Appropriate for March. Based on historical analysis of NCAA men’s basketball games.

Posted by SavantDJ

11 comments

  1. Is the x-axis supposed to start at 0? The hotspot at the top left seems cool but now it confuses me.

  2. I’m confused. Is this saying that the team trailing by 1 in the first minute of the second half has a 90% chance to win? That doesn’t seem right

  3. This makes March seem a lot less mad… great analysis!

    Could you share the code you used to make this? I’d like to learn how to do similar work.

  4. dont use a continuous color palette. itd better to have discrete colors for each probability.

  6. Teams down by 1 with 6 minutes left win ~25% of the time? That’s definitely not right 

  8. I think “win probably” should probably be interpreted as “added win probability” or to be a normalized 50% Win vs Loss at the beginning of the game. Or another way to say it is your win probability based solely on score and not ability.

  9. That doesn’t make any sense. There shouldn’t be anything above 50% in a point deficit. Otherwise this means you’re better off losing at the beginning

  10. My HS friends and I all had older sisters who played basketball growing up. We used to try to determine the “point of insurmountability” for a HS girls basketball game. We determined that it was like 13 points.

  11. If I’m reading this right you are almost guaranteed to win if you are behind by 1 point when the second half starts.

    Its a bold strategy cotton, let’s see if it pays off for them.

Comments are closed.