Roulette Statistics from 4.0M Spins
Every spin played on Roulette Simulator since 2011 has been recorded. The numbers below are the live picture of our European wheel over the trailing six months — recomputed each day, no smoothing, no editorial selection.
Window: 19 February 2026 → 18 August 2026 · Last refreshed: 18 August 2026 · Source: solo + multiplayer + tournament play, European wheel only
Red, Black, Green
On a fair European wheel, red and black should each land 18 times in 37 spins (48.65%), and green (zero) should land once in 37 (2.70%). After 4.0M spins, here's what we actually see:
Within a couple of hundredths of a percent of where a fair wheel says they should be. That gap will shrink further the longer the dataset runs — that's how randomness works at scale. Over a hundred spins you can easily see red 60% of the time. Over a million you can't.
Consecutive Outcomes
The longest streaks ever recorded on a single wheel — one continuous run of spins on the same solo game, multiplayer table, or tournament. These are record runs, not the everyday norm: on a fair wheel each extra repeat is roughly half as likely as the last.
Frequency by Number
Each of the 37 pockets on a European wheel should hit 2.70% of spins in the long run. The bars below show our observed rate; the small number to the right shows how far above or below the fair-wheel expectation each pocket has landed.Most-hit so far: 18 (2.72%, +0.013 pp) · Least-hit: 21 (2.68%, -0.022 pp)
None of the deviations above are statistically meaningful — they're well within the noise band you'd expect from 4.0M independent draws. The "most-hit" number isn't due, isn't lucky, and isn't rigged. It's just the one that randomness happened to favour by 0.013 percentage points this window.
Dozens and Columns
Each dozen and each column covers 12 numbers, so a fair wheel should show each one landing 12/37 of spins (32.43%). The remainder is the green zero.
Dozens
Columns
Even / Odd and Low / High
The classic even-money outside bets. Each side covers 18 numbers, expected 48.65%; the green zero loses to all four.
Even vs. Odd
1–18 vs. 19–36
What These Numbers Actually Mean
The temptation when looking at a stats page like this is to treat the numbers as tea leaves. The 7 has come up 0.04 percentage points more than expected. Should I bet on it next? No — it's the wrong way around.
A roulette wheel has no memory. The ball doesn't know what landed last spin, never mind what landed across the last hundred million. Every pocket has the same odds on every spin: 1 in 37 on a European wheel, 1 in 38 on American. The long-run distribution you see above isn't causing future spins to balance out. It's the consequence of a billion independent draws all happening to follow the same fair probability.
That's the gambler's fallacy in one paragraph: assuming a number is "due" because it hasn't appeared lately. Or that it's "hot" because it has. Neither is true. The wheel resets to 1/37 on every spin, and over enough spins that's exactly the rate each pocket converges to.
What this page does tell you, honestly: our wheel behaves the way a fair wheel should. Every pocket has landed within a fraction of a percentage point of the theoretical 2.703%. Red and black are within rounding of 48.65% each. Dozens and columns are within rounding of 32.43%. After a billion spins, that's the picture of a clean RNG.
If you want to see the same convergence on smaller samples, run an auto-spin session of 500 spins on the European table. Watch the red/black ratio bounce around 48% in early hundreds and settle down as the count grows. That's the lesson the long-term data here is teaching, only faster.
For the maths behind why the casino still wins despite a fair wheel, see the payouts guide — the short version is that 35-to-1 payouts on a 1-in-37 event is the gap that becomes the house edge. Honest wheel, dishonest odds.