How the experiment works

To measure, in public, what happens when ordinary people wager imaginary money at real sportsbook odds over a whole season. We aren’t trying to prove gambling is stupid. We’re trying to find out.

1. Most participants will finish the season with a negative return on what they staked.
2. How confident someone feels about a bet will barely predict whether it wins.
3. On average, participants won’t beat the closing line, the final price before kickoff.

Everyone gets 500 credits per season. They have no cash value and there are no refills. A single wager is capped at 100 credits, and there’s a published maximum payout per slip so one freak parlay can’t distort the results.

Prices come from a licensed feed of real US sportsbooks, refreshed hourly and again shortly before kickoff. The price you take is locked when you bet and used for grading, even if the line moves. At typical −110 pricing you need to win 52.4% of bets just to break even. That gap is the house edge. What −110 actually means.

Wins, losses and pushes follow standard sportsbook rules. Postponed and canceled games void the wager and refund the stake. A parlay with a pushed or voided leg pays as if that leg weren’t there. The full rules are published on the betting rules page.

Not betting keeps all 500 credits. Every bettor is measured against that line: how many end up below where they started.

The random bettor copies each participant’s stakes and odds but picks sides at random. It separates picking skill from the cost of the house edge.

Fake money changes behavior. People who sign up for an experiment like this aren’t a random sample. And one season is short: profit over a few months is mostly luck. That’s why we report closing line value alongside profit.

We publish aggregate statistics only, and only for groups big enough that nobody can be singled out. Your individual history is never shown unless you share your own report. You can delete your account at any time.