Same strategy, two data sets, two different equity curves

[Shadow Trader]

A backtest on one-minute bars shows a smooth climb. The same logic run on real tick data produces something messier and, often, a good deal flatter. Neither run is a bug. They're measuring different things.

Bar data compresses everything that happened within a minute into four numbers. The path between them is gone. A strategy with a three-pip stop and a four-pip target can't be evaluated properly when a single bar spans six pips, because the order of the high and low is unknown.

Tick data preserves the sequence. It shows whether the stop or the target came first, how the spread behaved during the move, and whether the entry price existed at all.

For anything holding for minutes or longer, bar data is usually fine. For scalping, where stops and targets are a few pips, bar-based testing can be wildly optimistic.

The gap often shows up as a sharp difference in win rate. Strategies that "win" 70 percent of the time on bars may win 50 on ticks.

How big was the gap when a strategy was first tested on both kinds of data?