Crazy Time strategy: what the arithmetic says

The rule that cannot be broken

Each spin is independent, and every bet has a negative expected value. The sum of negative expected values is negative, whatever order you place them in and however you size them. A staking plan changes the shape of the outcome, many small wins and rare big losses, or the other way round. It cannot change its average. With a house edge of 3.92 to 5.59 per 100 staked, 1,000 units wagered costs 39 to 56 units in expectation under any plan.

Why Martingale feels like it works

Doubling after every loss wins one unit each time the streak ends, and streaks end quickly: after a loss, the next bet wins 38.9% of the time. So most sessions are a string of small wins. The catch is the tail. A run of 7 losses, which takes a 127-unit hole and a 128-unit next stake, is 3.2% on any given sequence but 72.9% somewhere in 100 spins. At 8 losses it is 53.4%. Table limits and your bankroll will usually stop the sequence before it recovers.

The expected loss is unchanged. The plan trades many small wins for a few very large losses, and the average remains negative.

What does help

Three things reduce what you lose, none of them a system. Stake less per spin, because the expected cost scales with the stake. Play fewer spins, for the same reason. And choose the cheaper bets: on Crazy Time the published range is 94.41% to 96.08%, so the bets at the upper end cost about 30% less per unit staked than those at the lower end (3.92 against 5.59). Set the budget before the session, and treat it as spent when it is gone.

If you want the details of how spreading across spots changes the variance but not the edge, read the bet-spread maths.