How Scratch Cards Really Work, and the Statistician Who Cracked One
A scratch card feels like a tiny lottery in your palm: rub off the latex and chance decides on the spot. Except that is not how it works at all. The winners were chosen before the card was ever printed, the number of them is fixed in advance, and the whole deck is manufactured to feel winnable. Usually none of that matters to the person holding a coin. But in 2003 a statistician in Toronto looked at a scratch game and realised the losing tickets were quietly announcing themselves - and that he could sort winners from losers without scratching at all.
A Scratch Card Is Not a Mini Draw
The key misconception is baked into the gesture itself. Scratching feels like the decision is happening right now, under your coin. In truth the result was sealed long ago - fixed at printing, alongside thousands of other cards in the same run. You are uncovering an answer, not generating one.
That distinction is not merely academic. It explains why top prizes can be exhausted while the cards keep hanging on the rack - and why the very structure that makes a card exciting once made it vulnerable.
Three Truths Behind the Latex
Winners are chosen at the factory
The print file for a scratch game already encodes exactly which serial numbers win and how much. Buying the card doesn't roll any dice; it just reveals a result decided weeks earlier on a printing press. A scratch card is a sealed answer, not a live event.
The prize pool is finite
A game is a fixed deck. When the back says '1 in 4 wins', that ratio holds across the entire print run, not your handful of cards. Crucially, the big prizes are a fixed count too - once they're claimed, the remaining cards are guaranteed losers for the top tier, even as they keep selling.
They're engineered to feel winnable
Designers lace the deck with frequent tiny wins and deliberate near-misses so the experience stays exciting. That engineered structure is the whole point - and, as one game showed, that same structure can accidentally leak which cards are which.
The Man Who Read a Scratch Card
A gag gift and an idle question
Mohan Srivastava, a geological statistician in Toronto, was handed a couple of scratch tickets as a joke. He won a small amount - and then, out of professional habit, asked a question almost no one asks: if the winners are printed in advance, is the ticket really as random as it looks?
The singleton pattern
The game was a tic-tac-toe grid. Studying the numbers printed openly on the card - the ones you don't scratch - he found that certain digits appearing exactly once, 'singletons', clustered on winning tickets. The visible face of the card carried a fingerprint of the hidden result.
Sorting winners without scratching
Testing the rule, he could predict which unscratched tickets were winners about 90 percent of the time. In principle he could walk into shops, pick out the good cards and leave the rest. He worked out it might net a few hundred dollars a day - real money, but not worth abandoning his career for.
He reported it instead of exploiting it
Rather than quietly cash in, Srivastava phoned the Ontario Lottery and Gaming Corporation to warn them. Convincing a lottery that its own game leaked took some doing, but the vulnerable game was pulled. The story reached a wide audience through a 2011 Wired investigation.
Srivastava's find was about one badly designed game, not scratch cards as a whole. It rhymes with another story we've told: the weak link wasn't the draw, it was the design around it. Are lottery draws rigged?
The Part That Still Matters: Prizes Remaining
The 2003 loophole is closed. The consequence of the fixed prize pool is not. Because a game contains a finite number of big prizes, the last top prize can be claimed while cards keep selling for months. Anyone buying after that pays full price for a ticket that can no longer contain the headline prize at all.
That is exactly why many lotteries now publish, for each instant game, how many top prizes are still unclaimed. It is the one genuine piece of information you can use to sharpen an otherwise random decision - and surprisingly few buyers ever look at it.
Five Points for Smarter Scratching
- 1Check the 'prizes remaining' page before you buy. Most modern lotteries publish, per game, how many top prizes are still unclaimed - a game with none left is a bad buy at any price.
- 2Trust the overall odds printed on the back. That ratio is the real one, set across the whole print run - not the vibe of the display stand.
- 3Drop the 'this machine is due' instinct. There is no machine and no streak; each game is a fixed deck being sold down.
- 4Don't expect a repeat of Srivastava's trick. After 2003, printing and number design were reworked precisely so the visible card no longer leaks the result.
- 5Remember it's still a negative-EV bet. Better information changes which card you'd pick, never the fact that the house keeps its margin. Treat it as entertainment.
The Honest Framing
It's tempting to turn this story into a hack. There isn't one anymore. The lesson is different and more sober: instant tickets aren't a game against chance at the moment of purchase, they're the sell-down of a pre-set deck - with an operator margin that no amount of observation optimises away.
The only legitimate edge left is transparency: read the overall odds and check the prizes remaining. Everything past that is entertainment - and should be budgeted as nothing more.
Prefer Draws to Latex? Check Your Numbers
With the big draw-based lotteries you can at least run the numbers: see how a fixed combination would have done across years of real draws, across every major lottery we cover.
Disclaimer: This article summarizes publicly available information about instant lotteries and the documented case of Mohan Srivastava. Rules and transparency tools differ by lottery and country. Lottery remains a game of chance. Play responsibly. LottoROI is not affiliated with any lottery operator.