Every Tick vs Real Ticks vs 1-Minute OHLC: Why the Same EA Wins on One Modelling Mode and Loses on Another
31 August 2026, 01:00
2
382
Two traders backtest the same EA over the same years and get opposite results: one report is profitable, the other loses. Usually nobody made a mistake. They chose different modelling modes, and the modelling mode decides what price data the EA actually saw.
What the three modes actually do
"Every tick based on real ticks" replays tick history recorded by the broker — the closest thing to what happened inside each bar, including the gaps and bursts real feeds have. "Every tick" (without real ticks) generates synthetic ticks from one-minute bars using an interpolation pattern; the path inside each minute is invented, smooth, and friendlier than reality. "1 minute OHLC" checks only four prices per minute, and everything between them simply does not exist. Each step down is faster to run and further from the market.
Which strategies the choice breaks
Anything decided inside the bar is exposed. A scalper with a tight stop lives or dies on whether price touched a level between minute closes — generated ticks routinely miss touches that real ticks hit, and OHLC misses almost all of them. Pending orders near the market, breakeven moves, tight trailing stops, and take-profits a few points away are all in the sensitive group. An interpolated path can also resolve the classic "which was hit first, SL or TP inside one bar" question in the EA's favour when the real feed resolved it against it.
Which strategies it mostly does not break
An EA that evaluates on bar close and uses stops many multiples of the spread away is far less sensitive: the prices it acts on are real in every mode, because bar opens and closes are real everywhere. Slow trend followers on H1 or H4 often produce broadly similar results across modes. That is not a virtue of the mode — it is a property of the strategy, and it is worth knowing which kind you own before you trust any report.
A backtest without its mode stated is not a claim you can check
A published equity curve that does not say which modelling mode produced it cannot be reproduced, and a result you cannot reproduce is marketing, not measurement. Our own testing standard is every tick based on real ticks, and we have retired ideas that looked profitable on generated ticks and fell apart on real ones — that failure is common, and any vendor who has never seen it is not testing hard enough. When you evaluate any EA, ours included, run the free demo yourself on real ticks before believing anything.
An example where the modelling mode is stated with the results: https://www.mql5.com/en/market/product/187329
All products and free utilities: https://www.mql5.com/en/users/app.develop.sk/seller
Our measured backtest data for every EA (profit factor, equity drawdown, trade count, year-by-year results) is published at fxea365.com/ea/ranking
