Mega Roulette Live Stats: Reading the Multiplier Column, Number Heatmap and Droughts Across 291,000 Spins

The live page at /games/mega-roulette shows where the ball landed, which numbers carried a Mega Multiplier, and what a straight-up bet on the winning number would have been paid. Our tracker has stored 291,106 spins of Pragmatic Play's Mega Roulette since 28 December 2025, with 111,542 of them in the last 90 days. This page walks through each field and widget, explains the one column that confuses almost everyone, and sets out what the totals show. For the game itself, read the Mega Roulette guide first.

What does "multiplier 100" mean in a Mega Roulette round record?

It means the ball landed on a Mega Number carrying 100x, and straight-up bets on that number were paid 100:1 instead of the base 29:1.

Every spin we store has four core fields. The result is the winning number, 0 to 36. The color is red, black or green. The bonus_triggered flag is true when the winning number was one of that spin's Mega Numbers. The multiplier is the Mega Multiplier that was paid on the winning number, so it is 50, 100, 150, 200, 250, 300, 350, 400 or 500 when the flag is true, and 1 when it is false.

Three things the multiplier column is not. It is not the highest multiplier that appeared on the board that spin; a 500x can sit on number 14 while the ball drops into 22, and the record then reads 1. It is not added on top of the base payout; the 100x replaces the 29:1, it does not stack. And it does not apply to anything except Straight Up bets, because splits, corners, dozens and even-money bets pay standard European odds regardless of the board.

The full board is kept too. The result_details field stores the list of Mega Numbers for the spin with the multiplier attached to each, plus an is_mega_win flag. That is what lets the multiplier stats panel separate "what was offered" from "what was paid", which turns out to be the most useful comparison on the page.

How many Mega Numbers appear each spin, and how big are they?

Almost always two or three. Over the last 90 days the count per spin split as follows.

Mega Numbers on the board	Share of spins (90 days)
1	2.75%
2	36.40%
3	54.42%
4	3.67%
5	2.76%

The average is 2.67 Mega Numbers per spin, and the tails are thin: a one-number spin is as rare as a five-number spin.

The values on those numbers cluster at the bottom of the 50x to 500x range. Of every Mega Number shown in 90 days, 35.75% carried 50x and 37.50% carried 100x. So roughly three boosted numbers in four are worth 100x or less before the ball is even released. The 150x tier took 12.21%, 200x took 8.03%, and everything from 250x upward shared the remaining 6.5%, with 500x at 1.85% and 300x at 1.86%.

That is the board before the spin. The Mega Multipliers page covers the rules behind it; here the point is that the multiplier stats widget on the live page shows two histograms, and this one, the offered distribution, is the baseline the second one should be judged against.

Does the ball land on Mega Numbers more often than plain ones?

No. With 2.67 Mega Numbers on a 37-pocket wheel, the expected hit rate is 2.67 divided by 37, or 7.2%. The observed rate over 90 days was 7.24%, or 8,073 mega wins in 111,542 spins.

The 30-day figure was 7.19%. The all-time figure across 291,106 spins is 7.23%, from 21,034 mega wins. Three windows, all within a twentieth of a point of the arithmetic. The ball is a physical object dropping into a physical wheel, the Mega Numbers are drawn by an RNG after betting closes, and the two have no way of talking to each other.

Four mega wins in ten spins looks like the feature "heating up"; it is a 7.2% coin landing heads a few times running. There is also nothing to be gained from waiting for a five-number board and betting then: by the time you see the board, betting has already closed.

What has the paid multiplier actually been when a Mega Number hit?

The median paid multiplier over 90 days was 100x. The average was 115.1x, the 90th percentile was 200x, and the maximum was 500x.

The distribution of paid values tracks the distribution of offered values almost line for line, which is the second half of the fairness check.

Multiplier	Share of Mega Numbers offered	Share of mega wins paid
50x	35.75%	35.92%
100x	37.50%	37.25%
150x	12.21%	11.94%
200x	8.03%	7.87%
250x	0.92%	0.90%
300x	1.86%	1.92%
350x	0.95%	1.16%
400x	0.93%	1.02%
500x	1.85%	2.02%

If the ball somehow avoided the big multipliers, the right-hand column would sag at the bottom rows. It does not. A 500x number on the board is hit at the same rate as a 50x number, within sampling error. The slightly higher paid share for 350x to 500x is a few dozen spins either way and moves from month to month.

A mega win at 50x is a good result compared with 29:1, but it is the modal outcome, not the jackpot. Roughly 73% of all mega wins pay 50x or 100x. The headline 500x is 2.02% of mega wins, and mega wins are 7.24% of spins, so a straight-up player has about a 0.15% chance per spin that the winning number pays 500x at all, and a far smaller chance that it is their number. The RTP page shows how the 29:1 base and the multipliers combine into the published 97.30%.

How rare is a 500x, and when did the last ones land?

A 500x was paid on 163 spins in the last 90 days, which is 0.15% of spins or about one every 684. Since tracking began there have been 418.

The highest multipliers panel on the live page lists them newest first with the date and winning number. The recent run: 27 August 2026 (six hits), 26 August (five), 25 August (seven), then 23 August, 22 August and 20 August (two). Three consecutive days with five or more 500x results is a cluster, and a cluster is exactly what a Poisson process produces now and then. With about 1,240 spins a day and a 0.15% rate, the daily expectation is under two, so a seven-hit day is a tail event that will keep recurring across a year of 24/7 play.

For your own bet the arithmetic is stricter. The chance that the ball lands on your straight-up number and that number is a Mega Number is 1/37 multiplied by 7.2%, about 0.19%, or one spin in 510. The chance that it is a 500x on top is 0.19% multiplied by 2.02%, about 0.004%, or one spin in 25,000. At 40 to 50 seconds per spin that is roughly two weeks of nonstop betting on one number for a single expected 500x, and the same number of losing 29:1 spins in between.

What does the number heatmap say about wheel bias?

Nothing that a fixed betting plan could use. Over 111,542 spins the busiest number was 7 with 3,159 hits (2.83%) and the quietest was 12 with 2,857 hits (2.56%), against a theoretical 2.70% for each of the 37 pockets.

The rest of the top three were 33 (3,127) and 21 (3,106); the bottom three were 12, zero (2,880) and 13 (2,909). The whole spread from coldest to hottest is 0.27 percentage points. At this sample size a perfectly balanced wheel throws up gaps of that order between its luckiest and unluckiest pockets simply because 37 numbers are being compared and one of them has to come first.

The heatmap widget shades each number by its share in the selected window. Shorten the window to 24 hours, about 1,240 spins, and the shading becomes dramatic: a number that landed 50 times sits next to one that landed 20. Extend to 90 days and the colors flatten out. Nothing about the wheel changed between those views; only the sample did.

A mechanically biased wheel would show one pocket or sector pulling ahead consistently across windows, and the 90-day ranking does not repeat the 30-day one. The strategy page deals with hot-number betting in more detail; the summary is that it changes which number you lose 2.70% on, not whether you do.

How should I read red, black and zero shares and the streak detector?

As a fairness check for the even-money bets, and as a reminder that long streaks are routine.

Outcome	Theory	Observed (90 days)
Red	48.65%	48.86%
Black	48.65%	48.56%
Zero	2.70%	2.58%

Zero landed 2,880 times in 111,542 spins. The slight shortfall against 2.70% is about 130 spins, well inside normal variation, and zero was also among the three coldest numbers in the same window, which is the same fact seen twice.

The streak detector reports the longest run of one color. In 90 days that was 17 blacks in a row and 15 reds in a row. A run of 10 on a given color has a probability of 0.4865 to the tenth power, about 0.07%, from any starting spin, which sounds small until you multiply by 111,542 starting points. Runs of 15 and 17 are the expected extremes for a sample this size, not evidence that the wheel got stuck.

Two errors follow from streaks. One is betting against the run because it is "due to end"; the next spin is 48.65% red whether the last 17 were black or not. The other is betting with the run because the wheel is "in a black phase". Both are the same fallacy pointed in opposite directions, and both cost the same 2.70% per spin. Even-money bets in Mega Roulette are plain European roulette: no Mega Multiplier ever touches them.

How does the drought tracker count a mega win?

It counts spins since the last time bonus_triggered was true, and separately since the last paid multiplier of 100x or better.

The averages over 90 days were 13.8 spins between any mega win and 21.6 spins between 100x-plus mega wins. The longest gaps were 156 and 162 spins respectively. The widget shows the current gap next to those two reference figures so you can see whether the present wait is ordinary or unusual.

Unusual is not the same as informative. The probability of no mega win in N spins is 0.9276 to the power N: 10.5% for 30 spins and 1.1% for 60 spins. A 156-spin drought is the kind of extreme a 111,542-spin sample will contain once or twice, and none of it raises the chance of the next spin above 7.2%. If the counter reads 80, the wheel has no record of that number.

Note what the tracker cannot measure. It knows the table went 40 spins without hitting any Mega Number. It does not know whether your number was among the Mega Numbers on any of those spins, and it does not know whether your number came up at 29:1 in the meantime. A personal drought and the table drought are different quantities, and only the second is on the page.

What does the probability calculator work out for a straight-up bet?

It turns a fixed per-spin probability into "at least one hit in N spins", using theory rather than history.

For a single number the per-spin chance is 1 in 37, 2.70%. For a mega win on any number it is 7.2%. For your number being both hit and a Mega Number it is 0.19%. The calculator applies the same formula to each: the chance of at least one hit in N spins is 1 minus the chance of missing every time. Enter 100 spins for the 0.19% case and it returns about 17%, which is a useful figure to sit with before a long straight-up session.

This is the only forward-looking widget on the page, and only because it ignores the heatmap, the streaks and the droughts entirely.

One caution on interpretation. "At least one mega hit on my number in 100 spins" at 17% is not "17% chance of profit". Around 83% of the time you go 100 spins without one, and most of the hits that do come are 50x or 100x. Whether a session is ahead depends on how many 29:1 wins and how many losing spins surround that one boosted result; the how to play page covers the payout table if you want to work an example through.

When is the table busiest, and what does the data miss?

The heatmap by hour peaks between 13:00 and 17:00 UTC over the last 30 days. Round volume by hour is the one time-based figure on the page that carries real information, because it reflects how many spins the studio completed, and the studio runs 24/7 at about 40 to 50 seconds a spin.

Everything else by hour is noise. The mega-win rate in any given hour wanders around 7.2%, and a "lucky hour" on the widget is a small sample.

The data has gaps. Our collector reads the results feed, and a feed outage, a network fault or a collector restart leaves spins unrecorded rather than estimated, so a day's count can fall short of the true number of spins. The multipliers are recorded as the feed rounds them, which is why we cite Pragmatic Play's 97.30% RTP rather than deriving a return figure from our own hit rates: a small rounding effect across 8,073 mega wins is enough to nudge a computed number away from the published one without meaning anything. The RTP monitor across games discusses that distinction.

The tracker also follows one Mega Roulette table. Operator bet limits, maximum payouts and racetrack options are not in our records. And a mega win in our data means the winning number was boosted, not that anyone had a straight-up bet on it. Whether the multiplier was paid to a player at all is invisible to us.

How do I use Mega Roulette stats without fooling myself?

Use the page to check that the wheel and the feature behave as published, to understand how often the good outcomes really occur, and to size a session to survive the gaps.

The two fairness checks are the number heatmap (every pocket within 2.56% to 2.83% against 2.70%) and the offered-versus-paid multiplier table (every tier hit in proportion to how often it was on the board). Both say the game runs as Pragmatic Play describes it. That is worth knowing, and it is also the end of what the stats can do for your betting. Our cross-game fairness index applies the same reasoning to other tables.

The frequency figures are the practical ones. A straight-up bettor should expect a mega win on their own number about once in 510 spins, a paid multiplier of 50x or 100x in three cases out of four when it comes, and stretches of 30 spins with no mega win on the whole table one time in ten. An outside bettor is playing 2.70%-edge European roulette with a streak detector attached and should ignore the multiplier panels entirely, since they never apply.

Evolution's Lightning Roulette uses the same structure from a different studio: 29:1 base, 50x to 500x on up to five numbers, 97.30% RTP. The same reading rules apply to both trackers.

Set a loss limit before the first spin, treat the drought counter as a description rather than a countdown, and stop at the limit. The observatory has more on session limits and on why a game that pays 500x needs a smaller stake, not a bigger one.

Frequently asked questions

What does the multiplier number mean on the Mega Roulette tracker?

It is the Mega Multiplier that was paid on the winning number in that spin. A value from 50 to 500 means the ball landed on one of that spin's Mega Numbers and straight-up bets on it were paid at that multiplier instead of the base 29:1. A value of 1 means the winning number was not a Mega Number. The field never shows the largest multiplier that was on the board without being hit.

How often does Mega Roulette hit a Mega Number?

About one spin in fourteen. In the last 90 days of our data the ball landed on a Mega Number in 7.24% of 111,542 spins, and the all-time rate across 291,106 spins is 7.23%. That is exactly what you expect from an average of 2.67 Mega Numbers on a 37-pocket wheel, so the feature is random rather than weighted toward the boosted numbers.

When was the last 500x in Mega Roulette?

Recent 500x results in our tracker fell on 27 August 2026 (six hits), 26 August (five), 25 August (seven), 23 August, 22 August and 20 August (two). Across the last 90 days there were 163 spins that paid 500x, roughly one every 684 spins, and 418 in total since tracking began on 28 December 2025. The live page lists the newest ones first.

Is the Mega Roulette wheel biased toward certain numbers?

Our 90-day heatmap says no. Across 111,542 spins the most frequent number was 7 at 2.83% and the least frequent was 12 at 2.56%, against a theoretical 2.70% for every pocket. That spread is what random sampling produces at this volume. Colors match too: red 48.86%, black 48.56%, zero 2.58% against 48.65%, 48.65% and 2.70%.

How long can Mega Roulette go without a Mega Number hit?

Longer than most players expect. The average gap between mega wins over 90 days was 13.8 spins, but the longest recorded drought was 156 spins. For a hit of 100x or better the average gap was 21.6 spins and the longest 162. The chance of no mega win in 30 straight spins is 10.5%, so a wait of that length happens daily on a table spinning around the clock.