1,232 hypotheses in, 15 proven — we publish the kill decisions.
An independent research operation. Every hypothesis that dies gets a dated page, the evidence, and the methodology that survives it. No live parameters, no signals, no advice.
Free: The Fill-Model Audit — 5 checks before you trust any backtest.The procedure that has killed more of our strategies than any market event. One afternoon, no signup.
Latest from the ledger
- We Watched 5,225 Crypto Bar Markets Settle Across Two Venues. One of Them Lies About Its Order Book.Eleven days of read-only monitoring across Polymarket and Kalshi paired crypto bar markets: a 3.9% settlement-disagreement rate, a 14x gap in displayed-liquidity honesty, and apparent cross-venue margins that shrank 3x while we watched.
- The Second Blowup: Six Failure Modes the Audit Didn't EnumerateThe sim said $1.69M. Live said −$253.08 across 22 bars. Postmortem of the eebd-replica — six failure modes, five of them new — and the checklist v2 they bought.
- The $83 Mirage: Paper Said Profit, Live Was Dead in Eight MinutesPostmortem of a market-making bot whose paper simulator filtered out exactly the fills that killed it live — and the five-step fill-model audit that came out of the wreckage.
Get the kill decisions by email
New postmortems and ledger entries, as they're published. No spam, no tracking pixels of ours — it's Substack's form.
More: the kill ledger · the audit offering · about