Archive
Rejected experiments
Every entry is a completed experiment that failed its predefined criteria. Filter by market, open any record for the full method and numbers, or take the whole set.
Showing 1–10 of 45 experiments
| ID | Experiment | Market | Mechanism | Horizon | Result | Verdict |
|---|---|---|---|---|---|---|
| E904 | After a head of state's posts, crypto shows no reliable follow-through event-study, news, political-posts, bitcoin, ethereum, gold-token, continuation, unscheduled-news | Crypto & DeFi | Public intent | 15m-4h | +67.9 bp in 2025, -5.8 bp in 2026 | KILL |
| E889 | Options-expiry weeks and Fridays no longer beat a normal S&P 500 day equity-index, sp500, calendar-anomaly, options-expiry, day-of-week, publication-decay | Equities & volatility | Calendar | 1d | +0.3 bp and +3.1 bp per day | KILL |
| E888 | Buying the S&P 500 after a VIX spike rests on three crisis days equity-index, sp500, vix, panic-reversal, liquidity-provision, overnight | Equities & volatility | Forced flow | 1d | +9.4 bp, carried by 3 days | KILL |
| E886 | A patient maker resting 2% off the index loses once fills are verified market-making, liquidity-provision, adverse-selection, fill-model, thin-markets, perpetuals | Crypto & DeFi | Forced flow | 15m-1h | -14.2 bp dollar-weighted per fill | KILL |
| E878 | Protocol fee growth showed no detectable next-day edge protocol-fees, fundamentals, cross-sectional, long-short, perpetual-futures, alt-tokens, daily | Crypto & DeFi | Public intent | 1d | +4.3 bp/day net, interval includes 0 | KILL |
| E877 | Large exchange deposits do not predict price drops on-chain-flow, exchange-deposits, perpetual-futures, alt-tokens, event-study | Crypto & DeFi | Public intent | 4h-24h | -41.9 bp, same as control | KILL |
| E876 | Stocks that gap at the open do not reverse during the day stock-perpetuals, overnight-return, retail-attention, opening-auction, intraday-reversal, cross-section | Equities & volatility | Behavioural | 1d | -15.2 bp gross per day | KILL |
| E872 | Round-number price levels cannot be told apart from placebo levels round-numbers, stop-orders, liquidation-clusters, perpetual-futures, placebo-levels, hourly-bars | Crypto & DeFi | Forced flow | 1h-4h | -252.5 bp; round = placebo | KILL |
| E871 | Lending-rate spikes do not pay a short after costs lending-rates, constrained-shorts, spot-margin, perpetual-futures, alt-coins, event-study | Crypto & DeFi | Forced flow | 4h-24h | +3.4 bp net, interval -33.7 to +37.0 | KILL |
| E859 | No detectable lead from Korean memory-chip moves to US chip perps cross-market, read-through, stock-perpetuals, korea, semiconductors, lead-lag | Equities & volatility | Behavioural | 1h-24h | Slope +0.077, t +0.54 | KILL |