← All calls
ElectionActive★ featured
Olivia Chow re-election (Toronto Mayor 2026)
Will Olivia Chow win the 2026 Toronto mayoral election?
premise verified 0h ago
EHIQ
33%
(64-72%)
Edge
-37.5pp
EHIQ below market
The call vs the market, dated
EHIQ called this at 65% on May 10, 2026 when Polymarket was at 77%. (open reconstructed from the dated thesis record; the original stamp is preserved unedited)
Revised to 33% based on new analysis — see thesis updates below for revision history.
Market has since moved down to 71% · -6.5pp from open
EHIQ vs market over time · 53 data points · refreshed daily
EHIQMarket
Thesis
FLIPPED 2026-10-12 (stamped): 62% → 33% — the CBC debate (Oct 8) was Chow's worst night of the cycle on the language read: confidence 5.9→3.0, certainty 3.4→1.7, desperation 4.1→6.2, self-repair 0.7→2.2, momentum −0.34 against Bradford's +0.83 (margin −1.17, 4× the no-signal band). The caption-artifact control says it was the performance, not the instrument. The spot model maps the flip to Bradford 0.653; Chow carries the complement at 0.33 after the residual tail. The October polls (weighted +6.2 Chow, narrowing) are the lagging indicator the instrument claims to lead — the falsifier is stated on the Bradford card and holds here: if Chow's poll lead holds ≥ +4 through the final wave, this flip was wrong at full size and reverts with the receipt. Resolves Oct 26.
- 1.PROBABILITY REVISION TIMESTAMPED 2026-05-22: EHIQ 65% → 68% after SEM v1.2 build (`sem-1.2-2026-05-22`). SEM v1.2 Chow probability = 70.45% — down from v1.1 96.86% (-26.4pp) after three mechanisms applied: incumbency_adj_pp 3.0 → 1.5 (Liaison "traffic frustration" + Miller-2010 precedent), Canadian municipal MAE 5.0 → 8.0pp (Toronto retrospective), withdrawal_prob_estimate 0.20 (Chow undeclared per "I don't know yet" / "I haven't wrapped my head around it" register). Per migration note: "model and market are now in agreement at ~70-75%. No call." TIQ specificity Chow 10.34 (AMBIGUOUS) not load-bearing. §4 OOS Phase C FAIL (ρ=+0.18, p=0.35) — TIQ research-internal. Range tightened to 64-72% reflecting mid-window narrower uncertainty (5 months to October).
Opened
May 10, 2026
Expected resolution
October 26, 2026
Get an alert the moment this call resolves.
Every EHIQ call resolves on a date, in public. Subscribe free and we email you when this one grades, plus the weekly divergence leaderboard.