Lightning Roulette Live Stats: Hot Numbers, Multiplier Frequency and Droughts From 261,000 Spins

Lightning Roulette is single-zero European roulette with a twist: before each spin, one to five numbers are struck with multipliers of 50x to 500x, and a straight-up bet on the winning number collects that multiplier instead of the standard payout. Our Lightning Roulette tracker has stored every spin from the official results feed since 28 December 2025, 261,832 rounds as of 2026-08-27. This page explains what the tracker keeps for each spin, how to read the number frequency, multiplier, color and drought widgets, and where the data stops being useful. For the rules themselves, the Lightning Roulette guide is the place to start.

What does a Lightning Roulette round record contain?

Five things: the winning number, its color, whether it was odd or even, the list of Lightning Numbers struck that round with their multipliers, and a flag for whether the winning number was one of them.

That last field is what makes this game's statistics interesting. In standard roulette a round is one number and nothing else. Here every round also produces a small RNG draw (how many numbers were struck, which ones, and what multiplier each carried) before the ball is released, and we store the full draw, so the multiplier stats on the live page are built from every strike, not only the ones that paid.

The multiplier column in our database follows one rule. If the winning number was a Lightning Number, the column holds that number's multiplier (50 to 500). If not, it holds 1, meaning the straight-up bet paid the reduced 29:1. This is why a "lightning win" in our widgets means exactly one thing: the ball landed on a struck number.

For scale, the last 30 days contain 48,036 spins and the last 90 days 97,698. Volume is fairly even around the clock, with the busiest stretch at 06:00 to 09:00 UTC. Rounds appear in the tracker within minutes of the ball settling.

How often has each number landed?

Between 2.61% and 2.80% of spins over the last 90 days, for every one of the 37 numbers, against a theoretical 2.70%.

The number frequency panel lays this out as a strip of 37 bars, and the hot and cold panel sorts the same data by count. Here is what the extremes looked like across 97,698 spins.

Number	Hits (90 days)	Observed share	Theory	Difference
23 (hottest)	2,740	2.80%	2.70%	+0.10
10	2,737	2.80%	2.70%	+0.10
9	2,730	2.79%	2.70%	+0.09
0 (zero)	2,674	2.74%	2.70%	+0.04
30	2,566	2.63%	2.70%	-0.07
7	2,562	2.62%	2.70%	-0.08
18 (coldest)	2,551	2.61%	2.70%	-0.09

The expected count for any number in 97,698 spins is about 2,640. The hottest number sits 100 hits above that and the coldest 89 below it, which after nearly 100,000 independent 1-in-37 draws is ordinary sampling noise. A number at 3.5% over a 90-day window would be worth a hard look; 2.80% is not.

The window selector matters more here than on any other widget. At 24 hours each number has only around 40 hits to work with, so shares of 1.5% and 4% sit side by side purely by chance; at 90 days the bars flatten. Only the sample size changed between those views.

Do hot and cold numbers mean anything for the next spin?

No. Each spin is a fresh 1-in-37 draw, so number 18 was exactly as likely after a long miss as number 23 was after a hit.

The hot and cold panel is a useful way to see variance. It cannot separate a "trend" from noise, because in a memoryless process there are no trends, only samples that happen to be high or low. The observed 2.80% for number 23 will drift toward 2.70% as more spins arrive, not because the wheel corrects for it but because new spins at the true rate dilute the early surplus.

A second trap is specific to this game. Because Lightning Numbers are chosen by the RNG every round, some players track which numbers get struck most often. Even if that distribution were uneven, a strike does nothing unless the ball also lands on the number, so chasing a "frequently struck" number is chasing two independent events at once. The strategy page goes through why betting systems built on either idea do not move the house edge.

What do the Lightning multiplier stats show?

Most strikes are small. Across all Lightning Numbers struck in the last 90 days, 37.5% carried 50x and 42.0% carried 100x, so 79.5% of strikes were at the bottom two values.

The multiplier panel splits into three views. The first is how many numbers get struck per round.

Lightning Numbers per round	Share of rounds (90 days)
1	4.0%
2	33.6%
3	53.5%
4	6.9%
5	2.0%

Three struck numbers is the normal case, and the average works out at about 2.7 per round. Five-strike rounds are 1 in 50, and they are no more likely to pay for any particular number you have bet; they only widen the set of numbers that could pay.

The second view is the value assigned to each strike, regardless of whether the ball landed there.

Multiplier assigned	Share of all strikes (90 days)
50x	37.5%
100x	42.0%
150x	6.0%
200x	8.5%
250x	0.5%
300x	1.5%
350x	0.3%
400x	0.7%
500x	3.0%

The 500x value is assigned more often than 250x, 300x, 350x or 400x, which surprises people. The game can afford to show its headline number fairly often because it only pays when the ball also lands on that pocket. Seeing 500x on a number you bet is a 3% event per strike; collecting it needs the 1-in-37 landing on top.

The third view is what was actually paid, which is where the two probabilities multiply together.

How often does a straight-up bet actually collect a multiplier?

On 7.33% of spins. In the last 90 days the winning number was a Lightning Number 7,164 times in 97,698 spins, with an average gap of 13.6 spins and a longest gap of 150.

The arithmetic checks out. If k numbers are struck, the chance the ball lands on one of them is k divided by 37: 2.70% for a single strike, 8.11% for three, 13.51% for five. Weight those by the observed distribution of k and you get about 7.3%, which is what the tracker recorded. That agreement between the RNG draw and the physical ball is the roulette equivalent of a fairness test, and you can compare it across games on our fairness index.

Here is how the paid multipliers broke down.

Paid multiplier	Rounds (90 days)	Share of lightning wins	Share of all spins
50x	2,655	37.1%	2.72%
100x	3,449	48.1%	3.53%
200x	688	9.6%	0.70%
300x	123	1.7%	0.13%
400x	45	0.6%	0.05%
500x	204	2.8%	0.21%

First, 85% of lightning wins paid 50x or 100x. A player who remembers Lightning Roulette as "the 500x game" is remembering the rare tail. Second, a 500x payout landed 204 times in 90 days, once every 480 spins or so, and 105 times in the last 30 days. At 60 to 70 spins an hour that is a 500x somewhere on the table every seven to eight hours. On 2026-08-27 alone the tracker logged 500x on 9 and on 14, and 400x on 23, 29, 3 and 0. Big multipliers are not rare for the table; they are rare for you, because you also have to hold the straight-up bet on that exact number.

The remaining piece is what all of this costs. Straight-up bets pay 29:1 instead of 35:1 on the 92.67% of spins with no multiplier, and Evolution balances that reduction against the multipliers so the whole game returns 97.30%. The RTP page walks through the calculation; the Lightning Numbers and multipliers page covers the mechanics in depth.

What do the color and dozen splits show?

Red 48.71%, black 48.55%, green 2.74% over the last 90 days, against 48.65%, 48.65% and 2.70% in theory.

The color panel is the quickest sanity check on the page because the theoretical split is so simple: 18 red pockets, 18 black, one green. A difference of 0.06 points on red and 0.10 on black after 97,698 spins is sampling noise. The odd and even split, and the high and low split, follow the same 18-18-1 structure and land in the same place.

The dozen and column panel does the same job for 12-number groups. Each dozen covers 12 of 37 pockets, so its theoretical share is 32.43%, with the zero belonging to none of them. Over a long window the three dozens sit close to one another; over a 24-hour window one dozen will often lead by a few points and swap places the next day.

The caveat for all outside bets is that Lightning multipliers never touch them. A red bet on a round where the ball lands on a 500x red number pays 1:1, exactly as on any European roulette table. The color and dozen panels confirm the wheel is behaving; there is no multiplier route through outside bets.

How does the drought tracker work in roulette?

It counts spins since the last occurrence of an event and compares that count with the average gap and the longest gap on record. For Lightning Roulette the events are a lightning win of any size, a lightning win of 200x or more, a specific number, and each color.

The reference figures for lightning wins are 13.6 spins on average and 150 at the longest. For a single number the average gap is 37 spins, and in 97,698 spins some number somewhere will have gone well over 150 without landing. For a color the average gap is barely over two spins, and yet runs of eight or more of the opposite color are routine: the chance of eight straight losses on an even-money bet is (19/37)^8, or 0.48%, which sounds small until you realize a table producing 1,500 spins a day gives that event several chances daily.

The drought tracker is a variance display, not a countdown. A number at 120 spins without a hit is not closer to landing than it was at spin 1; the pocket does not know it has been missed. The same goes for a stretch of 30 spins without a lightning win, which the tracker will show a few times a week. Where the tracker earns its place is in expectation setting: if you know the longest recorded wait for any lightning win is 150 spins, you can decide before you sit down whether your bankroll can absorb a wait like that on 29:1 base payouts. The strategy page turns those gap figures into session sizing.

How do I use the probability calculator?

Pick an event and a number of spins, and it returns the chance of at least one occurrence, using the fixed wheel probabilities rather than recent history.

For a single straight-up number the formula is 1 minus (36/37) raised to the number of spins. The calculator does this for you, but a few reference points help.

Spins	Chance a chosen number lands at least once
10	24.0%
37	63.7%
50	74.6%
100	93.5%
200	99.6%

The 37-spin row is the one to remember. Playing a single number for 37 spins, the "average gap", still leaves a 36% chance of not seeing it at all, which is why the drought tracker regularly shows numbers past 100 spins.

For lightning wins the calculator uses the observed 7.33% per-spin rate. Over 13 spins, roughly one average gap, the chance of at least one lightning win on the table is about 63%. Note that a lightning win on the table is not a lightning win for you: you still need a straight-up bet on that number, so your personal chance per spin of collecting any multiplier is 1 in 37 times the multiplier-carrying share, not 7.33%.

This is the only widget on the page that looks forward, and only because it uses theory. It never adjusts for what has already happened, which is correct for a game with no memory.

What does the tracker add over the in-game statistics?

Depth and windowing. The Evolution client shows a strip of recent results and a small hot and cold view over the last few hundred spins; the tracker holds every spin since December 2025 and lets you set the window from 24 hours to all time.

Several questions on this page cannot be answered from the client at all. How often 500x actually pays needs tens of thousands of spins, because it happens on 0.21% of them. Whether the number distribution is flat needs a 90-day sample. The longest drought on record is by definition something a 500-spin history cannot hold.

The tracker also records the full set of struck numbers per round, not only the winner, which is what makes the "multiplier assigned" table above possible, and it stamps every round with a UTC time so the hourly view can show when the table is busiest. Because our figures come from the results feed rather than the client, a spin appears a few minutes after it happens. For a quick check of the last few results the client is faster; for anything statistical, the tracker is the tool.

Where does the data fall short?

In three places: feed gaps, operator-side differences, and the fact that nothing in it predicts a spin.

Feed gaps come first. A collector restart or an upstream outage leaves a stretch of rounds unrecorded, and we do not backfill estimates, so a 24-hour count that looks low usually reflects a gap rather than a quiet studio; the table runs 24/7 from Riga at roughly 60 to 70 spins an hour. Gaps are random with respect to outcomes, so they do not tilt the long-run shares, but they can make a short window look odd.

Operator differences come second. The feed reports the multiplier assigned and whether it paid. It does not know your bet size, your operator's bet limits, or the 500,000 per player per round maximum payout that applies at most operators. A 500x result in our data is a 500x result on the table; what it would have paid you depends on your stake and your operator's rules.

The third limit is the one this page keeps returning to. Every widget describes spins that have already happened. A hot number, a long drought, a run of red, a cluster of 500x results in one afternoon: a fair wheel produces all of these constantly, and none of them changes the 1-in-37 odds of the next pocket or the 2.70% house edge on every bet. The only good use of the data is to understand what a normal session looks like so that you can budget for it. If you find yourself increasing stakes to chase a number that is "due", that is the signal to stop, and our responsible gambling observatory has tools for setting limits before you play.

Official rules, payouts and the 97.30% RTP figure are Evolution's, published in the game client help and on the Lightning Roulette page at evolution.com. Everything else on this page is our own observation from 261,832 spins, and you can check any of it live on the tracker. For the game's bets and payouts in detail, see the how to play guide.

Frequently asked questions

What are the hot numbers in Lightning Roulette right now?

The hot and cold panel on /games/lightning-roulette ranks all 37 numbers by how often they have landed in the window you select, from 24 hours to all time. Over the last 90 days the top three were 23, 10 and 9 and the bottom three were 18, 7 and 30, but the spread between them was only 0.19 percentage points. A hot number is a description of past spins, not a forecast.

How often does Lightning Roulette hit a multiplier?

In our last 90 days of data the winning number was one of the struck Lightning Numbers on 7,164 of 97,698 spins, which is 7.33%. That matches the maths: with an average of about 2.7 Lightning Numbers per round, the chance the ball lands on one of them is roughly 2.7 divided by 37. The average wait between lightning wins was 13.6 spins and the longest wait was 150 spins.

How often does Lightning Roulette pay 500x?

A 500x payout landed on 204 of 97,698 spins in the last 90 days, or 0.21% of spins, about one in every 480. At 60 to 70 rounds an hour that is one 500x result every seven to eight hours on the table as a whole. It pays only if you had a straight-up bet on that exact number, which is a 1 in 37 shot on top of the 500x being assigned in the first place.

Are red and black equally likely in Lightning Roulette?

Yes. There are 18 red and 18 black pockets plus a single green zero, so each color has a 48.65% chance per spin. Our tracker recorded red on 48.71% and black on 48.55% of spins over the last 90 days, with zero on 2.74% against a 2.70% theory. The zero is what gives even-money bets their 2.70% house edge; Lightning multipliers never apply to red or black.

Can Lightning Roulette statistics predict the next number?

No. The ball and wheel have no memory, so a number that has not appeared for 150 spins has the same 2.70% chance on the next spin as one that hit twice in a row. Our 90-day data shows every number between 2.61% and 2.80%, which is what an unbiased wheel produces. The statistics are useful for understanding variance and budgeting a session, not for choosing a number.