Counts, classifications and test results produced on this site. Every number here comes with the method used to get it.
Research
We optimised MoonDog EA on XAUUSD M15 over one window and then ran the winning inputs, untouched, over a later one the optimiser never saw. The fitted window returned +28,742.7%; the unseen window returned +165,035.3%.
· 8 min read
Research
We optimised Quantum Titan Mt5 on XAUUSD M5 over one window and then ran the winning inputs, untouched, over a later one the optimiser never saw. The fitted window returned +461.0%; the unseen window returned +3,900.1%.
· 8 min read
Research
The fitted window returned +1,311.1%. On data the optimiser had not seen, the same inputs returned +301.3% — still positive, but far enough down that the first number was mostly fit.
· 8 min read
Research
You do not need the source code. A martingale or grid system has to expose itself in its input parameters, in the shape of its equity curve, and in which statistic the vendor chose to advertise — and any one of the three settles it.
· 8 min read
Research
Modelling quality measures how the strategy tester guessed the price path inside each bar. It is a statement about the data, not about the strategy — and on MT4 the number stops at 90% because the tester is inventing ticks from M1 bars.
· 7 min read
Research
A count, not an opinion. Of 1,180 expert advisors currently on sale, 880 rely on martingale, grid, averaging or loss-recovery logic with no hard stop. And the products making the most attractive claims are the most likely to be in that group.
· 8 min read