Sortino = (mean return − target return) / downside deviation. Same shape as Sharpe but it doesn't count upside volatility as risk.
Why this matters: trading P&L isn't symmetric. Big winners are a feature, not a bug. Sharpe treats a great month and a terrible month as equally 'volatile;' Sortino only cares about the terrible ones.
Sortino is usually higher than Sharpe for the same strategy by a factor of 1.3–1.6. Comparing your Sortino to other traders' Sharpe is apples to oranges — pick one metric and stay consistent.
Like Sharpe, Sortino needs sample-size discipline: <30 trades and the number is mostly noise.