The best odds of winning any prize — and what those prizes are worth

Every division of a lottery is a separate event, and because the divisions are mutually exclusive — you cannot match exactly 3 and exactly 4 with the same line — their probabilities simply add. That gives the "win any prize" figure operators love to advertise, and it is genuinely computable rather than a marketing invention.

P(any prize) = P(division 1) + P(division 2) + … + P(lowest division)

Everything in this article is that sum, taken over the divisions in each operator's published prize table. The formula for each individual division — the hypergeometric one — is set out in jackpot odds vs any-prize odds.

The ranking

Two things jump out. First, the ranking is almost unrelated to the jackpot ranking: Powerball has among the worst jackpot odds and respectable any-prize odds, while Saturday Lotto has excellent jackpot odds and dreadful any-prize odds. Second, the spread is 16-fold, from 1 in 5.70 to 1 in 92.

Why the numbers differ so much

It is almost entirely a design decision about how low the lowest tier reaches.

Games with a separate second barrel can pay for that barrel alone. US Powerball pays $4 for matching only the Powerball, and that single tier has probability 1/38.32 = 0.026093. Powerball's total any-prize probability is 0.040211, so that one tier is 64.9% of it. Strip it out and the remaining probability is 0.040211 − 0.026093 = 0.014118, or 1 in 70.8.

Games that require several main numbers pay rarely. Australia's Saturday Lotto has no prize below "three winning numbers plus a supplementary" (1 in 297) and "one or two winning numbers plus both supplementaries" (1 in 144). Its jackpot is 36 times more likely than Powerball's, and its any-prize odds are three times worse.

Match-2 tiers are the single biggest lever. UK Lotto's Match 2 alone is 1 in 10.26 — which is essentially the whole of its 1 in 9.23 headline. The National Lottery's own game procedures list it as a fixed £1 prize.

The part the advertising leaves out

Here is the arithmetic that matters. A "win" is only a win if it exceeds the ticket price.

UK Lotto is the cleanest illustration. The most likely non-losing outcome by a wide margin is Match 2, which since June 2026 returns £1 against a £2 line — the ticket "wins", and the player is fifty per cent down. Before that change the same tier returned a free Lucky Dip: a win that paid in more lottery tickets. Canada's Lotto Max still does exactly this, awarding a Free Play at Match 3.

There is a real behavioural phenomenon here — a near-miss or a below-cost return keeps players engaged more effectively than a clean loss does — but the mathematics is simple enough to state without any psychology: most winning tickets lose money.

Doing the sum properly

The honest question is not "how often do I win something" but "what is a ticket worth". That is the expected value: sum, over every division, of prize × probability. For US Powerball the eight fixed-cash divisions contribute a combined $0.3199 per $2 ticket — about 16% of the price — before the jackpot is counted at all. We work that decomposition out in full in lower-division prizes are the real returns, and you can run it for any jackpot size in the EV calculator.

And if you want the opposite of this article — the single most likely outcome of any lottery ticket ever bought — it is matching nothing at all, which we cover in the odds of matching zero numbers.