Median Win vs Average Win: The One Stat That Tells You What a Slot Mode Feels Like

Published 2026-07-14 · Tags: rtp, math, slots

The average is the RTP and no spin pays it. The median is what a typical spin returns, and for most modes it's zero. Now on every Slot Cruncher mode card.

Slot Cruncher got a new stat this week, one number per mode: the median win. I think it's the most honest number on the page and I want to explain why it's different from the one everyone already looks at.

The average lies by telling the truth

A mode's average payout is just its RTP. A 96% mode pays back, on average, 0.96x the stake per spin. Everybody knows this and nobody feels it, because no single spin pays 0.96x.

The average is being dragged in two directions at once. Most spins pay zero. A few spins pay enormous amounts. The average sits somewhere in between, at a value that almost never actually occurs. It describes the game over a million spins; it says nothing about what spin number 47 is going to feel like.

The median is what a spin feels like

The median win is the 50th percentile of the outcome table. Half of all spins pay less than it, half pay more. For most slot modes the median is zero: more than half of spins pay nothing at all. That's not a criticism, that's how slots are built, but it's worth seeing written down next to the RTP.

Where it gets interesting is the modes where the median isn't zero. A cluster-pays game with a high hit rate might have a median of 0.2x: a typical spin gives you back a fifth of your stake. A bonus-buy mode might have a median of 15x against a buy cost of 100x: the typical bonus returns a seventh of what you paid for it. Those are the numbers that tell you what the grind is going to feel like, and the RTP hides them completely.

Two modes, same RTP, different median

Here's a real comparison from the catalog (names withheld because the point isn't the games):

Mode A pays something on nearly four spins in ten and a typical spin returns a bit of your stake. Mode B pays nothing on four spins in five, the typical spin is a total loss, and the RTP is being held up by a tail that reaches ten times further. Same average. Completely different evening.

How we compute it

Stake Engine games publish their full outcome tables, so this isn't a sample. We sort every outcome in the mode by payout, weight by probability, and read off the value at the 50% mark. It's exact. For bonus-buy modes we divide by the buy cost so the median is in units of "what you paid".

How to use it

Look at median alongside hit rate. If the hit rate is 30% and the median is 0x, the game pays often but tiny; those hits are mostly below your stake. If the median is above zero at all, the game is genuinely giving back on a typical spin, which is rare and worth knowing.

And if you're choosing between two modes with the same RTP, the median is the question "which one will I enjoy losing 4% on". It's in every mode card on Slot Cruncher now, and in the Payout DNA strands as a marked point.