projects/decisions
0%
load trades.csv (1,247 entries · 90d)
Parsed 1,247 fills across 38 tickers. [ok]
analyze decision_quality
--window 90d
Thought for 1.2s · clustering reaction times
edge_per_trade →
+0.42R (p < 0.01)
flag revenge_trading after losses > $200
47- if (loss > 200) { trade.next(immediate) }
47+ if (loss > 200) { coolingPeriod(15m) }
3 patterns surfaced — review in dashboard
Build
fills → edge → pattern → flag
Explore →
Decisions
A trading platform that decodes the psychological patterns inside your decisions.
Most trading tools tell you what happened. Decisions tells you why you keep doing it. Upload your trade history and the system clusters every fill into behavioral patterns: revenge trading after losses, late-entry FOMO chasing, holding winners too long because of the dopamine, sizing up after a streak.
Each flagged pattern comes with the exact moment it fired, the cost in basis points, and a cooling-off rule you can adopt. The goal isn’t to write more code over the chart, it’s to make the trader holding the mouse a little less predictable to themselves.