Tick_Forge/market/FORUM.txt

Free tool: test an Expert on the same real ticks with a different calendar and price level

A perfect backtest proves nothing if the Expert only knows the past it was tested on. Some Experts carry a memory of that history: a list of dates, a table of times, hard-coded price levels. On the original data they look brilliant. Live, there is no list for the future.

I built Tick Forge to check for that. It makes a custom symbol from your own broker's history, with every timestamp shifted and every price passed through a function you type (p*2, p*0.5, p0*p0/p, a time offset in hours or days). You then test the Expert in the normal Strategy Tester.

What it does

- Real broker ticks, same order, same spreads. Only the date and the price level change.
- Audit set in one click: control, +7, +14, +28 days, prices x0.5, x2, mirrored. It asks first and can delete the set again.
- Free runner script in the repo that runs your Expert across the set and writes a comparison report.
- Free, MIT licence, no DLL, no internet, never trades.

What it does not do

- It is not a Monte Carlo of your trade list and it does not invent random price paths.
- It cannot start the Strategy Tester from inside MetaTrader (MQL5 does not allow it), hence the separate script.
- A pass does not prove an Expert is safe or profitable. A fail is a question, not a verdict: some honest Experts depend on sessions or weekdays.

Quick example: a 7-day shift keeps every weekday and time of day, only the date changes. An Expert driven by price and time of day trades much as before. An Expert that relies on a list of dates loses its trades.

Product: [PRODUCT_URL]
How-to article with the tests and an example: [ARTICLE_URL]
Source: https://forge.mql5.io/joshuadunnink/Tick_Forge

I would like feedback. Which transformations would you add? Which Experts did you find depended on dates or price levels?