How Scratch-Off Tickets Are Made and Printed

Not random in the way you think

Here is something most people do not realize about scratch-off tickets: the prizes are not randomly distributed. They are algorithmically placed across the entire print run so that the total payout matches the game's published prize structure exactly.

If a game prints 10 million tickets and promises a $5 million top prize with an overall return rate of 65%, the number and placement of winning tickets is calculated in advance. The manufacturer knows exactly how many winners exist before a single ticket reaches a store.

The printing process

Scratch-off tickets are printed by specialized security printing companies. In the US, the two dominant manufacturers are Scientific Games (now Light and Wonder) and Pollard Banknote. International Games Technology (IGT) also produces tickets for some states.

The printing process has more in common with currency production than regular commercial printing. Each ticket goes through multiple layers:

The base layer is standard card stock. On top of that goes the variable data layer, which contains the actual play information, the numbers or symbols that determine whether you win. This layer is unique to each ticket and is generated by the prize distribution algorithm.

Over the play data goes a thin opaque coating. This is not regular ink. It is a latex-based compound specifically formulated to be scratch-removable. The coating must be opaque enough to hide the data beneath, resistant to see-through methods (holding up to light, using solvents), and still easy to scratch off with a coin.

On top of the scratch coating goes the decorative top layer: the game graphics, price, branding, and barcode.

The security layers

A scratch-off ticket has between 5 and 12 security layers, depending on the game and the manufacturer. These include:

An overprint pattern that makes the ticket appear uniformly coated, even though the underlying data varies. Without this, you could potentially tell winning tickets apart by subtle differences in the scratch coating.

Confusion patterns printed between the data layer and the scratch coating. These are random visual noise that prevents someone from reading the data by shining light through the ticket at certain angles.

A validation number encoded separately from the visible play data. When a retailer scans a ticket, the terminal checks this validation code against a central database, not the visible numbers. This means even if someone could somehow read the play data through the coating, they could not validate the ticket without the separate code.

Void-if-removed features embedded in the scratch layer. If someone tries to chemically remove the scratch coating rather than physically scratching it, the data beneath is destroyed.

UV-reactive inks that are invisible under normal light but appear under ultraviolet, allowing lottery officials to verify ticket authenticity.

The prize distribution algorithm

The algorithm that distributes prizes across a print run is one of the most closely guarded secrets in the lottery industry. The general approach works like this:

The total print run is divided into packs, typically 300 tickets per pack. Each pack is guaranteed to contain a minimum number of winning tickets, though the specific prizes vary. This is why some players track which pack a retailer is currently selling and how many winners have already been scratched.

Within the full print run, the algorithm ensures that prizes are distributed unevenly enough to prevent pattern detection but evenly enough that every geographic region receives a fair mix. A state cannot have all the top prizes concentrated in one city.

The algorithm also ensures that top prizes appear at irregular intervals in the production sequence. If a $1 million winner appears as ticket number 50,000, the next one will not appear at ticket number 100,000. The spacing is randomized within constraints.

The manufacturing numbers

A popular scratch-off game in a large state like California or New York might have a print run of 30 to 60 million tickets. A $1 game typically has an overall prize payout of about 60 to 65 cents per dollar. A $30 game often has a payout of 72 to 78 cents per dollar.

Higher-priced tickets generally have better return rates. This is partly because the prizes scale faster than the ticket cost, and partly because the fixed production cost per ticket is spread over a higher purchase price.

The production cost of a single scratch-off ticket, including all security features, printing, shipping, and the retailer commission, is approximately $0.10 to $0.25 depending on complexity. On a $1 ticket, that is a significant percentage. On a $30 ticket, it is negligible.

The distribution chain

After printing, tickets are shipped to a central warehouse operated by the lottery commission or its contracted distributor. From there, packs are allocated to districts and then to individual retailers.

Each pack has a unique identifier, and each ticket within the pack has a unique serial number. The lottery commission knows at all times which packs have been shipped to which retailers, which have been activated (scanned into the retail terminal), and which individual tickets have been validated as winners.

When a pack is activated at a store, the retailer can begin selling from it. The lottery terminal tracks which tickets have been sold and which have been validated as winners. This data feeds back into the statewide database in real time.

Why the "hot store" theory is mostly wrong

Some players believe certain stores sell more winning tickets than others. The data shows this is almost entirely explained by volume: stores that sell more tickets produce more winners because they sell more tickets. The percentage of winners per ticket sold is roughly the same everywhere.

There are statistical outliers. Occasionally a single retailer will sell a disproportionate number of large winners. Lottery commissions investigate these anomalies, and in some cases they have found fraud (a retailer scanning losing tickets as if checking for a customer, but actually pocketing winning tickets). The Eddie Tipton case showed that insider manipulation is possible, though his scheme involved draw games, not scratch-offs.

But for the average player at the average store, the retailer has no influence on your odds. The tickets arrived in sealed packs from a central warehouse, and the prizes were determined months earlier by an algorithm.

What this means for players

The overall odds and prize structure are published for every game. Before buying a ticket, you can look up how many top prizes have been claimed and how many remain. Most state lottery websites publish this data and update it daily.

If a game's top prize has already been claimed, the expected value of that ticket drops significantly, but tickets from that game may still be on store shelves. Checking the remaining prizes before buying is one of the few genuinely rational things a scratch-off player can do.

The second rational thing is choosing games with higher return rates. A $10 ticket with a 70% return rate gives you back $7 on average per $10 spent. A $1 ticket with a 60% return rate gives you back $0.60 per dollar. Neither is a good investment, but the return difference is real.