Raising stakes after a winning run

A run of wins is weak evidence of edge and strong evidence of variance. Increasing stakes on it maximises exposure precisely when the justification is thinnest.

Winning streaks of six or eight occur regularly at any realistic strike rate, and they feel qualitatively different from the losing ones only because of their direction. The mathematics does not distinguish them.

Increasing stakes as the bankroll grows is correct and is a different thing. The rule is the same percentage of a larger pot, applied on a schedule rather than in response to results.

Here is the catch. It is the mirror image of chasing and attracts none of the same suspicion, because it happens while everything feels good.

What to do with it. Change unit size only when the bankroll crosses a pre-defined threshold, never in response to a sequence of results.