I have been playing scratch off games for quite some time now. I have always been fascinated by math, and after watching Jerry and Marge Go Large (the true story of a retired guy who found a loophole in a lottery game and legally made millions) I got obsessed with one question: is there a real, pure math edge in the lottery?

Quick background: I am an aerospace engineer with a Master's degree, and I have spent the last 13 years working on highly complex math problems in the industry. I like building things to answer questions I cannot stop thinking about. So instead of guessing, I decided to just measure it.

It turns out that most state lotteries publish how many prizes are still unclaimed for each scratch off game. That is the raw material for expected value (EV). If you know the prizes left and roughly how many tickets are left, you can calculate how much of every dollar comes back, on average, right now.

So I had an idea: what if I built a web app that tracks the expected value for every active scratch off game? That way, I would know exactly which games are statistically the best to buy and which ones are the worst.

I ended up building 42 scrapers (one for each state lottery site, which includes 41 states plus DC) that pull this data daily. Then, a calculator runs the EV on every active game. For states that do not publish complete prize data, I model the missing pieces with Bayesian estimates rather than guessing. The system is currently tracking 2,745 active games.

One thing that annoyed me while researching this: almost every lottery site and scratch off tool ranks games by "best odds," which is highly misleading. Odds tell you how often you win; they say nothing about how much you win. A game can have fantastic odds and still be a terrible deal, because most of those wins are just getting a fraction of your money back.

What actually matters is expected value. For context, the national average EV is $0.68 per dollar spent (meaning a typical scratch off keeps around 32% of your money).

However, a handful of games actually compute above breakeven right now. Here are the top games by expected value per dollar for a few major states (an EV of $1.50 means you statistically get $1.50 back for every $1 you spend):

Florida

1. Scorching Hot 7s (EV/Dollar = $1.03)
2. Florida 300x The Cash (EV/Dollar = $1.02)
3. $2M Gold Rush Multiplier (EV/Dollar = $0.90)

California

1. Instant Prize Crosswords (EV/Dollar = $0.92)
2. 7's (EV/Dollar = $0.89)
3. Celebrate 2026 (EV/Dollar = $0.87)

New York

1. $1M Gold (EV/Dollar = $1.44)
2. Plus the Money (EV/Dollar = $0.91)
3. New York Millionaire (EV/Dollar = $0.91)

Note on the high EV games: These are usually end of life games where most of the tickets have sold, but a top prize is still floating around out there, causing the mathematical EV to spike.

You can check out the entire dashboard and see the data for other states here:https://scratchiq.io?utm\_source=reddit

TL;DR: I am an aerospace engineer who got obsessed with lottery math. I built a dashboard that tracks and calculates the daily updated Expected Value (EV) for scratch off games across 41 states plus DC so you can see which games are actually worth playing.