— hunchlog
blog
may 06 · 2026
Brier scores, for people who don't trade for a living
Every weather forecaster has a Brier score. Most retail traders don't, and they think they're better at this than they are. The math is high-school simple. The implications aren't.
may 06 · 2026
Paper trading is a coping mechanism
Paper trading tests whether you can click buttons in a simulator. It does not test whether your reads on the world are any good. Those are different muscles, and only one of them compounds.
may 06 · 2026
What a trading thesis actually looks like
Most 'trade theses' online are a chart screenshot with arrows on it. A real thesis has four parts — claim, mechanism, horizon, falsifier — and the missing piece is almost always the last one.
may 06 · 2026
What 'I called it' actually requires
The phrase 'I called it' is everywhere in retail finance. Half the time it's true, and half the time the brain is filling in the parts the prediction was missing. Here are the four conditions that separate them.
apr 30 · 2026
Why your PnL doesn't tell you if you can trade
PnL is an outcome metric, not a skill metric. A trade journal that captures the thesis — before the trade — is what separates lucky calls from repeatable ones.
apr 30 · 2026
When 'just buy the index' stops being the safe answer
The 2010–2021 case for SPY-and-chill rested on conditions that are visibly breaking. Most investors are already hedging. They just have no track record of whether their hedges work.