We all know how the Olympics basketball ended. The outcomes make for a pretty tiny sample though. How the oddsmakers rated the countries would bring us much closer to their actual talent levels. Before we get to that, I’ll begin by using linear regression on actual scores to rank countries:

Now I’ll use linear regression on the betting lines. However, the oddsmakers’ power ratings are not static — they update them after every game. I found the best fit when I use 90% of the betting line and 10% of the final margin as the independent variable and I put more weight on the latter games.

By comparing the tables, you can get some sense which teams either overachieved/underachieved or were overrated/underrated. Also noteworthy is that France likely benefited from home field advantage (which the bookmakers surely factored in), but I can’t quantify how many points it was worth.

Sorry if some parts of this part are confusing. I’m on a road trip and am behind the schedule so I have to go now. Hopefully some of the more statistically savvy fans here will be able to fill the gaps.

Edit: I’ll post the same stuff for the 2020 Olympics in the comments tomorrow.