8 Comments
User's avatar
Evan Zamir's avatar

Good stuff. I believe the idea has always been that the underdog should take more 3s because expected value otherwise is worse than the favorite. Your assumption that expected value is the same between 2s and 3s makes this a moot point. Perhaps do a simulation where expected value for one side is lower than the other and see if they win more often by increasing the rate of 3s. Or maybe I should get off my lazy ass and do it lol.

Binomial Basketball's avatar

That’s a great point! I tried something along those lines near the end of the article but definitely a lot more to explore.

NBA Insights's avatar

Great analysis.

thw's avatar

Each team gets to make 100 shots or take 100 shots? Take, right?

ritmica's avatar

That last graph has one of the most devious axes in recent memory.

Also appreciated the footnote.