Tools chosen: polymarket_direct kalshi_related polymarket_related claude_news claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a Polymarket election question about the 2026 North Carolina Senate race, likely Roy Cooper (D) vs Michael Whatley (R) for Thom Tillis's open seat. The market price is the primary anchor; polling averages, fundraising, national environment (generic ballot, presidential approval), and NC's historical partisan lean are the key empirical drivers. Cross-venue comparison with Kalshi provides an independent crowd signal.

# Current state The 2026 NC Senate race pits Democrat Roy Cooper (former governor) against Republican Michael Whatley (former RNC/NC GOP chair); both are confirmed nominees (Wikipedia). As of late August 2026, polling shows Cooper leading by 5-11 points depending on poll, and nonpartisan forecasters (Cook, Sabato, Pollsmax) rate the race Lean/Likely Democratic — a shift from earlier "Toss-Up" status. Resolution requires AP+Fox+NBC to call the race for the same candidate, or official certification; election is Nov 3, 2026 and market closes at question close time same date. # Timeline of key events - 2025 (date unspecified): Sen. Thom Tillis (R) announces he will not seek re-election after opposing the "One Big Beautiful Bill Act" (confirmed, Wikipedia). - Early 2026: Whatley wins GOP nomination with 64.6% (over Don Brown) after Trump endorsement; Cooper wins Dem nomination with 92% (confirmed, Wikipedia). - 2026-06-11: Cook Political Report and Sabato's Crystal Ball move race from "Toss-up" to "Lean/Leans Democratic" (reported, ncnewsline.com). - 2026-07-14: Cooper campaign reports $14.8M raised in Q2 (reported, WSOC-TV). - Late Q2 2026 (period ending 6/30): FEC filings show Cooper $34.9M raised/$20.7M cash-on-hand vs. Whatley $11.2M raised/$3.4M cash-on-hand (reported, thecentersquare.com/carolinajournal.com). - 2026-07-30: Fox News poll: Cooper +9 (53-44) (reported, Yahoo/Fox). - 2026-08-11/13: Elon University poll: Cooper +11 (53-42) (reported, WRAL/WXII12); Silver Bulletin generic ballot D+8.1 with LV adjustment. - Mid-Aug 2026: High Point University poll: Cooper +5 (50-45); Carolina Forward/Change Research: Cooper +7 (reported, thehill.com). - 2026-08-22: Multiple outlets ("Cooper slipping") note tightening race per newer polls (reported, Townhall/WCBM). - 2026-08-25: Senate Leadership Fund launches new ad targeting Cooper, part of pledged $71M NC investment (reported, WBT/predictionedge.com). - Ongoing: Senate Majority PAC (D) has committed $31M+$11M additional to flip the seat (reported, WRAL). # Event Will Republicans win the 2026 North Carolina U.S. Senate race (Whatley vs. Cooper), per AP/Fox/NBC race call or official certification. # Outcomes to forecast - Yes (Republican win) - No (Democrat win) # Kalshi market anchor No direct Kalshi YES price for this specific ticker was returned by kalshi_direct in the research provided — this is a **gap**. The closest Kalshi proxy is a *different* market (SENATENC-28-R, the 2028 NC Senate Republican contract) priced at 27% (7d +2%, 30d -6%, 21 days of data) — not directly comparable since it's a different race/year. Polymarket's equivalent 2026 contract (same event) trades YES(R) at **8.5%** (7d +1%, 30d -4.5%, range 7.5-16.5%, $59K volume, 90 days data) — this is the closest true cross-market anchor available and should be treated as the primary external benchmark absent live Kalshi data. # Sub-question answers 1. **Polymarket price/90-day trend** — Currently 8.5% for Republican win; down from a 90-day high of 16.5%, recent 7-day uptick (+1%) but 30-day decline (-4.5%). (polymarket_direct) 2. **Kalshi price vs. Polymarket** — No direct Kalshi 2026 NC Senate market price was retrieved; only a 2028 NC Senate market (27% for R) was found via kalshi_related, which is not the same event. Cannot confirm divergence. (kalshi_related — gap) 3. **Head-to-head polling** — Cooper leads in all recent polls: Elon +11 (53-42), Fox +9 (53-44), High Point +5 (50-45), Carolina Forward +7, Carolina Journal/Harper +13 (52-39); earlier polls (May-June) showed +14-15. Average recent margin ~+7-9 D, with signs of modest tightening in August. (claude_news, gdelt_news, WRAL, thehill, cbs17) 4. **Fundraising** — Q2 2026: Cooper raised $34.9M ($20.7M cash-on-hand, or ~$27.4M incl. joint committees); Whatley raised $11.2M ($3.4M cash-on-hand, ~$7.8M incl. joint committees) — roughly 3x fundraising advantage for Cooper. (thecentersquare.com, carolinajournal.com) 5. **National environment** — Generic ballot: D+6 to D+8 (RCP, Decision Desk HQ, Silver Bulletin). Trump approval weak: 33-40% approve depending on poll (Reuters/Ipsos 33%, Decision Desk 38.7%, Gallup 40%), net approval around -19 to -20 (Silver Bulletin). This is a classic midterm-penalty environment against the president's party (R). (claude_news) 6. **Historical base rate** — Republicans are 5-for-5 in NC Senate races since 2008; Democrats haven't won a NC Senate midterm race since 1998. This cuts against a Democratic win structurally, but code_execution model estimates president's party wins only ~35-40% of true toss-up midterm Senate races, and a polling-based normal model (Cooper +5, 6pt SD) yields P(R wins)≈20%, blended to ~25%. (thecentersquare.com; code_execution) 7. **Primary challenges/third parties** — None credible identified; Whatley won GOP primary 64.6% over Don Brown, Cooper won Dem primary 92%, both are settled two-way nominees. No independent/third-party threat noted. (Wikipedia) # Key facts (high-confidence, factual) 1. [Wikipedia] Whatley (R) and Cooper (D) are confirmed nominees; Tillis (R incumbent) retired. 2. [claude_news/thecentersquare] Cooper leads Q2 fundraising ~3:1 and cash-on-hand ~6:1. 3. [claude_news] Cook/Sabato moved race to Lean/Leans D in June 2026. 4. [gdelt_news/WRAL] Multiple August 2026 polls show Cooper ahead by 5-11 points. 5. [thecentersquare] Republicans 5-for-5 in NC Senate races since 2008 (strong structural GOP lean historically). 6. [claude_news] SLF committed $71M; Senate Majority PAC $42M+ — heavy spending both sides. # Cross-market signals - Kalshi related: only 2028 NC Senate market found (27% R) — different cycle, limited comparability. - Polymarket: 8.5% YES(R), down from 16.5% high, modest recent uptick. - Sportsbook implied: none found. - Model estimate (code_execution): ~20-25% P(R wins) via polling+base-rate blend. # Analyst opinions and speculation - Cook Political Report: NC is the GOP seat Democrats are "most likely to flip." (cookpolitical.com) - Pollsmax model: Cooper 81.5% win probability, Likely Democratic rating. - Some outlets (Townhall, WCBM) frame tightening polls as "Cooper slipping," suggesting media narrative diverges somewhat from underlying poll aggregates, which still show consistent Cooper lead. # Directional lean per outcome - **Yes (R wins)**: Supported by historical GOP dominance in NC Senate races (5/5 since 2008), heavy outside GOP spending ($71M SLF), NC's 2024 presidential lean (Trump +3.2%), and polling tightening in August. Opposed by consistent multi-poll Cooper lead (5-14 pts), weak Trump approval, D-favorable generic ballot, fundraising disadvantage, and forecaster consensus (Lean/Likely D). - **No (D wins)**: Supported by consistent, converging polling averages (Cooper +7-9 avg), strong fundraising edge, favorable national environment, forecaster ratings. Opposed by NC's recent Senate-race history and large late GOP outside spending that could close the gap by November. # Gaps / unknowns - No live Kalshi YES price was retrieved for this exact ticker — a critical anchor gap; only Polymarket (8.5%) available. - Uncertain how much of the outside-spending ($71M SLF, $42M+ SMP) has already been reflected in current polling. - No data on early voting/turnout modeling or late-race momentum past Aug 25, 2026. - Polling error models are illustrative/placeholder (code_execution explicitly flags this). # Calibration anchors - Polymarket YES(R): 8.5% (current, closest true market anchor). - Kalshi 2028 NC Senate (R) proxy: 27% (different cycle, low comparability). - Model-based blended estimate: ~20-25% P(R wins). - Historical precedent: NC Senate seats: R has won 5 consecutive contests since 2008; last D win at midterm was 1998.

1. Both forecasters converge tightly (14%/15%) around a midpoint between Polymarket (8.5%) and the model blend (20-25%), but neither justifies why the true probability should sit closer to the market than the model — historically, thin/low-volume prediction markets ($59K volume noted) can be poorly calibrated for tail events, and the brief explicitly flags this as an anchor gap, so leaning so heavily on Polymarket may be a reference-class error. 2. Neither forecast adequately weighs the NC-specific historical base rate (5-for-5 GOP wins since 2008, no Dem midterm win since 1998) beyond a cursory mention — this is a strong structural prior for a purple state with consistent recent polling misses favoring Republicans (noted as "persistent pro-GOP polling error in NC" but not quantified or incorporated into the final number). 3. Both forecasts underweight the timing gap: current polling/fundraising data is from June-August 2026, but the market resolves in November — three additional months allow the $71M SLF and other late GOP spending to materially close a 5-11 point gap, especially given the "Cooper slipping" narrative already emerging by Aug 22-25, yet both forecasts treat this as a minor tweak rather than a major uncertainty driver given ~10 weeks of unmodeled campaign dynamics remain. 4. Neither forecaster addresses the divergence between poll-based estimates (Cooper +5 to +11) and the Pollsmax model's explicit 81.5% Cooper win probability (implying ~18.5% Republican) — this data point sits within their range but isn't explicitly reconciled, and no forecaster discusses why they're closer to Polymarket's more bullish-for-Dem 8.5% than Pollsmax's more moderate 18.5%. 5. Both forecasts assign similar confidence levels (0.62, 0.72) despite acknowledging significant data gaps (no live Kalshi price, "illustrative/placeholder" polling error models per the brief) — this confidence seems too high given the explicit admission that the anchor and error-model inputs are unreliable or missing. 6. Neither forecast considers scenario-splitting for how outside spending timing interacts with early voting — if NC has substantial early/mail voting before major ad blitzes take effect, the window for GOP spending to move numbers could be narrower than the November 3 close date suggests, a nuance absent from both rationales.

16% (67% confidence)

Yes: 16% No: 84%

Cooper leads every August 2026 poll (average ~D+7-8), holds ~3:1 fundraising and ~6:1 cash-on-hand advantages, and sits in a D+6-8 national environment with Trump approval near 33-40%; forecasters have moved the race to Lean/Likely D. The closest same-event market anchor (Polymarket, 8.5%) is thin ($59K volume) and likely too aggressive for a race that resolves ~10 weeks after the latest data, so I weight it less heavily than the two prior forecasts implicitly did. A normal model with Cooper +7 and a 6-7 point effective error SD yields ~13-16% for a Republican win, and NC's structural GOP lean (5-for-5 in Senate races since 2008, persistent pro-GOP polling error), plus $71M in late SLF spending and an emerging tightening narrative, argue for the upper end of that band — close to Pollsmax's implied ~18.5%. I therefore set Yes at 16%, slightly above both submitted forecasts and well above the thin market line.

Total pipeline time: 152.1s