Wednesday, December 31, 2008

"Stop Taking Threes!"

That's been a common refrain in sports bars, living rooms and chat rooms for Hoya fans since John Thompson, III has become coach. For the most part, it's been justified. There's no doubt that when the Hoyas become lazy on offense, they tend to jack up the easy shot, which is often a three.

I've long been a defender of shooting the three in college ball. Because of the added point and a relatively close line, it's not a bad shot. It's a better shot than anything from 15' to 20', provided they are both open.

That said, there's no doubt that it should be part of a team's arsenal, not all the team can get. A great offense takes open threes it wants. It does not settle for the shot.

If you look at the last few national champions, you'll also notice a bit of a trend in the teams percent of three-point shots attempted per FG attempt:

Year   Champ         3PA/FGA
2008 Kansas 29.3%
2007 Florida 34.0%
2006 Florida 34.8%
2005 North Carolina 30.4%
2004 Connecticut 25.8%
(median is 33.0%, but % range from 19% to 52%)

For some reason, the next couple of winners have slipped my mind. Needless to say, though, National Champs don't rely exclusively on the three, and often almost ignore it (relatively speaking). Florida's title winners were much more three minded than most, and even their ranks were in the mid-100s. In other words, average for DI, but probably above average for a major conference team (mid-majors and low majors tend to be a bit more dependent on the three).

It's not that teams who do not rely on the three are necessarily better because of that choice; it's that better and more talented teams rely on it less because they have other, better options. If you have a great low post player (which champions tend to have), great offensive rebounding or penetrating point guards, you tend to get a lot more dunks and layups and need to shoot threes less. Of course, you also draw more fouls, but we'll get to that.

Under Thompson, the Hoyas have been somewhere between 35% and 42%. When Jeff Green and Roy Hibbert were both on the team and playing big minutes together, the Hoyas had a 3PA/FGA of 37% and 35%. In the other years, it's been 43% and 40%.

The Hoyas are currently sitting at 39%.

This is not particularly surprising. With only one post player and a burgeoning crowd of elite perimeter players, this shouldn't seem odd.

Now this might be a bit discouraging to Hoya fans. After all, none of those prior National Champs had a number that high. And it's nowhere near the number in 2006-7, the most successful season under Thompson.

But wait! The number is also a bit deceptive, because it doesn't include free throws. Free throws don't register as a field goal attempt, but they are most often the equivalent of a shot taken, and when they are, they are almost always on a two point attempt. In other words, the more a team draws fouls, the more their 3PA/FGA is overstated relative to a team that doesn't draw fouls.

Here's another way to think about it. Using the proxy of .43 shots per FTA (adjusting for one and ones, hoop and the harms, etc.), here's how the % of shot attempts breaks down for the last three years:

Year3 PT2 PTFT

Now, some of those free throws are one and ones, end of game fouling, etc., so they aren't all off two point shots. Still, the percentage of shots taken from three is actually much closer to the Final Four year than it is to last year.

It doesn't mean the Hoyas are going to the Final Four, but it does indicate that the Hoyas have a more diversified offense this year. And that should pay off when the shots aren't falling.


  1. Hey SF,

    Great post, as always.

    Just and FYI that you can actually get the point distributions from here, and his stats go back to 2004 (Esh's last season).

  2. I didn't realize that was there. Interesting.

    I wanted to get attempts, though, not just points because that's affected by your ability to actually make shots.

    One of these days, I'll figure out how to do all the formatting.

  3. Nice post. This stat (3FGA/FGA) is one of my most watched. Keeping this percentage low was actually one of my keys to the UConn game, as it is for most games, and for precisely the reasons you mentioned: exploiting better opportunities.

    I also like the FT correction. I'd even consider thinking about this stat as a percentage of possessions as a 3FG usage%.