Immersive Roulette Live Stats & Results: Reading 240,000 Spins From a Dealer-Run Wheel

Immersive Roulette is Evolution's single-zero European wheel presented by a live dealer, cut between multiple HD camera angles and finished with a slow-motion replay of every ball drop. Our live tracker has logged 240,626 of its rounds since 28 December 2025. This page explains what each stored round contains, how to read every panel on the live page, and what 88,756 spins from the last 90 days actually show. For rules and bet spots, start with the Immersive Roulette guide.

What makes Immersive Roulette's data different from every other wheel we track?

The pace. Every other part of the wheel behaves like a standard single-zero table, but Immersive Roulette is the only one in our tracker where a real dealer sets the cadence.

A round here takes about 47 to 49 seconds from betting open to settlement, and a new round starts roughly every 52 to 55 seconds, based on the 957 most recent rounds that carry feed timestamps. That is dealer-driven, so it drifts by a few seconds from round to round rather than landing on a fixed clock. Compare that with Evolution's other standard tables:

Table	Round-to-round cadence	What sets the pace
Speed Roulette	about 25 s	fastest cycle, minimal presentation
Auto Roulette	about 51 s	automated launch, no dealer
Immersive Roulette	52 to 55 s	dealer pace, camera cuts, slow-motion replay

Fewer spins per hour than a fast automated table means a lower hourly cost at the same flat stake, since the house edge only bites once per completed spin. It does not change the edge on any single bet: every spot still returns 97.30%, whatever the clock says. The dealer, the camera work and the replay are worth watching for their own sake, but none of it is information the strategy page would ask you to use.

What does the tracker store for each dealt round?

One winning number as text, its color and type, and a timestamp, taken from the feed once the round settles and saved within minutes.

The result field holds a number from "0" to "36". The result_details field adds the color (red, black or green) and the type (Even, Odd, or None for zero). The multiplier column exists on every game in our database for consistency, but Immersive Roulette never uses it: our collector writes 1.00 on every row because there is nothing to multiply. If you see a claimed "average multiplier" for this game anywhere, it is not a real figure; there are no multipliers on this table. Bonus fields are likewise always null or false, because there is no bonus round.

One thing the tracker deliberately does not store: the feed also reports how many players shared a round and the total amount they were paid, but we do not save either figure, so no widget on the live page will show a player count or a payout total. Everything you see is built from the winning number alone.

Field	Immersive Roulette value	Notes
result	"0" to "36"	Winning number as text
result_details	color, type	Even, Odd or None for zero
multiplier	1.00 on every row	No multipliers exist in this game
bonus fields	null or false	No bonus round exists
created_at	UTC timestamp	Basis for windows and the hourly chart

How do I read each panel on the live page?

Each panel on /games/immersive-roulette describes past spins; none of them predicts the next one.

Live results strip. The most recent winning numbers with their color, newest first. Good for confirming a result you just watched land in the video feed.

Number and color shares. Observed frequency for each pocket and for red, black and zero, with the gap from theory shown in points. Over a short window this jumps around; over 88,756 spins it settles close to flat, which is what the table below shows.

Sector heat. The same 37 numbers grouped by their physical position on the wheel instead of by number, using the Voisins, Tiers and Orphelins racetrack sectors. This is the panel most likely to expose a genuine mechanical lean, because a tilt favors an arc rather than scattered numbers.

Streak and repeat tracker. The current run of one color and the current run of a single number repeating, alongside the longest of each on record.

Zero and number drought. Spins since zero last hit and spins since any given number last hit, next to the average gap and the longest gap in the window. It tells you how long the current wait has run, nothing more.

Probability calculator. Enter a number and a spin count and it returns the chance of at least one hit, using the fixed 1-in-37 probability. This is the one panel that is genuinely forward-looking, because it uses theory rather than history.

Hourly heatmap. Round volume by hour of day in UTC. Useful for knowing when the table is busiest; not useful for predicting an outcome, since the dealer deals at the same odds around the clock.

Which numbers have run hottest and coldest over 90 days?

Number 18 led with 2.85% of 88,756 spins, 2,527 hits, and number 22 trailed at 2.60%, 2,310 hits. Every one of the 37 pockets landed inside that narrow band around the 2.70% theory.

Rank	Number(s)	Observed share (90 days)
Hottest	18	2.85%
	19, 16, 35, 1	2.78%
	10, 3, 28	2.76%
Zero	0	2.66%
	20, 24, 14, 4	2.63%
	5	2.61%
Coldest	22	2.60%

The gap between the hottest and coldest number is 217 hits across 88,756 spins, a difference of well under a tenth of a percentage point once expressed as a share. That is the size of gap a fair wheel produces by chance when 37 pockets absorb nearly ninety thousand spins; it is not the signature of a lean or a dealer habit. The hot and cold ranking on the live page reorders itself as the window moves, and that reshuffling is expected, not suspicious. Our fairness index applies the same reasoning across every game we track.

How do color, parity and dozen bets compare with theory?

Red landed 48.79%, black 48.55% and zero 2.66% over 90 days, with the group bets all sitting within half a point of their 32.43% or 48.65% theory.

Bet	Theoretical share	Observed (90 days)
Red	48.65%	48.79%
Black	48.65%	48.55%
Zero	2.70%	2.66%
1-18	48.65%	48.83%
Odd	48.65%	48.89%
1st dozen	32.43%	32.36%
2nd dozen	32.43%	32.46%
3rd dozen	32.43%	32.51%
Column 1	32.43%	32.56%
Column 2	32.43%	32.24%
Column 3	32.43%	32.54%

These figures should not be read as an observed return. Every bet on this table pays true odds minus the 1-in-37 edge, so the return is fixed at 97.30% by rule, and a share of 48.79% against a 48.65% theory is ordinary sampling noise rather than a hint that the edge has drifted. The RTP page covers how that 97.30% figure is calculated and how our RTP monitor tracks it across games.

What do the Voisins, Tiers and Orphelins sectors show?

Grouped by wheel position rather than number, the three racetrack sectors held close to their published shares over 90 days: Voisins du Zero 46.15%, Tiers du Cylindre 32.39%, Orphelins 21.46%.

Sector	Numbers covered	Theoretical share	Observed (90 days)
Voisins du Zero	17	45.95%	46.15%
Tiers du Cylindre	12	32.43%	32.39%
Orphelins	8	21.62%	21.46%

Sector shares matter for a different reason than the number table above. Because the racetrack groups cover contiguous arcs of the wheel rather than a scattered set of pockets, a mechanical tilt would show up here first, as a persistent excess on one arc rather than one lucky number. A gap of 0.20 points on Voisins is well inside normal variation for a sample this size. A full walkthrough of the racetrack bets and how the sector, streak and hot/cold widgets fit together lives on the streaks, sectors and hot/cold page.

How long can a color streak or a repeated number run?

Fourteen. That is both the longest red run and the longest black run recorded in the last 90 days, with the longest run of consecutive zeros at three.

The average run of a single color is close to 1.95 spins across roughly 22,200 red runs and 22,200 black runs, so a run of fourteen is a genuine outlier, and it is also exactly the sort of outlier a fair wheel throws up over tens of thousands of spins. The chance that a streak reaches k spins from a given start is (18/37) raised to the power of k minus 1; for fourteen reds that works out to about 0.0086% per starting spin, which multiplies out to several expected occurrences across 88,756 spins. Consecutive zeros follow the same logic at (1/37) squared for two in a row, about 0.073%, and roughly 0.002% for three, which lines up with the single three-zero run we recorded.

Repeats behave the same way. The same number landed twice in a row 2,263 times over 90 days, 2.55% of spins against a 2.70% theory, three times in a row 67 times, and the longest identical run reached four. None of that changes the next spin's odds: the chance that any given number repeats is 1 in 37 regardless of how many times it has already repeated. The strategy guide works through why betting systems built on chasing or fading a streak lose to the same 2.70% edge either way.

How rare is a 300-plus spin drought, and has one happened here?

Not rare at all once you account for how many numbers are watching for one. The longest zero drought in 90 days ran 344 spins, and the longest drought for any single number reached 467.

Spins without a hit	Probability for one number	What it means at this scale
37	36.3%	Routine, several times a day across 37 numbers
100	6.5%	Several times a week, per number
200	0.42%	A few times a quarter, per number
344 (our longest zero drought)	roughly 0.001%	Rare for one number, expected somewhere across 37
467 (number 24, our longest overall)	roughly 0.0003%	Same logic, larger sample needed

The last column is the point. With 37 numbers each capable of a long absence and 88,756 spins in the window, a 300-plus spin gap for at least one of them is close to a certainty, not a coincidence worth chasing. Zero's drought gets the most attention because it is the number every roulette player watches, but the same math applies to 24, which sat out for 467 spins, and to 12 and 27, which had their own long gaps of 394 and 388 spins. None of it shortens the wait: the odds of zero, or 24, or any other number landing on the very next spin stay at 1 in 37 no matter how long the current gap has run. The belief that a long absence makes a hit more likely is the gambler's fallacy, and our responsible gambling observatory tracks exactly this pattern across every game because it is where chasing tends to start.

Does the hour of day or the dealer's pace change the numbers?

No. Round volume varies by hour, but the odds do not. Over the last 30 days, 14:00 and 15:00 UTC were the busiest hours, at roughly 2,010 spins each, followed by 10:00, 13:00, 17:00 and 12:00.

Because the dealer table runs at a near-constant cadence around the clock, those hourly counts differ only marginally; a busier hour mostly reflects the feed's polling pattern rather than a rush of extra players. What the pace does affect is cost per hour at a fixed stake: at roughly 65 to 70 spins an hour, a flat bet loses on average stake times 2.70% times the number of spins played, so the slower dealer cadence here means fewer of those losses per hour than on Speed Roulette and a similar count to Auto Roulette. It does not change what a single spin costs on average.

Where does the data fall short?

Two gaps are worth knowing before you lean on these numbers. The collector was offline from 2026-06-30 to 2026-07-09, only 40 rounds logged in that stretch, which sits inside the current 90-day window and slightly understates spin counts for that period without biasing the shares.

Second, pace figures (the 47 to 49 second bet-to-settlement window and the 52 to 55 second cadence) come from only the 957 most recent rounds, the ones that carry the feed's own start and settlement timestamps. Older rows only have our insert timestamp, batched from periodic polling rather than the exact moment of each spin, so streak and drought lengths from that era are accurate to within the order of a batch rather than to the second. Neither limitation changes the long-run number and color shares reported above, since gaps and batching are random with respect to which number lands.

How should I use this data at a dealer's table?

As a record of variance, not a forecast. Every bet on Immersive Roulette returns 97.30% regardless of what the wheel just did, the dealer has no more influence over where the ball lands than an automated launcher does, and the slow-motion replay is entertainment rather than information about the next spin.

Knowing that a color can run fourteen deep, or that zero can hide for over 300 spins, is useful for one thing: sizing a bankroll to survive ordinary variance rather than reading a system into it. Decide on a stake and a number of spins before you sit down, treat the tracker as a log rather than a signal, and step away if a session starts to feel like chasing a drought. Support for anyone who finds that hard is linked from our observatory page.

Frequently asked questions

Does a live dealer change the odds on Immersive Roulette?

No. The wheel is the same single-zero European layout used across Evolution's roulette tables, so every bet keeps a 2.70% house edge and a 97.30% return. The dealer, the multi-camera cuts and the slow-motion replay change the pace and the presentation, not the physics of the ball. Our 90-day data, with every number inside a quarter point of theory, reflects that.

How often does zero hit on Immersive Roulette?

Zero landed on 2.66% of 88,756 tracked spins over the last 90 days, close to its 2.70% theoretical share. The average gap between zeros was 37.5 spins, with a median of 25. The longest wait was 344 spins, ending 2026-06-07, with two more long droughts of 307 and 238 spins in July. The zero drought widget on /games/immersive-roulette shows the current gap live.

What is the longest number drought recorded on Immersive Roulette?

Number 24 went 467 spins without landing, a drought that ended 2026-08-17. Number 12 went 394 spins without hitting, ending 2026-07-14, and number 27 went 388 spins, ending 2026-08-11. With 37 numbers and 88,756 spins in the window, a gap this long for at least one number is expected, not evidence the wheel favors or avoids any pocket.

How is Immersive Roulette's data different from Auto Roulette's?

Both are single-zero wheels with a 97.30% return and no multipliers, so their long-run number, color and sector shares converge to the same theory. The difference is pace and presentation: Immersive Roulette runs on a dealer-driven cycle of about 52 to 55 seconds with a slow-motion replay of the ball drop, while Auto Roulette launches automatically roughly every 51 seconds with no dealer at all.

How do I check today's Immersive Roulette results?

Open /games/immersive-roulette and set the window to 24 hours; the results strip, number shares and streak panels then cover only today's spins. Timestamps in our data run on UTC, so a session played in the evening in Europe may fall under the previous UTC day. The dealer table runs continuously, so there is no separate "today" results page to publish.