Is that track record luck?
“68 % winning calls.” “7 of the last 10.” Type in any record you’ve been shown and see how often pure luck does just as well. It takes the win rate, the number of calls, and how a win is counted.
- Exact binomial maths
- Runs in your browser
- Nothing sent or stored
Type a record on the left, or try one of the examples. The answer shows how often a random record, with the same number of calls, does at least as well.
Six ways a record fools you
Each one makes luck look like skill. Most published track records have at least two.
Too few calls
A coin gets 7 or more of 10 right 17 % of the time. At 60 %, you need about 140 calls before luck becomes a 1-in-100 explanation.
The targets do the winning
With take-profit 1.2 % away and stop-loss 4.5 % away, random entries hit the take-profit first 79 % of the time, and need that just to break even.
The best of many
If 500 channels post calls at random, a few of them will have a great record. You hear about those, not the other 497.
The missing calls
A record with the losing calls quietly deleted, or edited after the fact, can show any win rate. No calculator can fix that. Only a record locked before the outcome can.
Win rate isn’t profit
Many small wins and a few large losses can lose money at 80 %. Ask for the average win, the average loss and the costs, not the win rate.
Calls that count twice
Ten calls on the same move in the same hour are one bet, not ten. Overlapping calls make a record look bigger and more certain than it is.
Every call, locked before the outcome
A track record is only evidence if every call was written down before its outcome, none were removed, and it beats a simple baseline over enough calls. That is what we test for a living. Try it on a record of your own with Lock your prediction, or run the seven tests on a signal file with the Quick Check.
The maths: the chance that a random record with the same number of calls gets at least as many wins (a one-sided binomial test). For take-profit/stop-loss, the base rate is stop-loss ÷ (take-profit + stop-loss), the chance a random walk hits the nearer target first; real prices drift and have fees, so treat it as a close guide. “Calls needed” uses the normal approximation for a 1-in-100 result at the same win rate.