I have been working on a machine learning model to predict NBA games. After tailoring the model to the postseason, it predicted something that I thought you guys might be interested in.
My model simulates the playoffs with features like elo, games played, game number, series margin, etc updating dynamically as the series progresses.
I ran 2 versions of the model 10000 times each, one with ELO and one without. In both models you guys came out as the most likely winner. (Not by a huge margin but still). I know a lot of this depends on JB staying healthy, not sure how much that knee will affect him. Still, I’m excited for you guys, I think your team composition is fantastic, and you will go all the way.
If anyone wants to see more about my predictions check out my substack: https://open.substack.com/pub/nbainsights/p/predicting-the-entire-nba-playoffs?r=5g57ct&utm_medium=ios
11 comments
Link the substack MFer!
I think your model is amazing. No notes.
STOP THE COUNT!
Found it in the Clippers’ sub:
https://nbainsights.substack.com/p/predicting-the-entire-nba-playoffs?r=5g57ct&utm_medium=ios&triedRedirect=true
i simmed a season of NBA2k and the Thunder beat the Cavs in the Finals
Wtf, I like AI now
If youre willing to share, what type of model did you use?
Where do you get the Elo metrics. I’m assuming you are talking about an Elo rating much like Chess. In which case it is named after Arpad Elo. It is not a acronym so it doesn’t need to be capitalized.
Nobody cares
*LISAN AL GAIB!*
Yah no shit
*