The $1 Scratch Ticket Is the Worst Value on the Rack

The cheapest ticket in the shop feels like the careful choice. It is the one you buy when you don't want to spend much, the one at the front of the rack, the one that costs less than a coffee. It is also, reliably, the worst value the lottery sells — and not by a little.

We took the published prize data for every scratch game on sale in 18 US state lotteries and worked out what each one actually returns per dollar. That's 1,275 games, priced from their own official figures: how many of each prize were printed, how many remain, and how many tickets were made. The result was the same everywhere we looked.

What 1,275 games look like when you line them up by price

Here is the average share of sales returned to players as prizes, grouped by what the ticket costs. These are launch figures — what each game was designed to pay across its whole print run — pooled across all 18 states.

The climb is almost perfectly monotonic. The one wobble is the $3 band, which sits a fraction below $2 — 65.9% against 66.2%, a gap of three tenths of a percentage point across 74 games. That is noise, not a counter-example, and we'd rather show it than smooth it away.

Put in money terms: on $1 tickets the lottery keeps about $37 of every $100 you spend. On $20 tickets it keeps about $24. That difference — roughly $13 per $100 — is the price of buying at the cheap end of the rack.

It isn't one state's quirk

A pattern in pooled data can hide a single large state dragging the average. So we ran it again within each state separately, comparing its $1–$2 games against its $20-and-up games. Seventeen of the 18 states sell both. The pattern held in all seventeen.

Seventeen lotteries, run by seventeen separate state agencies, using different vendors and setting their own prize structures. None of them coordinates with the others. All of them made the same decision.

Why cheap tickets pay back less

This isn't a conspiracy; it falls out of the arithmetic of selling a printed ticket.

Some costs are per ticket, not per dollar. Printing, distributing and stocking a ticket costs roughly the same whether it sells for $1 or $50, and the retailer's commission is a percentage of a much smaller number. On a $1 ticket those fixed costs eat a large slice of the price before a single prize is funded. On a $50 ticket the same costs are spread across fifty times more money.

There is also a structural limit. A cheap game cannot offer a large top prize without making the odds of hitting it absurd, so its prize table is necessarily dominated by $1, $2 and $5 payouts — and those small prizes are the ones players most often leave unredeemed. A $50 game can carry a seven-figure top prize and still fund a generous ladder beneath it. The expensive ticket has more room to be generous, so it is.

The part where we tell you not to act on this

It would be easy to read the table above as "buy the $50 ticket." That would be the wrong lesson, and we want to be blunt about it.

Every single game in this analysis returns less than it costs. The best price band on the rack pays back about 80 cents per dollar, which means it takes 20 cents from every dollar staked, forever, across all players. A better payout percentage on a losing bet is still a losing bet. What changes is the rate at which you lose, not the direction.

And the expensive ticket puts far more at risk per purchase. Someone spending $50 on one ticket loses their whole stake on a blank far more painfully than someone losing $1 — and the improved percentage does nothing to soften that. If the practical conclusion you draw from this article is "spend more," the article has done harm rather than good. The honest conclusion is narrower: if you are going to spend a fixed amount anyway, the cheapest tickets are the least efficient way to spend it.

Something that matters more than the price band

Payout percentage is the design of the game. It is not the whole story of what is on the shelf today, because a game's value changes as its prizes get claimed — and the ticket never tells you.

Of the 1,275 games in this analysis, 139 were still on sale with zero top prizes remaining. Not "few left." None. The headline number is still printed on the front, the price is unchanged, and the prize it advertises has already been won by somebody else. (Widening the net to every state that reports a top-prize count — 1,595 games — brings the total to 180, worth $124 million in advertised prizes.)

Some of them are stark. A $50 Missouri game called $5,000,000 Fortune advertises a five-million-dollar top prize and has none left at 58% sold — meaning roughly four in ten of those tickets are still out there to be bought, chasing a prize that is gone. Texas has a $100 ticket, $20 Million Supreme, whose $20,000,000 top prize is likewise fully claimed.

Within a single price band the spread is enormous. Among $5 games currently on sale, the weakest returns about 53 cents per dollar and the strongest about 83 cents — a 30-point gap between two tickets that cost exactly the same. Choosing well inside a price band matters more than moving up a price band. That is the whole reason we publish a page for every state, updated daily.

How we worked this out

Every figure here comes from the state lotteries' own published prize data — for each game, the number of prizes at each tier, how many remain unclaimed, the ticket price and the overall odds. We use those to reconstruct the print run and total prize pool, then divide prize money by ticket money. Nothing is estimated where a state publishes the real number, and where a state doesn't publish something we need, we leave the game out rather than guess it.

Three honest limits. "Unclaimed" is not "unsold" — a prize counted as remaining may already sit in a ticket someone bought and hasn't scratched, so late in a game's life the figures get unreliable; games more than 90% sold are excluded from the averages above. Free-ticket prizes are counted at face value where a state prices them and excluded where it doesn't, which makes a few states slightly conservative. And annuity top prizes are counted at their published cash value, never at the advertised total of all payments — a "$15,000,000" prize paid at $750,000 a year for 20 years is not $15,000,000 today, and treating it as such would badly overstate a game's payout.

Full detail is on our methodology page, and each state page documents what that particular lottery does and doesn't publish.

The bottom line

Across 1,275 scratch games in 18 states, the payout percentage rises with the ticket price, in every state, with no exception: about 63 cents per dollar at $1, about 80 cents at $50. The gap is roughly $13 per $100 spent.

But the useful takeaway isn't "trade up." It is that scratch tickets are a priced product whose value you can actually check before you buy — and that the number printed on the front tells you almost nothing about it. If you play, play for entertainment, decide the amount first, and spend it on games that still have their prizes. See what's left in your state on the scratch games hub, or read what expected value actually means and where the rest of your lottery dollar goes.

Figures are derived from state lotteries' published prize data and describe averages across all players — they are not a prediction about any individual ticket, and not a recommendation to buy any game. Scratch tickets are negative-expected-value by design: every game described here returns less than it costs. For information and entertainment only — not financial or gambling advice. You must be 18+ to play (21+ in some states). If gambling is a problem for you or someone you know, call 1-800-GAMBLER (1-800-426-2537).