Before you judge any strategy, learn the arithmetic of the product it is traded on. On a fixed-payout trade the arithmetic is simple, and it explains most of what happens to people who trade it.
The setup
You stake an amount. If you are right, you get the stake back plus the payout, which is a fixed percentage of the stake. If you are wrong, you lose the stake. If the price ends exactly where it began, the stake is returned.
The payout for each market is shown on the order ticket before you confirm. Use that figure, not a number from memory, because it can differ between markets and be changed by the operator. The examples below use 95% and then show other values.
Break-even win rate
How often must you win just to end level? Let the payout be p (0.95 for 95%). Over many trades, you break even when:
win rate × p = (1 − win rate) × 1
Solving gives:
Break-even win rate = 1 ÷ (1 + p)
| Payout | Break-even win rate |
|---|---|
| 95% | 51.3% |
| 90% | 52.6% |
| 85% | 54.1% |
| 80% | 55.6% |
A payout under 100% therefore means you must win more than half your trades merely to stay level. A prediction method that is right 50% of the time, no better than a coin, loses steadily. This gap is the cost of trading, the equivalent of a spread, and it is the first thing a serious trader checks.
Expected result per trade
For one trade with stake 1, win probability w and payout p:
Expected result = w × p − (1 − w)
With p = 0.95:
- At w = 50%: 0.475 − 0.500 = −0.025, a loss of 2.5% of the stake per trade on average.
- At w = 55%: 0.5225 − 0.45 = +0.0725, a gain of 7.3% of the stake per trade.
- At w = 60%: 0.57 − 0.40 = +0.17.
So the line between losing and winning is thin, and sits at a win rate that is hard to reach and hard to prove. Note that staking a larger amount does not change the sign of these numbers. It only magnifies them.
Luck is large in the short run
Suppose your true win rate is exactly 50%. Over 100 trades the number of wins is not 50; it varies. The standard deviation of the win rate over n trades is roughly:
√(w × (1 − w) ÷ n)
At 50% and 100 trades that is about 5 percentage points, so ordinary luck produces anything from about 40% to 60% wins. A 60% record over 100 trades is entirely consistent with having no edge at all. Over 30 trades, the spread is about 9 points. That is why the admin analytics on this platform flag any rate based on fewer than 30 decided orders.
A rough guide: to be reasonably sure that a win rate is above break-even by 3 or 4 points, you need hundreds of trades, not dozens.
Losing streaks are normal
With a win rate of 50%, the chance of losing ten times in a row at some point is not small if you trade often. In 1,000 trades, a streak of nine or ten losses is typical. If you win 55% of the time, streaks of eight or nine losses still appear.
This matters for two reasons. First, a streak is not evidence that something is broken. Second, your stake must be small enough to survive one. That leads directly to position sizing, the next lesson.
Compounding losses
Losses and gains do not cancel evenly. If you lose 50% of your balance, you need a gain of 100% to get back. Losing 20% needs +25%; losing 10% needs +11%. The deeper the hole, the steeper the climb, which is why avoiding large losses matters more than finding large wins.
What the maths tells you to do
- Know the break-even rate for the market you trade, and track your actual win rate against it, with the number of trades beside it.
- Prefer higher payouts when you have a choice, since each point lowers the bar.
- Do not trade more often just because you can. More trades with no edge only means a faster expected loss.
- Treat a good result as unproven until the sample is large.
- Stake small, because variance is not optional.
Common mistakes
- Quoting a win rate without the number of trades behind it.
- Believing a win rate of "about 50%" is acceptable. At any payout under 100% it loses.
- Increasing stakes after a few wins because the results feel like proof.
- Doubling the stake after a loss to recover it (the "martingale"). It works until one long streak, then costs everything, and trading limits eventually stop the doubling anyway.
Note: These are statistics about averages. They describe what happens over many trades, not what will happen on the next one, and none of this is a forecast of your results.
