A Retired Math Teacher Took One Look at the Lottery and Saw Something That Shouldn't Be There

The Kind of Boredom That Changes Things

Retirement, for people who spent their careers doing serious quantitative work, can be a strange adjustment. The problems don't stop being interesting. The tools are still there. The habit of looking at data and asking whether it makes sense doesn't switch off just because the professional context is gone.

This is how Mohan Srivastava — a geological statistician based in Canada whose case became one of the most widely discussed lottery integrity stories in recent memory — ended up staring at a scratch-off ticket and noticing something that the people who designed the game had apparently missed entirely.

Srivastava's story has been told before in fragments, but the full arc of it is still startling: a person with no insider access, no special equipment, and no particular agenda looked at publicly available information and found a flaw that undermined the integrity of a product millions of people were buying every week.

He did it with math. The math was not complicated.

What He Actually Found

The scratch-off ticket that started it all was a tic-tac-toe style game. Srivastava, as a statistical exercise more than anything else, started examining the visible numbers on the ticket — the ones printed in the game's grid that weren't hidden under the scratch-off coating.

He noticed that certain numbers appeared on the visible portion of the ticket only once. Others appeared multiple times. And when he looked at the relationship between the "singleton" numbers — the ones that appeared exactly once — and the winning outcomes, the pattern was consistent enough to be predictive.

He could identify, with a success rate far above random chance, which tickets were likely to be winners before scratching them. Not every time. But reliably enough that the math was unmistakable.

He tested his theory on a sample of tickets. It held.

Then, in a move that most people in his position might not have made, he contacted the lottery commission and told them what he'd found.

The Part Where Nobody Wants to Believe You

Lottery officials did not, initially, respond the way you might hope people would respond when informed of a significant flaw in a system that generates hundreds of millions of dollars and is legally required to be fair.

The first reaction was skepticism. The second was a more formal version of skepticism. Srivastava, to his credit, had documented everything carefully enough that the skepticism eventually gave way to something more productive — an actual investigation.

What that investigation confirmed was that the flaw was real. The game was pulled from sale.

But Srivastava's finding opened a door that was harder to close: if this flaw existed in this game, what did that say about other games? And more importantly, how had it gone unnoticed by the people who were paid to make sure these things didn't happen?

The Spreadsheet as Forensic Tool

The broader implication of cases like Srivastava's — and there have been others, less famous but structurally similar — is about what happens when someone applies rigorous quantitative analysis to data that the public has always had access to but rarely examines.

Lottery draw data, in most U.S. states, is public record. The winning numbers, the draw dates, the prize distributions — all of it is available. Most people look at it the way you look at a phone book: technically accessible, practically ignored.

A statistician looks at it differently.

When researchers and independent analysts have gone back through years of state lottery draw data — not scratch-off tickets, but the drawn-number games like Pick 3, Pick 4, and similar formats — they've occasionally found distributions that don't look the way a truly random system should look. Numbers appearing slightly more or less frequently than probability predicts. Sequences clustering in ways that random number generation shouldn't produce.

Some of these anomalies have innocent explanations: hardware imperfections in drawing machines, software quirks in random number generators, statistical noise that looks meaningful but isn't. Others have been harder to explain cleanly.

When the Pattern Is the Problem

In 2017, a Hot Lotto fraud case in Iowa — prosecuted after Eddie Tipton, a former security director at the Multi-State Lottery Association, was convicted of rigging jackpot drawings — demonstrated that the gap between "this looks statistically unusual" and "this was actually manipulated" can be bridged, but only when someone is willing to do the work of bridging it.

Tipton had allegedly used a rootkit to manipulate random number generator software. The wins he engineered were spread across multiple states and multiple years. The pattern existed in the data. It took investigators willing to look at that data systematically to find it.

The Srivastava case and the Tipton case are different in important ways — one involved a design flaw, the other alleged deliberate fraud. But they share a common feature: the evidence was in the publicly available numbers, and it took someone with the right combination of skills and curiosity to read what the numbers were saying.

What One Spreadsheet Actually Proves

Srivastava has been careful, in subsequent interviews, not to overstate what his finding means. He's pointed out that he reported the flaw rather than exploiting it, partly because the math made clear that the profit potential wasn't as large as it might seem — buying enough tickets to exploit the pattern at scale would have been logistically impractical — and partly because that's just what you do when you find something like this.

But the story's staying power comes from what it reveals about the gap between "this system is secure" and "this system has actually been tested by someone who knows what they're looking for."

Millions of Americans buy lottery tickets every week on the assumption that the game is what it says it is: random, fair, and properly audited. Srivastava's retired-statistician-with-a-spreadsheet story is a useful reminder that "properly audited" is only as good as the auditors — and that sometimes the person who finds the thing everyone missed is someone who just got curious on a Wednesday afternoon and decided to run the numbers.