Advertisement IC Markets

Win-rate vs expectancy on EURGBP: post anonymized stats

Document your personal trading journey. Track daily equity curves, review winning and losing streaks, share trade screenshots, and get constructive feedback.
Post Reply
LondonScalper
Posts: 965
Joined: Sat Sep 05, 2026 7:54 am

Win-rate vs expectancy on EURGBP: post anonymized stats

Post by LondonScalper »

EURGBP is the pair where my win rate and my expectancy tell completely different stories, so I thought it would make a good example for a thread on the difference. Here are my numbers. If you trade the pair, I'd like to see yours in the same format.

Period: January to September.
Trades: 143.
Win rate: 64%.
Average winner: 0.71R.
Average loser: 1.02R.
Expectancy: about plus 0.05R per trade, after costs.

That 64% looks great, and it's my highest win rate of any pair. But the expectancy is my lowest of any pair I still trade, and it's barely positive.

The reason is in the averages. EURGBP's small ranges mean my targets are small, so winners are modest. When it does move against me, it often moves quickly, so losers hit the full stop or worse.

If I compare EURUSD over the same period: win rate 48%, average winner 1.34R, average loser 0.98R, expectancy about plus 0.13R. Lower win rate, more than twice the expectancy.

If I judged by win rate, I'd trade more EURGBP. Judging by expectancy, I should trade it less, or only at its best times, which for me is 08:00 to 10:00 London.

Would you post yours? Format: period, trades, win rate, average win in R, average loss in R, expectancy. No account sizes needed.
Recommended broker for automated trading & scalping IC Markets
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Win-rate vs expectancy on EURGBP: post anonymized stats

Post by Shadow Trader »

Happy to share mine in the same format, and the attached chart is an illustration of how my EURGBP and EURUSD equity curves compare in R, with synthetic trades built to match these numbers.
img_13024.png
img_13024.png (105.17 KiB) Viewed 200 times
EURGBP, January to September: 96 trades, win rate 61%, average win 0.68R, average loss 0.95R, expectancy about plus 0.05R after costs.

EURUSD, same period: 158 trades, win rate 46%, average win 1.38R, average loss 0.97R, expectancy about plus 0.11R.

So almost the same pattern as yours. The curves make the point better than the averages. EURGBP looks smoother for months and then gives back a chunk in a few trades, while EURUSD is bumpier but ends well ahead.

Your 08:00 to 10:00 window idea is where I'd look first. When I split my EURGBP trades, the ones before 10:00 had an expectancy close to EURUSD, and the afternoon trades were negative after costs.

What's your average cost per trade in R on EURGBP? Mine is about 0.08R, which is twice my winner's margin.
Fairman
Posts: 6955
Joined: Tue Jul 21, 2026 7:11 am
Location: Abuja

Re: Win-rate vs expectancy on EURGBP: post anonymized stats

Post by Fairman »

LondonScalper wrote: That 64% looks great, and it's my highest win rate of any pair.
One number that shows how thin that EURGBP edge is: the break-even win rate. With an average winner of 0.71R and an average loser of 1.02R, you break even at 1.02 divided by 1.73, about 59%. So the 64% has a five-point cushion. Over 143 trades a win rate carries an uncertainty of about four percentage points either way, so the true rate could plausibly be close to break-even already, and a small rise in costs or one extra slipped stop a month would also eat it. EURUSD's break-even, 0.98 against 1.34, is about 42% against an actual 48%, so a similar cushion in points but a bigger edge per trade. Adding the break-even rate as a line next to the actual one would make the format in this thread even more useful.
It’s Fairman :geek:
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Win-rate vs expectancy on EURGBP: post anonymized stats

Post by Shadow Trader »

Fairman's break even line is a good addition, so here are mine with it. EURGBP breaks even at 0.95 divided by 1.63, about 58%, against my actual 61%, while EURUSD breaks even at about 41% against an actual 46%.

The sample size point hits my EURGBP harder than the original numbers, since I only have 96 trades. That gives an uncertainty of roughly five points either way, so my three point cushion sits fully inside the noise.

The break even rate also shows cost sensitivity nicely. If I add 0.03R of extra cost per trade, taking it off the average winner and adding it to the loser, EURGBP's break even rises almost two points, while EURUSD's rises a little over one.

So the pair with the thinner cushion is also the one that loses it fastest when spreads widen. That fits my morning and afternoon split, where the wider afternoon spreads were what pushed EURGBP negative.

For the format, I'd add one line after break even: average cost per trade in R. With those two side by side, anyone can see how much spread widening an edge can survive.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Win-rate vs expectancy on EURGBP: post anonymized stats

Post by Shadow Trader »

A strategy shows a 55% win rate over 100 trades. Is it better than a coin flip? Not necessarily.

The standard error of a win rate near 50% is about the square root of 0.25 divided by n. For 100 trades, that's 5 percentage points. A 55% result is one standard error above 50%, well within normal luck. To be reasonably confident, say two standard errors away from 50%, the standard error needs to be 2.5 points, which takes about 400 trades.

That calculation assumes a win rate matters in isolation. With unequal win and loss sizes, the relevant quantity is expectancy, whose standard error depends on the spread of results. Strategies with large variance in trade outcomes need even more trades.

The lesson is that most traders judge strategies on far too little data. Fifty trades says almost nothing. Two hundred says a little. Several hundred begins to say something, provided conditions varied.

Because scalpers trade a lot, that sample is achievable within weeks. The temptation is to draw conclusions sooner.

Anyone wanting a rule of thumb can remember this: to detect a 5-point edge in win rate with reasonable confidence needs hundreds of trades, to detect a 2-point edge needs thousands. That's why small edges are hard to verify, and why claims based on a few dozen trades should be treated as stories, not evidence.

How many trades are required before a strategy gets real money or a real verdict?
FXS Prop Desk
Posts: 1689
Joined: Wed Oct 07, 2026 9:52 am

Re: Win-rate vs expectancy on EURGBP: post anonymized stats

Post by FXS Prop Desk »

The break-even win rate calculation is a great way to show how thin the edge is, Fairman. A useful next step would be to see how the break-even point changes with a small increase in costs. If an extra tenth of a pip per trade pushes break-even close to the actual win rate, the edge is very sensitive to the broker's spread, and it might be worth testing the strategy on a different account type before scaling up.
FXS Official Team | The Idea Lab & Strategy
Post Reply