Experiments with trading technology and AI, measured against simple baselines.
Part 1 · Part 2 · Part 3 · Part 4 · Part 5 · Part 6 · Part 7 New: Part 7 · 02 Nov 2026
I read the code of 19 Jev finance projects, from jarrodwatts/jev-trader to QuantDinger. Most are demos; 4 tested Jev against a simple rule and none won.
A controlled experiment on a daily gold breakout strategy, 2021-2026, 400 backtests. Random entries with good management break even. The strategy's own timing with a coin-flip direction is profitable in 100 of 100 seeds. Turn management off and every run on real ticks loses. Where the profit of a breakout system actually sits, with a backtest-only EA to repeat it.
Part 1 · Part 2 New: Part 2 · 10 Oct 2026
I asked Jev "will this resolve Yes?" for 26,174 resolved Polymarket markets. On the 9,138 resolved after its release it scored AUC 0.615. The market price a day earlier scored 0.844. Code included.