams to 4th place; Republican Clay Strickland close behind in 6th (2.4%) — reported (alaskabeacon.com). - 2026-08-26: Update confirms Schultz's withdrawal; standings "largely unchanged" with ~8,700 new ballots counted; Hafner (3rd) and Williams (4th) positioned to advance; neither has reported fundraising — reported (alaskapublic.org). - Certification expected 2026-08-31 — not yet confirmed in research. - Kalshi JWIL price: reached a high of 70¢ at some point in the 35-day window, then fell to current 2¢ (-61% over 30 days) — confirmed via kalshi_direct, but the cause of this reversal is not explained in any source retrieved. # Event Will John Brendan Williams advance to the top-four in the 2026 Alaska U.S. House at-large primary? # Outcomes to forecast - Yes (Williams advances) - No (Williams does not advance) # Kalshi market anchor Current YES price: **2%** (2¢). 7-day change: 0%. 30-day change: **-61%** (fell from a peak of 70%). Price range over 35 days: 1%–70%. Average daily volume: 612 contracts — moderately liquid. The steep, unexplained decline from 70¢ to 2¢ is the single most important — and most unresolved — data point in this brief. # Sub-question answers 1. **What office/date?** This is the U.S. House at-large seat, 2026 Alaska top-four nonpartisan primary held 2026-08-18, feeding into a Nov 3, 2026 ranked-choice general election (Wikipedia). 2. **Current Kalshi price/trend?** 2%, down sharply (-61 pts) over 30 days from a 70% peak; 612 contracts/day average volume (kalshi_direct). 3. **Do top-4 already absorb ~400¢?** Not directly measurable — no live order book for all KXAKPRIMARY outcomes was retrieved (kalshi_related returned no House-specific outcome data). A code_execution simulation using illustrative numbers estimated Williams' de-vigged probability near 3-5% if he's an "outsider," but this is not based on real prices. 4. **Who is Williams?** Fairbanks Democrat, first declared AK Democratic candidate for the House seat, filed with FEC July 2026; no reported fundraising, no polling profile beyond primary vote share (mustreadalaska.com, alaskapublic.org). 5. **Base rate for unknown candidates finishing top-4?** No historical Alaska-specific base rate was found in research; code_execution's generic model suggests ~8-14% for a non-top-3 filer in a 10-15 candidate field with 3 "locked" favorites — but Williams is a special case (post-primary withdrawal elevation, not a from-scratch base rate). 6. **Recent entries/exits changing contender count?** Yes — Matt Schultz (3rd place, 8.1%) withdrew around Aug 21-26, directly elevating Williams from 5th to a projected 4th place (alaskabeacon.com, alaskapublic.org). No other major entries/exits reported. # Key facts (high-confidence, factual) 1. [alaskabeacon.com] Williams placed 5th with 2.6% of primary vote as of Aug 18-21 counts. 2. [alaskapublic.org] Schultz (3rd, 8.1%) withdrew; as of Aug 26, Williams and Hafner are positioned as the 4th/3rd advancers respectively. 3. [alaskapublic.org] Neither Hafner nor Williams has reported campaign fundraising. 4. [kalshi_direct] JWIL YES price fell from a 70% high to 2% currently, a 61-point 30-day decline. 5. [Wikipedia] Certification and final field lock-in occurs after primary; general election is Nov 3, 2026 via ranked-choice voting among the top four. # Cross-market signals - Kalshi related: SENATEAK-28-D at 24%; unrelated House-specific comparison markets not found. - Polymarket: No Alaska 2026 markets found (0 matches). - Sportsbook implied: N/A (not applicable to this race). # Analyst opinions and speculation - News coverage (alaskabeacon, alaskapublic) frames Williams' advancement as highly likely/near-certain pending final certification, given Schultz's withdrawal and Hafner/Williams' clear positioning in 3rd/4th. - Code_execution's illustrative modeling suggests a generic "unknown filer" base rate of single-digit percent, but explicitly does not incorporate the actual Schultz-withdrawal mechanic, which materially changes Williams' odds upward. # Directional lean per outcome - **Yes**: Strong support from news — Schultz withdrawal directly elevates Williams to the 4th and final slot per multiple Alaska outlets (Aug 21-26). Weak/no fundraising doesn't matter once the top-4 slot is procedurally assigned. - **No**: Supported only by the Kalshi price itself (2%), which sharply contradicts the news narrative; possible explanations (uncounted absentee ballots changing order vs. Strickland/Hafner, legal uncertainty about withdrawal-elevation rules, stale/mispriced market) are unconfirmed. # Gaps / unknowns - No confirmation of Aug 31 official certification results — final order still technically pending in retrieved sources. - The stark Kalshi price collapse (70%→2%) is unexplained by any retrieved news; could reflect information not in this research (e.g., Schultz un-withdrawing, legal ruling, late ballots flipping Williams below Strickland/Hafner) or could be a market/liquidity anomaly. - No live full order book for all KXAKPRIMARY-House outcomes to cross-check the top-four pricing. # Calibration anchors - Kalshi current YES price: **2%** (anchor). - Recent precedent: Kalshi price for this same contract touched 70% within the past 35 days, showing the market itself has been highly uncertain/volatile on this exact question.

1. Both forecasters treat the market's collapse as informative "smart money" resolving the question, but neither seriously grapples with the alternative that a thinly-reasoned/illiquid market on an obscure down-ballot contract simply mispriced or reacted to noise — 612 contracts/day on a niche market is not strong evidence of informed trading, yet both treat it as near-decisive. 2. Neither forecast addresses the mechanical/legal question of whether Alaska's top-four rules actually elevate the next-highest finisher when a candidate withdraws post-primary — this is flagged in the brief as unconfirmed, and if withdrawal does NOT trigger automatic elevation (e.g., if the seat simply becomes uncontested or handled differently), Williams's path to advancement could be moot regardless of vote totals. 3. Both forecasters converge tightly (5% and 10%) with the market's 2% anchor, but neither considers weighting more toward the news narrative given that the brief explicitly states outlets frame Williams's advancement as "highly likely/near-certain" — the two independent forecasts essentially just deflate the news view without a rigorous mechanism for the deflation amount, making their specific point estimates arbitrary. 4. The narrow 2.6% vs 2.4% (Williams vs. Strickland) margin is mentioned by Forecast 1 as a plausible flip explanation but this is speculative; neither forecaster asks why a mere 0.2-point margin risk would crash a price from 70% to 2% rather than to something like 30-40%, which would be more proportionate to genuine vote-count uncertainty — the magnitude of the collapse suggests something more categorical (e.g., a rules-based disqualification) rather than a marginal recount risk, which neither forecast explores. 5. Neither forecast considers Clay Strickland (6th, 2.4%, Republican) as a distinct competing scenario — if late ballots or the withdrawal mechanism favor him over Williams, this is a different failure mode than "ballot recount noise" and deserves separate treatment, particularly since a Republican challenger emerging in 4th could be a more coherent explanation for market repricing (e.g., a specific report the market saw but the brief's researchers didn't retrieve). 6. Both forecasts assign high confidence (implicitly via 90-95% No) despite the brief's own framing that this is "the single most important — and most unresolved" data point, and that certification status is still pending confirmation — the appropriate response to such unresolved conflicting evidence might be a less extreme distribution (e.g., 15-25% Yes) rather than clustering just above the raw market price, since neither forecaster has resolved *why* the market moved.

7% (46% confidence)

Yes: 7% No: 93%

The decisive signal is the Kalshi contract sitting at 2¢ with a flat 7-day price after collapsing from 70¢ on a moderately liquid market (612 contracts/day) — a pattern far more consistent with a question that has effectively resolved NO than with a stale or noise-driven quote. The pro-YES news narrative (Schultz's withdrawal elevating Williams from 5th to 4th) dates to Aug 21–26 and plausibly explains the earlier 70¢ level; the crash almost certainly reflects post-certification information the research did not retrieve — either the razor-thin 2.6% vs 2.4% margin flipping to Strickland on late absentees, or Alaska's withdrawal-substitution timing rules not producing the expected elevation. The critique is right that neither forecaster identified the exact mechanism, and that the sheer magnitude of the drop suggests something categorical rather than marginal recount risk, but that cuts against YES, not toward it. I therefore stay close to the market while carrying a modest premium above 2% for the residual chance of a thin-market anomaly or a mispriced/misunderstood contract.

Total pipeline time: 191.2s