EHIQ Accuracy
Every probability call EHIQ publishes is logged before resolution and graded against the prevailing market (Kalshi, Polymarket, or the closest comparable). Hits, misses, and calibration notes all live here. No edits. No deletions. The math below is the whole math.
The math
Clean hit rate = hits ÷ (hits + misses + calibration notes). We count a call as a hit only when both the outcome AND the EHIQ rationale resolve correctly. A right outcome reached via wrong reasoning becomes a calibration note — counted separately, never silently folded into the headline.
Edge at call time = absolute difference (in percentage points) between EHIQ's published probability and the prevailing market probability at the moment EHIQ posted. Averaged across all resolved real-time calls.
What's excluded: 3 historical validation runs (Brexit, Biden 2020, Carney 2025, etc.) are kept on the predictions page in a separate “Historical Validation” section. They never blend into the headline hit rate — only real-time, dated, pre-resolution calls are counted here.
Hit rate by category
Every resolved call · 20
See active calls →- HITElectionsJun 3, 2026
Steve Hilton — CA Governor primary first-place
Hilton leading the CA Governor primary at ~26.7% (vote counting ongoing, but Hilton has held the #1 spot since polls closed 2026-06-02). EHIQ called Hilton at 25% for first-place against Polymarket's 9.7% — a +16pp market fade. The call was contrarian (Hilton was the lowest-priced of the top-tier candidates on Polymarket) and the result has Hilton leading. Resolving as HIT on directional outcome; final order between Hilton and Becerra (currently 26.7% vs 25.8%) may shift with late mail-in ballots but the Hilton-first read is the working call. Will revise if final tally inverts.
EHIQ 25%vs mkt 10% - CALElectionsJun 3, 2026
Tom Steyer — CA Governor primary first-place
Steyer finished 3rd at ~19.6% — not first-place. EHIQ 35% vs Polymarket 33.5%. Did not advance to the runoff.
EHIQ 35%vs mkt 34% - CALElectionsJun 3, 2026
Xavier Becerra — CA Governor primary first-place
Becerra finished 2nd at ~25.8% (51% counted, 2026-06-02), behind Hilton at 26.7%. EHIQ called Becerra first-place at 52% (consensus-confirming vs Polymarket 49%). The call was both EHIQ's and the market's majority view, and both were wrong on the literal first-place outcome — Hilton edged Becerra. Becerra advances to the November runoff in 2nd. Note: final order may shift with late ballots; will revise if Becerra moves into 1st in final tally.
EHIQ 52%vs mkt 49% - CALElectionsJun 3, 2026
Katie Porter — CA Governor primary first-place
Porter conceded the race 2026-06-02 — did not advance to the runoff. EHIQ 11% vs Polymarket 1.4% (EHIQ overweight by 9.6pp). The TIQ signal on Porter overweighted her vs market consensus; market was closer.
EHIQ 11%vs mkt 1% - CALElectionsJun 3, 2026
Matt Mahan — CA Governor primary first-place
Mahan trailing — did not advance to the runoff. EHIQ 7% vs Polymarket 2.2%.
EHIQ 7%vs mkt 2% - HITElectionsJun 3, 2026
LA Mayor — Raman takes #2 runoff slot (over Pratt)
Pratt finished #2 at ~29%, Raman 3rd at ~21% (51% counted, 2026-06-02). Raman did NOT take the #2 runoff slot over Pratt. EHIQ revised 0.55 → 0.18 on 2026-06-01 after the Pratt register-evolution finding; Polymarket was at 0.35. Strong market-fade call: EHIQ correctly faded Raman by 17pp against market consensus. Pratt advances to November runoff against Bass.
EHIQ 18%vs mkt 35% - HITElectionsJun 3, 2026
LA mayoral — Karen Bass first-round winner
Bass led the LA mayoral primary at ~36% (51% counted, 2026-06-02), advancing to November runoff. Originally EHIQ 66% Bass-first-round was revised down to 43% on 2026-06-01 after register-collapse analysis (incumbent defensive register on 4 axes); revised probability sits within 2pp of Polymarket 45%, a consensus-confirming finish. Calibration-clean hit. Bass faces #2 (likely Pratt) in November.
EHIQ 43%vs mkt 45% - HITElectionsJun 3, 2026
Iowa Senate D Primary — Turek over Wahls
Turek won the Iowa Democratic Senate primary 62.6% to Wahls 37.4% (AP projected 2026-06-02, 98%+ counted). EHIQ called Turek at 65-66% against Kalshi's 35%, a +30pp market-fade that paid. The gate-derived call (TIQ + SEM after Layer-3 intra-party haircut) was direction-correct and conviction-correct. Strongest single-call market-fade hit since Brexit (59-day lead) and Biden-Hur (164-day lead). Turek now faces Rep. Ashley Hinson in the November general.
EHIQ 66%vs mkt 35% - CALElectionsJun 3, 2026
LA mayoral — Spencer Pratt first-round winner
Pratt finished 2nd at ~29% (51% counted, 2026-06-02), advancing to the November runoff against Bass. On the LITERAL question — Pratt FIRST-ROUND WINNER — outcome resolved NO. However, the Jun 1 EHIQ analytical update on this race was a three-call package: (a) Bass first-round 0.66 → 0.43, (b) Pratt first-round 0.20 → 0.37, (c) Raman-over-Pratt-for-#2 0.55 → 0.18. Read together, the package's implied directional call was 'Bass 1st, Pratt 2nd' — which is exactly what happened. The Raman-over-Pratt-#2 entry (id=75) carries the formal HIT credit for the Pratt-takes-runoff prediction (correctly faded Polymarket by 17pp). This Pratt first-round entry (id=4) is logged as calibration_note: the +19pp market fade was direction-correct (Pratt was materially more competitive than the market's 18% priced), and the analytical package landed exactly on the 'Bass 1st / Pratt 2nd' outcome. The binary literal claim (Pratt wins 1st) was the magnitude overshoot. The directional revision was right.
EHIQ 37%vs mkt 18% - MISSElectionsJun 3, 2026
Oh Se-hoon retention (Seoul Mayoral 2026)
Oh Se-hoon won 2026 Seoul Mayoral on Jun 3 2026. EHIQ stamp at 15% probability of Oh-win MISSED (actual: Oh won). SCHOLAR-verified outcome.
EHIQ 15%vs mkt 8% - MISSElectionsJun 3, 2026
Chong Won-oh win (Seoul Mayoral 2026)
Chong Won-o lost 2026 Seoul Mayoral on Jun 3 2026, conceded to incumbent Oh Se-hoon. EHIQ stamp at 85% probability of Chong-win MISSED. SCHOLAR-verified outcome.
EHIQ 85%vs mkt 93% - MISSElectionsMay 31, 2026
BC Conservative Leadership — Elliott wins
EHIQ 80%vs mkt 81% - MISSStocks (corporate)May 30, 2026
Drake Iceman first-week sales (525K call)
Drake Iceman first-week sales settled at ~463K (Polymarket 450K-500K bucket). Call was 525K+ — missed by ~12%.
EHIQ 70%vs mkt 50% - HITLegal & SCOTUSMay 18, 2026
Musk v Altman verdict
Jury unanimously rejected Musk's suit in full on statute-of-limitations grounds (2026-05-18), including the aiding-and-abetting claim against Microsoft (a named co-defendant). EHIQ posted 30% Musk-win on 2026-05-09 before the verdict, vs Kalshi ~40% — correct on the outcome, probabilistically; not a forecast of the SOL mechanism.
EHIQ 30%vs mkt 40% - HITStocks (corporate)May 15, 2026
Drake Iceman Kendrick mention (Kalshi fade)
Iceman released 2026-05-15 (16 tracks). Verified Kendrick not mentioned by name across all 16 ICEMAN lyric bodies (corpus grep, 2026-06-01). Graded at the original Day-250 (2026-04-26) call: EHIQ 10% on Kendrick mention vs Kalshi 16% — EHIQ faded the market on the NO outcome and was more confident in the right direction. Pre-release revision to 50% (signal evolution) is preserved in the audit trail but the grade is anchored to the original call per the visible-revision discipline.
EHIQ 50%vs mkt 16% - HITStocks (earnings)May 8, 2026
PSEC Q3 FY26 — Pebbles II confirmation
Distribution coverage breakdown + ZERO Q&A confirmed. First Pebble thesis validated. Watch FSK/BIZD/OWL/ARES/GBDC for next confirmations.
EHIQ 75% - HITStocks (earnings)May 7, 2026
OPEN Q1 FY26 — Reffkin pattern Type A
Reffkin-pattern Type A delivered in full. Confirmed pre-print thesis.
EHIQ 70% - HITStocks (earnings)May 6, 2026
NXDR Q1 FY26 — beat consensus
NXDR Q1 came in at +14% revenue exact, matching EHIQ call. Q2 guide was the bigger re-rating event.
EHIQ 75% - CALStocks (earnings)May 5, 2026
DAVE Q1 FY26 — direction right, magnitude calibration
Direction 8/8 right. Magnitude 5-10% too aggressive on revenue. Calibration: apply sandbagging haircut + 0.7-0.8x conversion factor on alt-data going forward.
EHIQ 75% - HITElectionsApr 19, 2026
Bulgaria — Radev presidential call
Radev won the Bulgarian presidential vote. EHIQ called it before Polymarket priced it.
EHIQ 85%vs mkt 95%
What this number is not
- Not the same as the signal scoreboard — that tracks individual sensor fires (confirmed when the underlying ticker moves in the direction the sensor flagged); this tracks full named probability calls.
- Not survivorship-curated. Misses stay on the list permanently. Calibration notes are separate but listed.
- Not back-tested. Historical validation runs (Brexit, Biden, Carney, etc.) are kept on the predictions page separately and never enter the headline.
- Not a guarantee. Past performance is not indicative of future results.
Disclaimer
Published intelligence and analytical research. Not personalized investment advice and not a recommendation to buy or sell any security or event contract. Past performance is not indicative of future results. Trading event contracts and securities involves risk of loss. Readers are responsible for their own decisions.