Thinking in Probabilities: The Mental Shift That Changes Everything
Beginners ask whether a trade will win. Experienced traders ask what the odds and payoffs are. The shift from prediction to probability is central.
No setup wins every time. Even a good one might win 55 percent of the time. The outcome of any single trade is uncertain, and judging a decision by one result mistakes luck for skill, as discussed earlier.
Probabilistic thinking has practical consequences. It makes losses less personal, since they are an expected part of the distribution. It focuses attention on expectancy: win rate times average win minus loss rate times average loss, measured over many trades. And it supports consistent execution, because you take every valid setup, knowing that the edge appears only over a series.
Skipping valid setups after a loss, or taking only the ones that feel certain, destroys the statistics. The trades that feel most certain are often the ones where the crowd agrees, which means less edge.
To practice, state each trade in terms of odds: I expect this setup to win about 45 percent of the time, with an average win of 2R and loss of 1R, for an expectancy of plus 0.35R. After a loss, remind yourself that the plan anticipated 55 percent losses.
Think of yourself as running a casino, not placing a bet. The house accepts losses, because the edge plays out over volume.
Practical step: for your main setup, write the estimated win rate, average win, and average loss, and compute the expectancy.
Thinking in Probabilities: The Mental Shift That Changes Everything
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
Writing the expected loss rate before each session helps. When I've already written about half of these will lose, the losing trade feels like part of the plan. I take it a step further and write the expected range for the week. For example, with a 45 percent win rate and three or four trades a day, a week with only five or six winners out of fifteen is completely normal. Having that written down makes it much harder to panic and start changing things after a bad Tuesday. It also keeps good weeks in perspective. Eight winners out of ten is great, but it's probably not the new normal, so size stays the same.
It’s Fairman 
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
Skipping trades after a loss is the hidden killer. Your stats assume you take every valid setup. Pick and choose, and the numbers no longer describe what you're doing. Worse, the trades you skip after a loss are often the ones that would have won, because the setups that feel scariest after a loss are frequently the clean ones against the recent move. Try marking skipped valid setups in your journal with what they would have done. After a month, add them up. Most people find the skipped trades would have improved the results noticeably. That's usually enough to make taking every valid setup feel less optional.
It’s Fairman 
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
The shift to probability is central. A practical habit: before each trade, write down the expected outcome in terms of probability (for example, this setup wins about 55% of the time). It reinforces the idea that a loss is part of the expected distribution, which reduces emotional reactions.
Reviewing results in batches of 20 or more trades, rather than one at a time, also supports probabilistic thinking.
Reviewing results in batches of 20 or more trades, rather than one at a time, also supports probabilistic thinking.
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
The estimate of the win rate is itself a probability, and that part often gets skipped. With 40 trades at a 45 percent win rate, the standard error is close to 8 percentage points, so the real rate might be anything between roughly 30 and 60 percent.
At 2R winners and 1R losers, the break even win rate is 33 percent. That means the plus 0.35R expectancy from the example sits on an estimate whose lower end is close to zero edge, which is a good reason to keep size small until the sample grows.
Streaks are the other thing to write down in advance. At a 45 percent win rate, the chance of at least one run of 7 or more losers somewhere in 100 trades is about one in two, and over 200 trades a run of 8 is more likely than not.
Knowing that number before the run starts changes how it feels. A run of seven then reads as a normal event inside the plan, not a sign that the setup has stopped working.
With batches of 20, how do you treat one that comes in at 6 winners? For a 45 percent system about one batch in eight lands there or lower, yet it is hard not to react.
At 2R winners and 1R losers, the break even win rate is 33 percent. That means the plus 0.35R expectancy from the example sits on an estimate whose lower end is close to zero edge, which is a good reason to keep size small until the sample grows.
Streaks are the other thing to write down in advance. At a 45 percent win rate, the chance of at least one run of 7 or more losers somewhere in 100 trades is about one in two, and over 200 trades a run of 8 is more likely than not.
Knowing that number before the run starts changes how it feels. A run of seven then reads as a normal event inside the plan, not a sign that the setup has stopped working.
With batches of 20, how do you treat one that comes in at 6 winners? For a 45 percent system about one batch in eight lands there or lower, yet it is hard not to react.
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
The average win in that expectancy sum can be more fragile than the win rate. A 2R average often comes from a handful of large winners sitting on top of many modest ones.
Say 40 trades produce 18 winners and 22 losers at 1R. If 16 of the winners made 1.6R and two ran to 5.2R, the average win is exactly 2R and the expectancy is the plus 0.35R from the opening post.
Take those two trades out and the remaining 38 return 3.6R, about plus 0.09R each. Most of the edge then rests on two outcomes, and whether trades like those come again matters more than any of the other 38.
This is where the casino comparison stretches a little. A casino knows its odds exactly and they never change, while a trader estimates them from a short record, and the payoff side is usually the shakier half of that estimate.
Recomputing the expectancy with the best two or three trades removed is a quick stress test. Does anyone here judge a setup by that trimmed figure rather than the full average?
Say 40 trades produce 18 winners and 22 losers at 1R. If 16 of the winners made 1.6R and two ran to 5.2R, the average win is exactly 2R and the expectancy is the plus 0.35R from the opening post.
Take those two trades out and the remaining 38 return 3.6R, about plus 0.09R each. Most of the edge then rests on two outcomes, and whether trades like those come again matters more than any of the other 38.
This is where the casino comparison stretches a little. A casino knows its odds exactly and they never change, while a trader estimates them from a short record, and the payoff side is usually the shakier half of that estimate.
Recomputing the expectancy with the best two or three trades removed is a quick stress test. Does anyone here judge a setup by that trimmed figure rather than the full average?
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Thinking in Probabilities: The Mental Shift That Changes Everything
Developing a mental model of order flow means forming an idea of who is buying, who is selling, and how urgent each side is. Done well, it sharpens timing. Done badly, it becomes a machine for finding stories in random movement.
Overfitting happens when a model has so many exceptions that it explains every past bar and predicts none of the next ones. A trader who answers every failed signal with a new rule, such as hidden absorption or a trapped crowd, is usually fitting the past instead of learning from it.
Safeguards help. A model should make a statement that could be wrong, for example that a fast move through a level on rising activity tends to continue more often than not. That statement can be logged and checked across many cases. If the claim never fails, it probably says nothing.
Simplicity is another guard. Models with two or three observable inputs are easier to test than those needing a dozen subjective readings.
Spot FX adds a limit: no single venue shows total activity, so any picture of flow is partial. Holding models loosely, writing predictions in advance, and counting misses alongside hits keeps the thinking honest and keeps confidence proportional to evidence.
Overfitting happens when a model has so many exceptions that it explains every past bar and predicts none of the next ones. A trader who answers every failed signal with a new rule, such as hidden absorption or a trapped crowd, is usually fitting the past instead of learning from it.
Safeguards help. A model should make a statement that could be wrong, for example that a fast move through a level on rising activity tends to continue more often than not. That statement can be logged and checked across many cases. If the claim never fails, it probably says nothing.
Simplicity is another guard. Models with two or three observable inputs are easier to test than those needing a dozen subjective readings.
Spot FX adds a limit: no single venue shows total activity, so any picture of flow is partial. Holding models loosely, writing predictions in advance, and counting misses alongside hits keeps the thinking honest and keeps confidence proportional to evidence.