The numbers on the back of a scratch-off are clear enough. "Overall odds: 1 in 3.5." "Odds of top prize: 1 in 12,000,000." Almost nobody actually reads them. The people who do mostly get them wrong.
That's not a knock on players. The numbers are just doing more work than they look like they're doing. Once you know what they actually count (and what they don't), the rest of how scratchers work clicks into place.
Here's the plain-English version of how scratch-off odds work, why they shift over time, and how to read them like someone who's read more than one ticket.
What "1 in 3.5" actually means
Most common misread: people see "1 in 3.5" and figure that out of every three or four tickets they buy in a row, one is a winner.
That's not what it means.
What it means: across the entire print run of the game, about 28.6% of all tickets are winners. Distributed randomly. Three losers in a row is normal. Eight in a row is also normal. So is three winners in a row, though that one is rarer. Each ticket is its own independent event.
This is the single biggest misconception in scratch-offs, and it's responsible for roughly half of all "I feel like I'm due" purchases that don't pay off. The ticket has no memory of the last one.
It's also worth knowing that "winner" in this stat includes every prize tier, including the smallest ones. Winning a $5 prize on a $5 ticket is technically a "win" by the published odds. It's also a refund. (More on this in a minute.)
The print run is fixed
Every scratch-off game is printed in a fixed quantity with a fixed prize structure. A typical $5 California game might look like:
- ~12 million tickets total
- 4 top prizes of $1,000,000
- A few hundred mid-tier prizes ($1,000 to $50,000)
- A few thousand small prizes ($25 to $500)
- A few million small wins and refunds ($5 to $20)
- A current rate of roughly 70¢ on the dollar (what we mean: about 70 cents of every dollar comes back to players as prizes, on average)
Once that game prints, the prize pool is set. The state can't add more winners. They can't remove them either. Every prize that gets claimed comes out of a finite pool.
This is the part that matters: a scratch-off game is a depleting inventory, not a slot machine. Slot machines re-roll each pull. Scratch-offs are pulling from a stack of pre-printed tickets where the winners and losers were determined at print time.
Why posted odds don't update (and what does)
The "1 in 3.5" on the back of the ticket was computed against the original print run. That number never changes, no matter what happens after the game ships.
But the reality underneath it absolutely does change.
Imagine a $20 game where all 4 top prizes were claimed in the first month. The remaining tickets still say "1 in 3.5" overall odds. But the implied value of the remaining pool is now meaningfully worse than it was at launch, because the four biggest prizes are gone. Same printed odds, materially different product.
The reverse also happens. A game that's seen a lot of sales but disproportionately many losers means more winners are still in circulation. The pool is now better than the printed odds suggest.
The number that does update, and the one actually worth knowing, is the current rate: a recalculated estimate of how many cents per dollar come back, based on what's still left in the pool. State lottery sites publish the remaining prize counts (usually). Computing the current rate from them is just math, and it's exactly what the Serious Score does for every active game, every day. The printed number doesn't move. The real one does.
Are you "due" for a winner?
This is where the gambler's fallacy gets confusing, because the usual "every flip is independent" answer is half right and half wrong for scratch-offs.
A coin flip is fully independent. The coin has no memory. "I'm due for tails" is a fallacy. Flips and tickets are not the same product.
A scratch-off pool is depletion. The pool has memory: prior claims change what's left. Over the whole game, "fewer winners remain, so the probability of any given remaining ticket being a winner has shifted" is mathematically true.
But the catch is that you don't pick your ticket from the entire game inventory. You pick a roll a clerk grabs from under the counter, which is essentially random from your perspective. So the depletion math applies to the game in aggregate, not to your specific purchase.
The honest version: a game where the prize pool has held up well (top prizes still out there, decent current rate) is, on average, a better bet than one where it's been gutted. "I personally am due" is still wrong. "This game still has favorable math" can be right, and you can actually check.
What counts as a "win" (and why Above Cost matters)
The published odds count every prize tier, including the breakeven ones. Winning a $5 prize on a $5 ticket is technically a "1 in 3.5" event. It's also not really a win in the way most people mean it.
A more useful question: did you win more than you paid?
That's what the Above Cost filter does on Serious Scratchers. It excludes breakeven and below-cost prizes from the calculation, so the current rate, the score, and the rankings all reflect tickets where you'd actually come out ahead. The number is meaningfully smaller than "1 in 3.5," and it's closer to what players intuitively mean by "winning."
For top prizes paid out as a jackpot, the cash option is also typically 50% to 60% of the advertised face value. A "$10 million top prize" is usually a $5–6 million cash payout. Nothing shady about it: it's standard present-value math for lump-sum vs annuity payouts. Just worth knowing before the headline number sets your expectations.
Top-prize odds vs. overall odds: different products
People conflate these all the time. They're different things.
A game with 1 in 3.5 overall odds and 1 in 12,000,000 top prize odds is a "lots of small wins, almost never the jackpot" game. You'll hit something often. The big one is mostly theoretical.
A game with 1 in 4.5 overall odds and 1 in 800,000 top prize odds is a "miss more often, but when you hit it could mean something" game. Different psychology. Different product.
Neither is "better" in the abstract. They're built for different players. A new player chasing the daydream of the big win should look at top-prize odds. A regular player playing for the small-win frequency should look at overall odds. Most people want some of both, which is why ranking systems weight multiple signals together.
Where to look up real, current odds
Two options:
- The state lottery site. Always the source of truth for prize schedules and remaining prizes. Almost always painful to navigate. You'll find the data; you won't find it compared, sorted, or current-rate-computed.
- Serious Scratchers. We pull the same source data daily, recompute the current rate as prizes get claimed, and rank every active game in your state. The number on the back of the ticket is what the lottery prints. The number on the site is what the math says today.
Be skeptical of anywhere else, especially blog posts with no published-date or no link to real data. "Best scratchers to buy right now" articles that don't show current numbers are almost always months stale. Stale data in this category is worse than no data, because it tells you the opposite of what's true.
The odds aren't a mystery. They're just doing more work than they look like they're doing. Once you can read them, the rest of the strategy follows. See today's top-ranked games in your state →