The payout ratio league table: why draw games return half and slots return 93%

Put every product on one axis — prizes returned per unit staked — and the spread is enormous. This is that league table, built only from regulator filings and audited accounts, followed by the structural explanation for the gap.

The table

The slot figures are the Nevada Gaming Control Board's monthly revenue report, which publishes a "win percent" that its own introduction defines as "the reported win amount divided by the total dollar amount played by patrons" — the exact complement of RTP. Statewide slot win was 7.12%, so RTP was 92.88%. Everything else is sourced in return to player by game.

Note the internal ordering within slots: the cheapest machines pay back the least. Penny slots hold 8.87% against $5 slots' 4.53%.

Why draw games sit at 50%

The answer is entirely visible in the UK's published breakdown of where each £1 of ticket money goes (Gambling Commission, FY2024-25):

Four of those five lines are absent, or radically smaller, in a casino:

1. Good causes. A lottery exists, legally, to raise money for causes. Twenty-three pence in the pound is a purpose, not an overhead. A slot machine has no equivalent obligation. This single line accounts for nearly half the gap between 55% and 93%.

2. Tax base. UK Lottery Duty is "12% of all stake money" with "no deductions" (HMRC). Remote Gaming Duty on casino games is charged on a completely different base: "your profits are the amounts due to you as stakes… less amounts paid out as winnings" (HMRC). Taxing turnover forces the payout down; taxing profit does not. A game taxed at 12% of stakes must find that 12p before a single prize is paid.

3. Retail margin. Physical lottery tickets are sold through tens of thousands of independent shops that take a commission on every sale — 3p in the UK, 5% by statute in Texas. Online products have no such layer, which is exactly why Michigan's online instants (89.43%) outpay its retail scratch cards (76.07%).

The same four lines explain why Australia sits ten points higher than Britain. Australian draw games are required to allocate 60% of subscriptions to prizes, and The Lottery Corporation reports that "27 cents in every dollar spent on a typical lottery ticket is returned to state and territory governments" — a smaller government take than the UK's combined 23p of good causes plus 12p of Lottery Duty, and the difference lands in the prize fund.

4. Prize structure and volatility. Roughly a third of a draw game's prize fund is committed to a jackpot that usually is not won, and therefore rolls. That money is eventually paid out, but the game is designed around a large, lumpy, deferred prize; a slot machine returns almost all of its money in small, immediate increments, which is cheap to administer and produces no rollover liability.

The trap in the league table

A 92.88% machine sounds nine times kinder than a 50% lottery. It is not, and the reason is that RTP is measured per stake, not per pound you walked in with.

Buy £100 of UK Lotto tickets. You stake £100 once. Your expected return is £50, your expected loss is £50, and the transaction is over.

Put £100 into a slot machine at 92.88% and play £1 spins. Each spin loses 7.12p in expectation, but the winnings return to the machine and get staked again. After 200 spins you have staked £200, not £100. Expected loss:

200 spins × £1 × 0.0712 = £14.24

After 1,000 spins — perhaps forty minutes at a modest pace — you have staked £1,000:

1,000 × £1 × 0.0712 = £71.20

And at 1,404 spins the expected loss reaches the entire £100:

£100 ÷ £0.0712 = 1,404 spins

The number that determines your loss is handle (total amount staked), not deposit. A high RTP simply means each individual stake is cheap, which permits an enormous number of stakes per hour. A lottery draw allows one stake per draw, two or three times a week. That structural difference — stakes per hour — matters far more to a player's outcome than the RTP percentage does.

expected loss = total amount staked × (1 − RTP)

Both halves matter, and gambling products differ in the first half by a factor of thousands.

What the table is genuinely good for

Comparing like with like. Within draw games, the spread is real and actionable: Mega Millions realised 48.75% in Michigan against Powerball's 51.39% in the same state and year, at two and a half times the ticket price. Within scratch cards, price points differ — Texas's closing analysis for one $2 game records a "Printed Payout Percentage" of 65.07%, well below the 76% category average, and cheap tickets systematically pay back less.

Sanity-checking marketing. "Over £100 billion paid out in prizes" is true and tells you nothing without the denominator. Prizes ÷ sales is the only version of that claim that means anything.

Understanding a roll-down. The one time a draw game's payout leaps is when accumulated jackpot money is forced into a single draw — briefly pushing a draw's return above 90%, into slot-machine territory. That mechanism, and the arithmetic of it, is in Must Be Won draws and roll-downs.

What the table cannot do is tell you which product is safer. That depends on how fast you can stake, not on how much comes back per stake — and on that measure the ranking inverts completely.