90 ball bingo explained: three prizes, a strip of six and what each ticket buys

90-ball in one minute: the short version

• One ticket, fifteen numbers, three prizes. The ticket is nine columns by three rows, 27 squares holding 15 numbers and 12 blanks, five numbers to a row. One line pays for any single row, two lines for two rows on the same ticket, and the full house for all fifteen. Each has its own winner, so missing the line costs you nothing for the other two.
• Buy a strip of six and you mark a number on every call. Six tickets dealt as a strip carry every number from 1 to 90 exactly once between them. That is why a strip feels so much livelier than a single ticket, and why it reaches a full house on call 79 instead of call 87.
• The pot is a share of ticket sales, so the room sets the prize. In our model room of 250 players on six 10p tickets, the full house is £52.50; the same six tickets in a room of 20 make it £4.20. Your stake is 60p either way.
• Buying more tickets only helps if the room does not. When everyone in a 120-player room moved from one ticket to twelve, this player’s chance of winning something crept from 2.56% to 2.77% while the stake went from 10p to £1.20. Your share is your tickets divided by all the tickets in play, and nothing else.
• The catch is that almost every game pays nothing. On six 10p tickets in a 120-player model room, 97.3% of games returned nothing at all, the average return was 42p against a 60p stake, and a win was worth £15.28.

One rule to keep: pick the ticket price first and the room size second, because the price sets what the hour costs and the room size only decides whether the prize is small and frequent or large and rare.

Which ticket price suits the evening you want

British rooms run from a penny to about a pound a ticket, with most of the lobby sitting between 1p and 10p; published room listings at the British names put most of their games at tickets from as little as 10p, and penny and 2p rooms are published by name elsewhere. Twelve games an hour is the cadence a real room keeps, so a strip of six is easy to price for a whole evening.

Read down the third column rather than the first. The jump that hurts is not from a penny to 2p; it is from six tickets to twenty-four. Ticket count, not ticket price, is what turns a quiet hour into a bill, and the same arithmetic says 96 penny tickets a game cost more across an hour than six 10p ones. Choosing a room and a site is mostly choosing which of those two dials you are turning.

The lobby lists are longer than the list of companies behind them, which matters when you are comparing 90-ball rooms across sites. Entain carries Gala Bingo and Foxy Bingo inside a bingo estate big enough that a single account reaches more than one of its brands, so a welcome taken at one is usually a welcome spent for the group. Gamesys runs Jackpotjoy and Double Bubble Bingo on one platform alongside Virgin Games, Monopoly Casino and Rainbow Riches Casino, and the first two were both active domains on Gambling Commission account 38905 when the public register was read in October 2026. bet365 keeps Heart Bingo next to a sportsbook-first casino, while Sky Vegas is the casino-first brand of the Flutter group.

The same pattern runs through the rest of the market. Grosvenor Casinos and Magical Vegas share one licence, held by Rank Interactive (Gibraltar) Limited on account 57924. Pink Casino is operated by LeoVegas Gaming PLC on account 39198, which puts it alongside LeoVegas, BetMGM and Bet UK. Sun Bingo and Fabulous Bingo both name their licensee, with its Commission account number, in the small type under the lobby, and that is where a check on an unfamiliar 90-ball room ought to begin; judging a new site is five minutes of exactly that kind of reading.

What a 90-ball ticket actually is

Nine columns, three rows, 27 squares. Fifteen of them carry numbers and twelve are left blank, with exactly five numbers on each of the three rows. The blanks are not decoration: they are what makes a line achievable in a sensible number of calls, and they are the reason a 90-ball ticket plays so differently from a 75-ball card where almost every square is live.

A column holds a fixed slice of the numbers, so the first column usually takes 1 to 9, the second 10 to 19, the third 20 to 29 and so on up to the ninth, which takes 80 to 90 and therefore holds eleven numbers where the others hold nine or ten. Say usually, because the ranges vary a little between suppliers: the first column may run 1 to 9 or 1 to 10, the second 10 to 19 or 11 to 20. Nothing about that changes the odds, and it is worth knowing only so that a ticket that looks slightly different from the one you saw last week is not a different game.

Online, the software daubs for you. Auto-daub marks the numbers as they are called whether or not you are watching the screen, and it claims on your behalf, which removes the entire category of mistake that haunts a hall. In a hall the claim has to be checked by hand, the last number called must be part of the winning line or house, and a claim made after the next number has been called is usually void; that checking is why hall games run longer than online ones. Playing from a phone changes nothing about the ticket, as the mobile bingo guide sets out, because auto-daub does the watching.

As for where it all came from, one operator traces the game to a sixteenth-century Italian lottery, Il Gioco del Lotto d’Italia, and says the British version took off after the Second World War. That is the operator’s account rather than settled history, and it is offered here as colour rather than as fact.

The strip of six, and why it covers all 90 numbers

Tickets are sold in strips of six, and the six are not six random tickets. Between them they carry every number from 1 to 90 exactly once. Six tickets times fifteen numbers is ninety, and the strip is built so that none of them repeats. Hold a full strip and there is no such thing as a dead call: every ball out of the machine marks a square somewhere on your screen.

6 tickets × 15 numbers = 90, which is every ball exactly once

That is the single most useful thing to understand about buying 90-ball. A strip is not six times one ticket in the way that six lottery lines are six times one line, because the tickets are deliberately disjoint. The effect shows up immediately in how early the strip gets there. Across 250,000 simulated games of our model room, a single ticket’s full house typically landed on call 87, with the middle eighty per cent of games falling between calls 79 and 90 and the earliest in the whole run at call 47. Across 208,333 simulated games, a strip of six got there on call 79, with the middle eighty per cent between 71 and 83 and the earliest at 39.

Eight calls earlier does not sound like much until you remember that somebody else in the room is also holding a strip. In a model room of 120 players all on strips, the first full house anywhere in the room typically landed on call 60, with the middle eighty per cent between 54 and 64 and the earliest in the run at 24. A single ticket against that field is simply not in the race, which is why rooms sell strips and why the strip is the unit everyone actually buys.

The other place the strip matters is in what the game feels like. One ticket gives you long stretches where nothing happens. Six give you a mark on every call, three rows apiece to watch and a constant sense of being almost there. That is the product, and it is worth being honest that the design is doing exactly what it is meant to do.

One line, two lines, full house: three prizes, three winners

Three prizes come out of every 90-ball game, in a fixed order. One line goes to the first player to complete all five numbers on any one horizontal row. Two lines goes to the first to complete any two horizontal rows on the same ticket. The full house goes to the first to cover all fifteen numbers on one ticket. Each stage has its own winner, so the player who takes the line is not out of the running for the other two, and the player who misses it has lost nothing.

The money behind those three prizes is a share of what the room spent on tickets. Operators say so in their own words: one brand’s bingo FAQ puts the payout down to the number of tickets sold and the ticket price, another says what you can win in a session depends on how many players are in it, and a third notes that higher-priced games tend to have bigger payouts while penny games pay smaller. What no British operator publishes is the percentage of ticket money that becomes prize money, so anybody quoting you one has made it up. Our model rooms use 70%, stated here as an assumption so you can argue with it, and split the pool 20% to one line, 30% to two lines and 50% to the full house.

Work the three rooms through and the shape is obvious. Twenty players on six 10p tickets sell 120 tickets for £12, which at 70% is an £8.40 pool: £1.68 for one line, £2.52 for two lines and £4.20 for the full house. Eighty players sell 480 tickets for £48, a £33.60 pool paying £6.72, £10.08 and £16.80. Two hundred and fifty players sell 1,500 tickets for £150, a £105 pool paying £21, £31.50 and £52.50. Your stake is 60p in all three.

Then look at what the simulation says about how often that money reaches you. In the twenty-player room this player took one of the three prizes in 14.77% of games, about 1 game in 6.8, and a win averaged £2.83; 85.23% of games paid nothing. In the eighty-player room it was 3.99%, about 1 in 25.1, averaging £10.47. In the room of 250 it was 1.33%, about 1 in 75.1, averaging £31.78, with 98.67% of games paying nothing. The expected return is 42p on a 60p stake in every one of the three, because 70% of 60p is 42p whoever else turns up.

£0.60 × 0.70 = £0.42 a game, in a room of 20 and in a room of 250 alike

So the room size is not a value decision at all; it is a temperament decision. A quiet room hands you small prizes often. A busy room hands you a large one rarely. If what you want is the hour, the quiet room gives you more to cheer about; if you are there for a number that would change your week, only the busy room has one, and you will see nine hundred games go by waiting for it. The same trade with the prize money removed entirely is on the free bingo page.

When each prize actually lands

Here is the part that nobody puts on a promotion page, and the part that explains why 90-ball feels the way it does. There are two completely different sets of call numbers in every game: the call on which the room claims a prize, and the call on which your own tickets would have been ready. They are nowhere near each other.

In a model room of 120 players, each holding a strip of six 10p tickets, the room’s one line is typically claimed on call 20, two lines on call 43 and the full house on call 60. This player’s own best ticket typically reaches one line on call 47, two lines on call 66 and the full house on call 79. Twenty-seven calls behind on the line, twenty-three on two lines, nineteen on the full house. You are not unlucky; you are one strip among a hundred and twenty.

Fill the room and those first three numbers come in. With twenty players on strips, the room’s calls were 28, 51 and 66. With eighty, 22, 45 and 61. With 250, 18, 41 and 57. A busier room finishes sooner, which is why a packed room feels fast and a thin one drags, and why the published game length of four to ten minutes covers both. Your own tickets, meanwhile, do not care at all: they reach one line on call 47, two lines on 66 and the full house on 79 whether the room holds twenty people or 250, because your tickets do not know who else is playing.

There is a practical use for all this. If you are in a room where the full house routinely lands around call 57, the game is effectively over for you by call 60 and an evening is really a sequence of short races rather than long ones. If the room is quiet and the house lands around 66, there is time for the tension to build and your own strip genuinely is in contention. Neither is better value. They are different evenings, and you can tell which one you are in from the lobby before you spend anything.

What buying more tickets really does

There is a published rule for this and it is one line long: divide the tickets in your hand by all the tickets in the game and multiply by a hundred. The operator publishing it walks through four cases of its own, in which 5 tickets among 100 comes to 5%, 10 among 300 comes to 3.33%, 8 among 200 comes to 4%, and 10 among 100 comes to 10%. All three British prizes follow that same fraction, so more tickets push it up and more players push it down.

That rule has a sting in it, and the simulation makes it plain. We held the room at 120 players and 10p tickets and changed only how many tickets everybody bought. With one ticket each the room sells 120 tickets, the pool is £8.40 and this player wins something in 2.56% of games. With three each, 360 tickets, a £25.20 pool, 2.64%. With six each, 720 tickets, a £50.40 pool, 2.70%. With twelve each, 1,440 tickets, a £100.80 pool, 2.77%. The stake went from 10p to £1.20, a twelvefold increase, and the chance of winning something moved by about a fifth of a percentage point.

6 tickets of 720 = 0.83%, but 12 tickets of 726 = 1.65%

That second sum is the honest version of the advice. Buying twelve tickets while everyone else holds six really does roughly double your share, because the denominator barely moves. Buying twelve because the whole room has moved to twelve buys you nothing but a bigger prize for the same chance, which is a perfectly reasonable thing to want and a terrible thing to be sold as better odds. The same point sits behind the fixed-ticket session rooms, where everyone holds exactly fifteen tickets for a flat £5: nobody can out-buy anybody, so your chance is simply one in the number of players.

What more tickets do change is how close you get. With one ticket each, this player’s own best ticket reached one line on call 66, two lines on 79 and the full house on 87. With three, calls 54, 71 and 82. With six, 47, 66 and 79. With twelve, 42, 63 and 76. More tickets buy you a better seat and more near misses, which is worth something as entertainment and nothing as value.

Price the evening before you buy it. This is set to a strip of six at 10p across twelve games, with the payback set to 83, which is the sector prize share rounded down rather than our model’s own assumption.

One cost sits outside every number above. Scratchcards and reels are sold in the same lobby as the tickets, the regulator writes its session guidance around a customer doing both at once, and the stake ceiling on a single online slot spin is £5 for anyone 25 or over and £2 for an 18 to 24-year-old. That is more than an entire hour of penny tickets, gone in one press. The rule forbidding several games on screen together covers casino play and not bingo, which is precisely why a reel can sit beside your strip while you wait for the next call.

Sharing a prize when two people call at once

With 720 tickets in play and ninety balls, two tickets completing on the same call is not a freak event; it is a regular one, and it is built into every figure on this page. The rule is simple and published: joint winners share the prize equally. Published rules say of one short format that if two players finish at the same time the prize is shared between them, and of a card-hybrid game that where more than one player has a claim the prize is split between the winners.

Community and network jackpots work differently and you should know which you are in before the game starts. A supplier’s 90-ball game with a community jackpot gives the full house winner half the jackpot and splits the rest equally among the other participants who also qualify. A network jackpot at a single brand pays the player who made the call a percentage of the prize, with the remainder split equally between everybody else holding a ticket in that game, and the same operator elsewhere puts the caller’s share at 50%. That is a genuinely different product from an ordinary prize: everybody in the room gets something, and the winner gets much less than the headline.

The technical standards back this up with a requirement rather than a suggestion. Jackpot rules have to describe what happens when two or more players trigger the same jackpot simultaneously, or appear to, including because of network latency. So if you cannot find out what a room does with a tie, that is a gap in the room’s published rules rather than a gap in your reading. Where those pots come from and what the win conditions cost is the subject of the jackpots guide.

One consequence worth carrying. Because ties are common, the chance of winning or sharing a prize is always a little higher than your bare share of the tickets, and the average win is correspondingly smaller. That is why the figures above show this player collecting something in 2.70% of games in a room where a flat share would suggest 0.83% for each prize: three prizes, each sometimes split, add up to more occasions and less money.

A full house in 38 calls, and why no single room can do it

Bingo jackpots usually carry a condition to win: a full house completed inside a set number of calls, or on a particular number, or in a named pattern. One published example sets the bar at 38 calls for a fixed cash prize described as varying from £10,000 to £20,000. The number of balls changes from room to room, which is what makes one jackpot pool bigger than another.

Our runs put a number on how tight 38 calls is. For a single ticket, the chance of all fifteen numbers landing inside the first 40 calls is 1 in 1,138,478; inside 38 it is about 1 in 2,960,043. These are exact combinatorial figures taken from the same run as the simulation, because the event is far too rare to count reliably even in 250,000 simulated games. A strip of six improves it a great deal and still leaves it hopeless on its own: inside 45 calls a single ticket is 1 in 132,792 and a strip of six is 1 in 22,132.

A whole room is a different matter, and that is the entire design. With 720 tickets in play, a full house inside 45 calls happened in 0.54074% of our simulated games, about 1 game in 185. At 37 calls that same room managed it in 0.01472% of games, about 1 in 6,793, and at 39 calls in 0.03952%, about 1 in 2,531; a 38-call condition sits between the two. Run a few hundred games an hour across every room on a network and a one-in-several-thousand event becomes a weekly occurrence, which is precisely why a condition that strict is always attached to a pot shared across many rooms rather than funded by one.

1 in 2,960,043 for your ticket; about 1 in 4,000 for a 720-ticket room

Two other shapes turn up. Some rooms set the condition on a particular number rather than a call count, so a full house claimed on the number 7 or the number 80 takes the jackpot. Others publish a must-drop prize whose call limit starts low and rises through the day until, at a stated evening game, the condition is set to 90 balls so that the prize is certainly awarded that night. None of that changes the arithmetic of your own strip; it changes who is competing with you and how often the pot resets.

90-ball against the shorter games

Most British lobbies run several formats side by side, and they are genuinely different products rather than cosmetic variations. The 75-ball card is a 5×5 grid of 25 squares with a free centre that counts towards any pattern, so you start a square ahead; its columns run B 1 to 15, I 16 to 30, N 31 to 45, G 46 to 60 and O 61 to 75, the winning pattern is announced before the game and, as one lobby puts it, there is only ever one winner. The 80-ball ticket is a 4×4 grid of 16 colour-coded squares, run in some rooms for a first line and a full house and in others on an announced pattern. The 30-ball ticket is a 3×3 grid of nine numbers with no free square and a full house as the only prize.

Published game lengths from one operator’s own comparison make the trade concrete: 30-ball runs one to three minutes, 75-ball three to seven, 80-ball five to eight and 90-ball four to ten. A supplier that holds a British licence lists its bingo range as 90, 80, 75, 50 and 30 ball plus a speed variant, and describes one shortened 90-ball game as three to four times quicker than the full version.

Where that leaves 90-ball is as the long game with the most to watch. Three prizes instead of one means three chances to collect from the same ticket, and the two-line stage in the middle is a genuinely distinctive thing that no other format has. Against that, a 90-ball game ties up ten minutes where a 30-ball game takes two, so an hour of 90-ball is a dozen games and an hour of speed bingo can be thirty. Cost per hour follows ticket count, not format, so the faster game is the dearer one at the same ticket price. The 75-ball guide works through the pattern game on the same arithmetic.

One warning about names. Speed Bingo is a room name rather than a format: one brand’s room of that name is a 75-ball game with tickets from 1p, while another brand’s lobby lists a Speed Bingo room as 30-ball. Thirty-ball is often called speed bingo, but a room called Speed Bingo may be either, so read the room’s rules rather than its title. The same goes for hybrid rooms, where a Slingo game can appear inside a bingo lobby while being, in regulatory terms, a reel-based casino game carrying slot stake limits.

What a 90-ball room must tell you, and what it never will

A bingo room is not required to publish a return-to-player percentage, and that is not an oversight. The technical standards require information that would reasonably let you judge your chances of winning to be available before you commit, and they accept, for bingo and for games where the chance depends on how many people turn up, a description of the way prizes are allocated instead of an RTP or a probability. The carve-out exists because your odds genuinely do depend on the size of the room, which cannot be known in advance.

What the room must do instead is tell you how the prizes are determined, including where the amount is not known before you commit. It must make the rules easily available, covering what counts as a winning outcome, any limits on play or maximum wins, and whether there are jackpot contributions and how the jackpot works. A remote licensee also has to provide a player’s guide to each gambling opportunity it offers. So a room that will not tell you its prize structure, its allocation rules or its jackpot condition has fallen short of something published, and you are entitled to go looking for it.

Three more standards apply to the game itself. You have to commit to each game cycle individually, which is why auto-daub is allowed and auto-play is not: auto-daub marks tickets you already hold rather than buying the next game for you. Reality checks apply to bingo, you set the frequency, and you must acknowledge one for it to leave the screen. And there are rules on interrupted play, which exist so you are treated fairly and know in advance what happens if your connection drops mid-game, which on a strip of six at call 50 is a question worth having answered before it happens.

Two more things follow a 90-ball ticket around whatever room you buy it in. Bonus wagering at a British-licensed site has been capped at ten times the bonus since 19 January 2026, which applies to a bingo bonus as much as to a casino one and narrows the gap that no-wagering brands such as PlayOJO and MrQ were built on. And a licensed site cannot take a credit card, banned for gambling since 14 April 2020, so an add-a-card prompt means a debit card. Long-standing names carry the same obligations as new ones: 888 Casino has been online since 1997 and William Hill joined the same group in 2022, and both sit on licences you can look up exactly as you would look up a brand you have never heard of.

What a room will never tell you is how many tickets sold, how big the pot got before the call, or what share of ticket money it keeps. None of that is required, and none of it is published by any British operator we read. The one figure that exists is the sector-wide 83.25% prize share, and it describes an entire financial year across every licensed room rather than the one you are sitting in. Everything else on this page is either the arithmetic of the game, which cannot be changed by anybody, or a model of a room with its assumptions on the table.

If an evening of bingo stops being an evening out, the tools are free and they work: self-exclusion across every British-licensed site through Gamstop, and free confidential advice from GamCare. Deposit limits are reachable from the homepage of every licensed site, a reduction takes effect at once, and an increase has to wait at least 24 hours.

How the draw was simulated

Every room figure here is our own simulation of a model room using this site’s bingo library, and a model room is not any operator’s published game. The 90-ball runs use bingo90, which deals tickets of fifteen numbers on a 9×3 grid off strips of six, draws the balls in a uniformly random order, awards one line, two lines and the full house to whoever completes each first, and divides a prize equally between every ticket completing it on the same call. Ticket money becomes prize money at an assumed payback of 70%, split 20%, 30% and 50%, and the room is re-dealt between blocks of games so the result averages over ticket layouts as well as over draws.

Session counts, because they decide how much weight a figure carries. The room-size runs at six 10p tickets cover 200,000 simulated games at 20 players, 200,000 at 80 and 100,000 at 250. The ticket-count runs at 120 players cover 200,000 games each at one, three and six tickets, and 104,166 at twelve. The call figures in the second chart come from the 200,000-game run at 120 players on six tickets. The full-house call distributions come from jackpot: 250,000 simulated games for the single ticket and 208,333 for the strip of six, both in a 120-player room.

Where a figure is exact rather than simulated it is marked as such. The expected return is exactly the payback share of the stake, because a symmetric room gives every player the same expectation, and the jackpot probabilities for tight call limits are exact combinatorics rather than counted hits, because a one-in-a-million event cannot be measured in a quarter of a million games. The simulation sits alongside those exact figures and agrees with them wherever it has enough hits to stand on its own.

Everything else is sourced. The ticket layout, the column ranges, the strip of six, the three prizes, the ticket prices, the game lengths, the format comparisons, the tie-splitting rules and the jackpot conditions are quoted from operators’ and suppliers’ own published pages read on 10 October 2026. The regulator supplies the rest: the prize-allocation standard, the player’s guide duty, the commit-to-each-cycle rule, reality checks, the interrupted-play requirement, the simultaneous-jackpot wording and the slot stake ceilings all come out of its licence conditions, its codes of practice and its remote technical standards. The 83.25% figure is worked out of its industry statistics covering April 2025 to March 2026, which were published on 17 September 2026. Every brand mentioned is drawn from a maintained sheet of permitted brand facts, and nothing here says that any of them is running a given offer today.

Questions about 90-ball bingo

Next on your sheet

• BingoHow a British room works, what a ticket buys and where the prize money goes
• 75 ball bingoThe pattern game, the free centre square and what a faster round costs
• Bingo jackpotsHow progressive pots are funded and what a call limit is really worth
• Free bingoThe same three prizes with the ticket money taken out of the sum
• Best bingo sitesTicket prices, room sizes and licences compared across brands
• SlingoThe reel-and-grid hybrid, its published RTPs and its slot stake limits