How we calculate expected return
Our rankings come from one straightforward idea, applied carefully to official data. Here is exactly how a pile of state prize tables becomes a single, comparable number.
The core idea
For any game, we add up the value of every prize that is still unclaimed, then divide by the number of tickets still unsold. That gives the average value of a remaining ticket. Compare it to the ticket price and you have the expected return per dollar — what the ticket pays back, on average, for each dollar you put in.
Expected return = (value of all unclaimed prizes) ÷ (estimated tickets remaining). We assume winning and losing tickets sell at the same rate — the standard, neutral assumption when the state doesn’t publish which specific tickets have sold.
Where the numbers come from
The hard part is the denominator — how many tickets are left — because states publish their data differently. We use whichever of four methods the available data supports, and we label the less certain ones as estimates:
- Exact. Some states publish the actual print run (total tickets made). This is the most accurate input.
- Estimated.When the print run isn’t published, we derive it from each prize tier’s total count and its odds, then take the consensus across tiers.
- Direct. When a state only publishes prizes remaining plus the overall odds, the print run cancels out of the math, letting us estimate tickets remaining directly.
- Not calculable.When a prize table is incomplete (for example, only the largest prizes are listed), we don’t fabricate the rest — we mark the game Return N/A.
What we’re careful about
- Annuity prizes (paid out over decades) are counted at their advertised face value, which can make a few end-of-life games look unusually generous. We flag those rather than hide them.
- “Free ticket” prizes are valued at the ticket price, not zero — ignoring them would distort the return.
- Stale data. States update on their own schedules and prizes are claimed between updates, so every figure can lag reality by days. We show when each game was last collected.
You can see the inputs for any game on its own page — the full prize table, tickets remaining, and the method used. Start by browsing the ranked database or reading our plain-English primer on expected value.
Frequently asked questions
What does “expected return” mean?
Expected return is the average amount a ticket pays back per dollar spent, based on the prizes still unclaimed in that game. A return of $0.75 per $1 means that, on average, a ticket gives back 75 cents for every dollar played. It is an average across all remaining tickets, not a prediction for any single ticket.
Why is the expected return almost always less than $1?
Scratch-off games are designed so the state keeps a share of every dollar wagered. Across a full print run, scratch-offs typically pay back 60 to 70 cents on the dollar. A higher-ranked game loses you less on average; it does not make you money.
How often is the data updated?
We refresh prize data every day from official state lottery sources. Because prizes are claimed continuously, a game's expected return changes over time, so daily updates keep the rankings current rather than relying on launch-day odds.
Can a game's expected return be above $1?
Rarely, yes. Late in a game's life, if a large top prize is still unclaimed among relatively few remaining tickets, the average value of those tickets can briefly exceed their price. These are uncommon, short-lived situations and still carry enormous variance — most individual tickets still lose.
Why do some games show “Return N/A”?
Some states don't publish enough data — for example, exact ticket counts or a complete remaining-prize table — to calculate a reliable current return. Rather than guess, we mark those games Return N/A.
Expected returns are statistical estimates built on assumptions and official data that may be delayed or incomplete. They are not predictions of what any individual ticket will pay, and nothing here is gambling or financial advice. See our full disclaimer.