Reading the win rate as the risk
A martingale cycle is defined to end on a win, so nearly every cycle wins. The win rate hides what the failed cycle costs: the account.
also: martingale sizing, doubling strategy, recovery multiplier
A position-sizing rule that multiplies the lot after each loss, usually by two, so that one win repays the whole losing run plus the original target. In an MT5 EA it shows up as a lot-multiplier input and as volume that rises after losing deals.
Lose, then trade twice as big; lose again, twice as big again. The next win pays for everything before it. It works every time until the losing run is longer than the account can fund, and then it takes the account.
Martingale is sold on the numbers it flatters: long stretches of small, regular gains and one loss the size of everything. That shape looks best in a short record and worst in a long one.
A $5,000 account, 0.01 lot on EUR/USD doubled after every loss, with a 20-pip stop worth $2 at the base lot.
At even odds, any trade has a 1 in 2,048 chance of opening eleven straight losses: rare per trade, routine across a long record.
Calculation 0.01 × 2^11 = 20.48 lots · (2^11 − 1) × $2 = $4,094 lost before trade 12
Result Eleven ordinary losses leave $906, and the recovery trade's stop costs $4,096
Use the term as a filter: read the inputs and the sizing rows of the tester report before the win rate.
| Range | What it means |
|---|---|
| One lot size, equity drawdown within about 1.3× balance drawdown | Fixed-lot sizing; all 18 current listings here sit inside this band. |
| Multiplier above 1.0 with a level cap | Base lot × multiplier^cap is the position to fund. |
| Volume rising after losing deals | Martingale, whatever the description says. |
| Multiplier with no stop and no cap | The loss is bounded only by the account. |
All 18 EAs in the current catalogue declare a fixed lot and no recovery method, and each publishes its trade ledger.
A martingale cycle is defined to end on a win, so nearly every cycle wins. The win rate hides what the failed cycle costs: the account.
A cap turns an unbounded loss into a large, known one. With doubling and a ten-level cap, the capped run loses 1,023 base losses; size the account for that.
A 1.3 multiplier survives longer runs, but after ten losses its win repays the run only when it earns about three times what a same-size loss costs.
Martingale is a sizing rule, so the evidence sits in the rows about size and open risk, not in the win rate. Every EA currently listed here trades one fixed lot, so their ledgers show those rows when size never changes.
| MT5 report row | 18 current fixed-lot ledgers | What a doubling ladder does |
|---|---|---|
| Volume column of the Deals table | One lot size per ledger over 5,966 closes | Rises after every loss, resets after a win |
| Equity Drawdown Maximal ÷ Balance Drawdown Maximal | 1.03–1.30 across all 18 | Held legs sink equity while balance counts closes only |
| Largest loss trade ÷ Average loss trade | 1.28–6.65, median 1.94 | The ninth loss in a row is 256 times the first |
| Maximum consecutive losses | 2 to 13, median 5.5; 10 or more on 3 of 18 | After 10 losses the next entry is 1,024 times base |
Two ledgers top their bands without any ladder. Matins reaches 6.65 from one GBP/JPY buy on 2020-03-19 that lost 151.9 pips at 0.1 lot. Halyard reaches 1.30 because it has no stop and can hold a buy and a sell at once. The Deals table settles any outlier.
Martingale shares its add-to-a-loser mechanics with grid trading; the difference is whether the added position is larger. The consecutive-loss row is the worst losing streak, and a ladder that runs out of margin ends in a margin call. The martingale strategy page replays a doubling ladder over published trades to show where it breaks.