Why winning scalps seem to fill worse than losing ones

[Amanhasnoname]

Plenty of traders swear that the fills on winners are always a touch worse than the fills on losers. Sometimes that is paranoia, but there is a real mechanism behind it. Many liquidity providers use a last look window, a few milliseconds where they can accept or reject an order after seeing it. If price has moved in the trader's favor during that window, the order is more likely to get rejected or filled at a revised level.

The effect is invisible on a swing trade where a 0.3 pip difference is noise. On a 3 pip scalp it is ten percent of the whole target.

Pull the last 100 trades and split them into winners and losers. Compare the difference between the requested price and the executed price for each group. If winners show a consistently larger negative gap, the broker's liquidity stack is probably shaving them.

One decent sample proves nothing, though. Anything under a hundred fills per group is mostly luck.

Has anyone actually run this comparison on their own statements, and did the gap show up or not?