Sports Betting Statistics: 11 Numbers That Hold Up
TL;DR: A quarter of active British betting accounts are in lifetime profit. Around one in twenty carries a commercial restriction, and restricted customers are nearly twice as likely to be winning as everyone else. The average football bet loses 7.8% of stake to bookmaker margin, more than the standard calculation warns.
Search for sports betting statistics and you get the same round numbers across twenty affiliate sites, none of which say where they came from. These eleven come from the Gambling Commission, HMRC and a University College Dublin working paper. Where we have done our own arithmetic on published figures, we say so.
1. A quarter of active betting accounts are in profit
The Gambling Commission asked most of Britain's largest online betting operators to break their customer base down by lifetime profit and loss on settled bets. Of 14,923,840 active accounts, 25.42% were in net profit and 72.54% were in net loss. The remainder sat at break-even or were missing from two operators' returns.
Worth being careful with that number. It counts accounts, not people, and plenty of customers hold five or six.
2. The average football bet loses 7.8% of stake
Tadgh Hegarty and Karl Whelan at University College Dublin tested the overround calculation that every betting guide recommends, using 84,230 professional football matches across 22 leagues in 11 countries between the 2011/12 and 2021/22 seasons, priced on average closing odds.
The overround said customers should expect to lose 6.5% of stake. They lost 7.8%, a fifth more. The gap comes from bookmakers pricing a fatter margin into longshots than into favourites, which the overround formula assumes away.
3. Tennis is worse than football
The same two researchers ran the test again on 58,112 ATP and WTA matches. The overround predicted 5.4%. Realised losses came in at 7.4%, two percentage points adrift and a wider gap than football showed.
4. Backing longshots costs eight times more than backing favourites
Sort every football bet in that dataset into ten bands by probability of winning. The least likely tenth lost 17% of stake on average, against 2% for the most likely tenth. Tennis runs wider, 23% at the bottom against 3% at the top.
Richard Griffith documented this pattern in horse racing in 1949 and bookmakers still price it in. If you favour outsiders and accumulators, the margin you pay has little to do with the one advertised.
5. 4.31% of active accounts carry a commercial restriction
From that same sample, operators reported 643,779 accounts restricted in some form. Stake factoring was the most common method, applied to 62.17% of restricted accounts, and commercial closures accounted for 51.69%. Those categories overlap, since one account can carry several restrictions at once.
How heavily operators lean on restrictions varies enormously between them. Some rely on stake factoring alone, and several told the Commission they never close accounts for commercial reasons at all.
6. 22.41% of stake-factored accounts were cut to 1% of normal stakes
Of accounts subject to a stake factor reduction, 22.41% were cut to 1% or less of the maximum available to an unrestricted customer. Another 36.22% were cut to between 1% and 9%.
In practice a customer who could stake £500 can now stake £5, and the account stays open, so it never appears in the closure figures.
Our own arithmetic on those percentages gives roughly 400,000 stake-factored accounts, of which about 90,000 sit at 1% or below and about 235,000 below a tenth of normal stakes. The Commission's bands add up to a little over 100%, because some operators counted the same account in two of them.
7. Restricted customers are almost twice as likely to be winning
Across all active accounts, 25.42% were in profit. Among restricted accounts, 46.78% were. The loss-making split runs the other way, 72.54% of the whole base against 51.29% of restricted customers.
Operators describe restrictions as liability management rather than a penalty on skill, and on their own terms that is a reasonable argument. The Commission's table is the first time anyone has been able to see the pattern across most of the market at once.
8. Winning is not a protected characteristic
Publishing the data, the Commission said that being a successful bettor is not a protected characteristic in discrimination law. It also said that telling operators how to manage their commercial liabilities sits outside its remit, a position it traces back to the 2023 Gambling White Paper.
It did commit to looking at whether operators could be clearer with customers about how, when and why an account gets restricted.
9. Sports betting revenue held flat while the number of bets fell 8%
Online real-event betting produced £600 million in gross gambling yield between January and March 2026, up 1% on the same quarter a year earlier. Over that period the number of bets fell 8% and average monthly active accounts fell 5%.
That works out at roughly 10% more revenue per bet than a year before. Whether it came from larger average stakes or from wider margins, the published data does not say.
For comparison, gross gambling yield across all online products rose 7% to £1.55 billion in the quarter, with slots up 12%.
10. The tax rise is expected to show up in the odds
Remote gaming duty rose from 21% to 40% on 1 April 2026. From 1 April 2027 a new 25% rate of general betting duty applies to online betting, up from the 15% that has covered bets placed online and in a shop alike. Remote bets on UK horseracing stay at 15%, as do bets taken on self-service terminals in licensed premises.
HMRC's own impact note says the measure will affect individuals if operators pass the increase on through a negative change in betting odds or return to player. The Exchequer take, certified by the Office for Budget Responsibility, runs from £810 million in 2026/27 to £1,155 million by 2030/31.
A bookmaker facing that bill can shade its prices to recover some of the cost. Prices on an order book come from the traders on either side, so there is nothing for an operator to shade.
11. The Commission is switching off the dataset behind number 9
The May 2026 release is the last one in that series. The Commission has collected the data monthly since March 2020, and it is the only public source showing bet counts, active accounts and yield by product on a quarterly cadence.
What none of these numbers show
No operator publishes what its winning customers win. The Commission can tell you how many accounts sit in profit and how many got throttled. It cannot tell you the largest position a winner had filled, or how long the payout took, because nobody collects it and nobody asks for it.
Pred runs a peer-to-peer order book. Prices come from traders on both sides rather than a margin set by a house, and we do not restrict, stake-limit or close accounts for profitability. Trade the Game.
Frequently asked questions
What percentage of sports bettors are profitable? The Gambling Commission found 25.42% of active online betting accounts in lifetime net profit on settled bets, from a sample of 14,923,840 accounts covering most of the British market. Figures like "only 3% of bettors win" circulate widely with no published source behind them.
What is the average bookmaker margin on football? Research at University College Dublin covering 84,230 matches found an average realised loss of 7.8% of stake, against the 6.5% predicted by the standard overround calculation.
How many betting accounts are restricted in the UK? Operators reported 643,779 restricted accounts from 14,923,840 active accounts, a rate of 4.31%. Stake factoring was the most common method, affecting 62.17% of restricted accounts.
Do bookmakers limit winning customers? Among restricted customers, 46.78% were in net profit, against 25.42% across the customer base as a whole. The Gambling Commission has said operators are entitled to manage their commercial liabilities and that successful betting is not a protected characteristic in discrimination law.
How far can a stake factor cut your maximum bet? Of stake-factored accounts, 22.41% were cut to 1% or less of the maximum available to an unrestricted customer, and 36.22% to between 1% and 9%. A £500 maximum can become £5.
Is UK betting tax going up? Remote gaming duty rose from 21% to 40% on 1 April 2026. A new 25% rate of general betting duty applies to online betting from 1 April 2027, up from 15%. Online bets on UK horseracing stay at 15%.
What is the difference between bookmaker margin and exchange commission? A bookmaker's margin is built into the quoted price and applies to every bet, itemised nowhere. An exchange charges commission on the trade, and the price comes from the order book. We covered this in more detail in sports prediction exchange vs betting exchange and in why your betting account was restricted.