March Madness 2026 has arrived with the NCAA Men’s Basketball Tournament running from March 17 through April 6.
Just how rare is a perfect bracket — and is it even possible?
Here’s a look at the odds and how prediction market company Kalshi is looking to reward the near impossible.
• DraftKings stock is trading near recent lows. What should traders watch with DKNG?
Kalshi Offers $1 Billion Prize
Prediction market company Kalshi is offering a $1 billion prize if anyone can complete a perfect bracket of the round-of-64 games through the national championship, which doesn’t include the First Four games set for Tuesday and Wednesday.
The contest is backed by Susquehanna International Group with no purchase or deposit required from users.
Along with the $1 billion prize for a perfect bracket, Kalshi will pay $1 million to the top-scoring bracket and award $1 million to charities and scholarships as part of the promotion. The company said in a video that a perfect bracket is “the holy grail of sports math.”
“Warren Buffett offered $1B for a perfect college basketball bracket in 2014. Kalshi is bringing it back. No deposit requirement. Just fill a bracket. Your odds of winning are 1 in 120 billion. Very low. But not zero. Good luck,” Kalshi co-founder Tarek Mansour tweeted.
Mansour’s tweet may have been conservative on the odds.
The Odds Of A Perfect March Madness Bracket
The odds to correctly predict the outcome of all 63 games during the tournament are 1 in 9.2 quintillion based on a 50-50 coin flip and all possible outcomes (that’s 1 in 9,223,372,036,854,775,808 in numerical form).
Late DePaul University professor Jeff Bergen projected the odds could be 1 in 128 billion for those who know about basketball and follow seed trends.
Bergen estimated it could take 2,300 years for every person on the planet, completing a bracket every minute, to cover all 9.2 quintillion outcomes.
The DePaul professor said that you would have a better chance of winning the Powerball and Mega Millions …
This post was originally published here
