Thursday, February 21, 2013

Is James Bell Really Clutch?

A narrative has emerged in the last month or so among Villanova fans that James Bell is a "clutch" three point shooter based on his three point shooting in the final seven minutes of games and overtime, when is 18 for 27.

First off, we must ask what qualifies as "clutch". Is any shot late in any game even if you are up by 40 considered clutch? Does a clutch shot have to happen near the end of a game?

What narratives like this tend to ignore is that those 27 shots that Bell has taken in the final seven minutes of a game or overtime represent only 22% of his three point attempts. That is nowhere close to an adequate sample size. If he were to do this over the course of 100 shots at the end of games, then there would be more to go on, but how can you make such a definitive observation after 27 shots?

There is also the fact that the final seven minutes of a game is an arbitrary starting point. What would his shooting percentage be if you started it three minutes earlier or two minutes later?

The chances are that Bell's shooting at the end of games, which has been excellent will regress to somewhere around his career average, because that's what tends to happen.

The data shows us that James Bell is a good shooter, which is fine, and what limited data we have shows that he's been good at the end of games, but don't call him "clutch" based on such conditional numbers.


  1. I would call him "clutch" for the season as some of those threes were huge in defining this season putting the team to be considered for the NCAA tournament.

  2. Thanks for calling me out dude!!! You the man!

    I think Bell is clutch, but I don't know that those are necessarily the best stats to back his case up. If I get bored I'm going to look up what % of his 3's made were to tie or take the lead. As I watched the Rutgers game I had to turn to my buddy and think "wow it seems like most of his 3's are in very tight score situations"

    How much of a sample size do you really think you're going to get from a college player? This isn't the NBA where you get 82 games +4-7 game series. You have to make assumptions with the sample sizes you've got.

    But I do agree 7 mins is a weird time to start to counting from, and could make his #s look better or worse. Was merely pushing along some findings I found else where for my friends and other followers to see, didn't mean to start a big uproar.

  3. No issues. I wasn't calling you out, it was just the first example I could find to readily link to.

    I think there needs to be a lot more examples to demonstrate this. I think it's acceptable to just say that he's been a good shooter.

  4. I've convinced myself that Bell may or may not be clutch, but when he hits 3's its at clutch times....

    Doing some analysis I'll share the results with you

  5. Check your email, found some interesting data as well as some pretty boring data