Recency bias is the tendency to overweight what just happened and treat it as representative of the whole. In trading, it means the last few trades — or the last hour — quietly hijack your judgment about a system built on hundreds of outcomes.

Two losses ≠ a broken system

After a couple of losses, recency bias screams that your strategy is broken, tempting you to abandon a sound process or tinker with it mid-stream (consistency over being right). But any positive-edge system has losing streaks — they're expected variance, not evidence of failure (risk of ruin). Quitting a good system at its normal drawdown is how traders serially rebuild the wrong wheel.

A hot streak isn't skill

The flip side: a few recent wins feel like proof the market is easy, feeding overconfidence and oversizing (overconfidence). Recent results — good or bad — are a tiny sample dominated by luck (expected value).

Your last three trades are a rounding error in your edge. Judge the system by the hundred, not the handful you can still feel.

The fix

Evaluate a strategy over a large sample, not a session — track expectancy across dozens of trades, not the last few (your equity curve, keeping a journal). A mechanical system is immune by design: it doesn't feel the last loss or the last win — it executes the same rules on trade 200 as on trade 1 (mechanical vs discretionary).