How the Streak Math Works
Single run: P = (1 - winRate)^N for a losing run of N, or winRate^N for a winning run.
Over T trades: P(at least one run of N) = 1 - (1 - p^N)^(T - N + 1)
Expected worst streak: the largest N whose in-sample probability is still 50% or more.
The over-T formula is a close approximation (it treats the runs as independent windows), not an exact Markov result. It is accurate for the trade counts traders actually use, and this tool reports the same numbers the previous version did.
Streaks Are Normal: the Math
A win rate is a long-run average, not a schedule. A 55% edge does not deal you wins and losses in a tidy alternating order; it scatters them, and that scattering produces runs. Over 100 trades at 55% you are more likely than not to see a 5-loss streak, and the expected worst run sits near 6 in a row. None of that means the edge has failed, it is exactly what a 55% process looks like up close.
The uncomfortable part is that higher win rates do not make streaks disappear. Even a 75% strategy carries a real chance of a 5-loss run over 100 trades, and a 30% win rate strategy that is highly profitable at wide reward to risk will hand you 10-loss runs regularly. Plan for the streak your win rate implies rather than the one you hope for.
Sizing So Streaks Cannot Kill You
Once you know your expected worst streak, sizing is arithmetic. If the worst run is 8 losses and you refuse to give back more than half the account to a normal streak, your risk per trade cannot exceed about 6%, and most traders should sit well under that. The max risk value above does this division for you so you can size to survive the drawdown, not just the average trade.
This is why streak math and position sizing belong together. A position size that looks fine against a single loss can still be fatal against a run of them, so size against the streak the sample will actually produce. Use the position size calculator to turn that risk percentage into a concrete lot size for each trade.
Tilt: the Real Cost of a Streak
The largest danger in a losing streak is rarely the drawdown itself, it is what the streak does to your decisions. After four or five losses the temptation is to size up to win it back, to widen stops, or to abandon the rules that gave you the edge in the first place. Each of those turns a normal, survivable run into an account-ending one.
The defense is to decide the plan before the streak, not during it. If the math above shows a 6-loss run is ordinary at your win rate, then a 6-loss run is not new information and does not justify a new plan. A systematic, repeatable entry model keeps you from improvising at the exact moment improvising costs the most.
