Here's the secret the gurus bury under chart patterns: trading is a probability game, and the math is what decides whether you survive. You can be right less than half the time and still get rich — or right most of the time and still blow up. The difference is expectancy and risk. This is "The Money."
Expected value — the only metric that matters
Every trade has an expected value:
EV = P(win) × Reward − P(loss) × Risk
If EV is positive over many trades, you make money. That's it. This is why a 35% win rate can beat a 70% win rate — if the winners are big enough. The break-even win rate for a given reward-to-risk (R:R) is 1 ÷ (1 + R:R):
- R:R 1:1 → you need a 50% win rate just to break even.
- R:R 1:2 → only 33%.
- R:R 1:3 → only 25%.
- R:R 1:4 → only 20%.
The higher your reward-to-risk, the lower the win rate you need. Quality of entry beats frequency of entry.
Risk management — the one rule that separates pros from gamblers
Risk only 1-2% of your account per trade. This single habit is why professionals survive long enough to let their edge play out and gamblers don't. Two reasons the math is brutal:
- Drawdowns are non-linear. Lose 30% and you need +42.8% just to get back to even. Lose 50% and you need +100%. Small risk keeps you out of the death spiral.
- Leverage amplifies the wrong thing. It doesn't just multiply gains; it shrinks the distance to liquidation. High leverage + a normal pullback = you're gone before your idea even had a chance.
Position sizing & the Kelly criterion
Sizing turns your edge into growth. The Kelly criterion mathematically maximizes long-run growth, but full Kelly is dangerous — a small error in estimating your win probability can wreck the account. The practical version: use a fraction of Kelly (half or less) and cap any single trade at ~3% no matter what the formula says. Bet too big and variance kills you before your edge saves you.
Do you even have an edge? The metrics
You can't feel whether a strategy works — you measure it:
- Profit Factor (PF) = gross profit ÷ gross loss. Above 1.5 = a real edge.
- Max Drawdown (MDD) = biggest peak-to-trough drop. Keep it under ~20%.
- Sharpe / Sortino = return per unit of risk. Sortino is better for trading because it only penalizes downside volatility — it doesn't punish you for big wins.
- Calmar = return ÷ max drawdown. Above 1.0 is healthy.
And you need a sample: ~30 trades for a first read, 50 for a pattern, 100 before you trust the numbers. Judging a strategy by its last three trades is recency bias, not analysis.
Thinking in probabilities
One trade is a coin flip; 50+ trades reveal the truth (the law of large numbers). The skill is updating your belief as evidence arrives — start with a base rate, then adjust up or down for each independent factor that's present. (That's Bayesian thinking. Turning it into a precise, weighted model is its own discipline; the principle is what matters here: more confirming, independent evidence → higher probability.)
Volatility, regime & correlation
- Volatility clusters — calm follows calm, chaos follows chaos. After a big shock, widen your stop and shrink your size (an ATR-based stop does this automatically).
- Regime — markets trend or mean-revert (the Hurst exponent measures which); a bull, flat, or bear regime should change how much risk you take.
- Correlation — in crypto, "diversifying" across 4 altcoins is an illusion: their correlation spikes toward 1.0 in a sell-off, so four positions behave like one. In a crash, everything falls together — size accordingly.
Stops & losing streaks
A stop is where your idea is proven wrong — not an arbitrary percentage. Set it with the trade and never move it against yourself; that one habit (moving stops) is the single most expensive mistake in trading. And accept that streaks are normal: at a 45% win rate, a run of 4-5 losses in 100 trades is essentially guaranteed. Protect yourself with pre-set limits — e.g. two stops in a day = close the terminal. A losing streak amplifies every bias at once; the answer is smaller and slower, not bigger and faster.
Key takeaways
- Expectancy (EV), not win rate, decides long-run profit.
- Risk 1-2% per trade — drawdowns are non-linear and leverage is unforgiving.
- Size with a fraction of Kelly, capped; never full Kelly.
- Measure your edge: PF > 1.5, MDD < 20%, over 50-100 trades.
- Never move a stop against yourself. Streaks are normal — pre-set your limits.
FAQ
What is expected value in trading?
Expected value (EV) is the average profit or loss of a trade over many repetitions: EV = probability of winning × reward − probability of losing × risk. A strategy with positive EV makes money over time even if it loses individual trades; it is the single metric that determines long-run profitability.
How much should I risk per trade?
The widely-taught rule is 1-2% of your account per trade. Because drawdowns are non-linear (a 30% loss needs a 42.8% gain to recover) and leverage shrinks the distance to liquidation, small per-trade risk is what keeps you in the game long enough for your edge to work.
What is the Kelly criterion?
The Kelly criterion is a formula for the position size that maximizes long-run growth given your edge and reward-to-risk. Full Kelly is too aggressive in practice because errors in estimating win probability can ruin the account, so traders use a fraction (half Kelly or less) and cap each trade at a few percent.
What is a good profit factor or Sharpe ratio?
A profit factor (gross profit ÷ gross loss) above 1.5 indicates a real edge. For risk-adjusted return, a Sharpe ratio above 1.0 is solid; the Sortino ratio is often more useful for trading because it only penalizes downside volatility, not big winning trades.
How many trades do I need to know if a strategy works?
Roughly 30 trades give a first read, 50 reveal a pattern, and 100 make the statistics trustworthy. Judging a strategy by its last few trades is recency bias — you need a meaningful sample to separate skill from luck.