Ready to go further
A high win rate is not an edge
Check what the average trade earns after the rare big loss is included.
What is included
- How the bot works, step by step
- The arithmetic of the ladder or grid, with tables
- How it fails and how to test it honestly
- The tools that check the numbers
Who it is for
- People thinking of using this kind of bot
- People already running one and wanting a stress test
- Readers comparing bot types before buying
- Not advice, and not a recommendation to use one
Key facts
- Martingale means making the next position bigger after a loss, usually double, so that one win recovers every earlier loss plus one original unit.
- It wins small amounts often and loses a very large amount rarely. It does not improve the average result of the trades.
- With a 50 percent win rate, a run of 7 losses starts about once in 128 sequences, but over 1,000 trades it appears at least once about 98 percent of the time (exact calculation, 27 Oct 2026).
- Position size doubles each step, so margin and capital needs grow faster than an account can follow.
- Grid and recovery bots use the same idea in different clothes. Test them the same way.
What martingale trading is
The idea comes from betting. Bet one unit, and if you lose, bet two. If you lose again, bet four. The first win pays back everything lost so far plus the original unit. In trading, the bot opens a position, and if price moves against it, opens a bigger one in the same direction hoping for a small reversal that closes the whole stack in profit.
Because the target is only a small bounce, a martingale bot closes in profit very often. That is why a backtest of one can look like a smooth, rising line.
The doubling ladder
Starting at 0.01 lot and doubling after each loss, the sizes run like this:
| Step | Lots at this step | Total lots open | Compared with step 1 |
|---|---|---|---|
| 1 | 0.01 | 0.01 | 1x |
| 2 | 0.02 | 0.03 | 2x |
| 3 | 0.04 | 0.07 | 4x |
| 4 | 0.08 | 0.15 | 8x |
| 5 | 0.16 | 0.31 | 16x |
| 6 | 0.32 | 0.63 | 32x |
| 7 | 0.64 | 1.27 | 64x |
| 8 | 1.28 | 2.55 | 128x |
By step 8 the bot holds 2.55 lots against a first trade of 0.01. Every unit of price movement now costs 255 times what it cost at the start, and the margin needed has grown by the same factor. The guide on margin calls shows what that does to a small account.
Why losing streaks are not rare
People judge martingale by how unlikely a long losing run looks. A run of 7 losses on a coin flip is a 1 in 128 event. But the question is not whether a 7-run happens to the next 7 trades. It is whether one appears anywhere in the trades the bot will take. The table shows the exact chance, assuming independent trades and a 50 percent win rate.
| Losing run | Chance a given start is one | At least one in 100 trades | At least one in 1,000 trades |
|---|---|---|---|
| 5 losses in a row | 3.13% | 81.0% | over 99.9% |
| 7 losses in a row | 0.78% | 31.8% | 98.2% |
| 10 losses in a row | 0.10% | 4.4% | 38.5% |
Calculated by us on 27 Oct 2026 as an exact count, not a simulation. Real markets are not coin flips. Trends make runs longer, not shorter, so treat these as a floor.
What stops the ladder
A real account has limits the maths ignores. The broker caps position size. Free margin runs out and orders are rejected. The bot has a maximum number of steps. When any of those arrives mid-ladder, the bot takes the loss on the whole stack, and that loss is far bigger than every small win it collected on the way. Many accounts that look healthy for months end this way in a single event.
Close relatives of martingale
- Grid bots place orders at fixed price steps and add as price moves away. Sizes may stay equal, but exposure still stacks up in a trend.
- Recovery and zone recovery open an opposite position to claw back losses, usually with a larger size.
- Averaging down and DCA add to a losing position at lower prices. Smaller steps, same direction of risk.
- Anti-martingale does the opposite: it sizes up after wins and down after losses.
How to test a martingale bot honestly
- Look at the single worst losing run in the test and the margin used at the deepest step, not only the final profit.
- Include spread, swap and slippage on every position in the ladder, not just the first.
- Test on a period the settings were not tuned on, then forward test on a demo account.
- Check what the bot does when an order is rejected or the broker caps the size.
- Run the average through the expected value calculator. If the average is negative, the ladder only delays the result.
Tested results of our own will appear here when they are ready, with the data, dates and costs shown. None are published yet, and until they are, no figure on this page is a result from a bot.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
Does martingale work?
It changes the shape of the results, not the average. It wins small amounts often and loses a large amount rarely. If the underlying trades have a negative average, the ladder cannot fix it, and if the average is positive the ladder adds risk the strategy did not need.
Is martingale illegal?
It is not illegal in itself. Brokers, exchanges and prop firms can restrict or ban it in their own rules, so read the terms of the one you use.
Is a martingale EA safe?
No strategy is safe, and martingale concentrates the risk in rare large losses that a short backtest may never show. Ask for the worst losing run and the margin at the deepest step before trusting one.
What is the difference between martingale and grid?
Martingale increases size after a loss. A grid places orders at fixed price intervals and usually keeps sizes equal, though many grids also scale. Both add exposure as price moves against the bot.
What is anti-martingale?
Sizing up after wins and down after losses. It keeps losses small and lets winners grow, at the cost of giving back gains when a streak ends.
Are the numbers on this page real?
Example numbers are made up to show how the maths works, and the page says so. A figure only counts as exact when the page says it was calculated on a stated date. Nothing you enter in a calculator leaves your browser.
Can I use this to decide whether to run a bot?
It can show whether a bot's numbers hold together and how much an account can lose. It cannot say what the market will do, and it is not advice. Use it to ask better questions, then read the terms of your own broker or exchange.
Is this financial advice?
No. It is general education about how a sizing method works. It does not take your situation into account.
Know the Risk Before You Run It
Free calculators and guides, added every week. Leave your details and we will email you when new ones go up.
Trading carries a risk of loss