A Losing Streak Tells You Nothing
Why a run of losses carries no information about what comes next, and why it feels like it does.
Five in a row is not unusual
Between two evenly matched players, losing five duels in a row has a probability of one in thirty-two. Play a hundred duels in an evening and you should expect to hit a run of five at some point; it would be more surprising if you did not.
Losing eight in a row is one in 256. Across a large enough population of players on any given day, some of them will. That is not evidence of anything being wrong. It is what randomness looks like from close up.
The mistake is not noticing the streak. The mistake is concluding that it means something about the next round.
Why the next round does not care
Each round draws its blast point fresh. The formula is applied to a new random value with no reference to any previous round, no accumulator, and no state carried between rounds — the derivation is set out on the fairness page.
There is no mechanism by which a loss could make a win more likely afterwards. Not because we have chosen not to build one, but because the derivation has nothing in it that could hold such a memory — the inputs are the seed, your client seed, and a counter, and none of them encodes how the last round went.
You can check this. Take two settled rounds from a session and verify each independently. Neither computation refers to the other. That is the whole design.
The feeling is real and it is not information
The sense that a win is due has a name — the gambler's fallacy — and knowing its name does not stop you feeling it. People are built to find patterns in sequences, and a run of losses is exactly the kind of sequence that triggers the machinery.
It is worth separating two things that get confused here. The belief that outcomes must even out in the short run is false. The observation that they even out in the long run is true, but it is true in a way that does not help: the long run is long, and averages converge because later results swamp earlier ones, not because earlier ones get repaid.
A hundred duels from now, your losing streak will have been diluted, not compensated.
Why this is the expensive moment
The reason this matters more than an ordinary statistical curiosity is what people do about it. The instinct after a run of losses is to raise the stake, to recover the deficit faster. That instinct is precisely backwards.
The expected cost of a duel is roughly half the rake on the pot. Raising your stake raises the pot, which raises the rake, which raises what each duel costs you on average. Chasing losses does not shorten the recovery; it accelerates the loss.
It also concentrates risk at the worst possible moment — when you are making decisions under the pressure of already being down, which is not when anyone makes their best ones.
What to do instead
Decide before you start what you are willing to lose, and treat that number as the cost of the evening rather than a float you are trying to protect. Once you are down, the decision about when to stop is being made by a version of you with a strong reason to get it wrong.
If you find yourself reasoning that a win is due, that is a good moment to stop for reasons that have nothing to do with the odds. The belief is false, and acting on it costs money at exactly the rate the arithmetic predicts.
This is a real-money game for adults. Over time the rake means an evenly matched player loses money. Stake only what you can afford to lose, and stop when it stops being a game.