Friday, February 23, 2007

Trying to Improve

Ken,

You seem universally skeptical of intra-season “team improvement” stories.  For example, with Duke you give a great alternate explanation of their possible future success.  What do you make of Vanderbilt’s improvement this season?  From a 1-3 start with losses to Wake and Furman to a 7-2 mark in their last 9 SEC games, beating Florida to lengthen their home winning streak to 12.  The “party line” of the Commodores is that after their loss at UGA, the team realized that it needed to pick up the defensive intensity, and Derrick Byars realized he needed to basically take over games on the offensive end at times.  Do the numbers back up the party line, or is Vanderbilt just another good luck / bad luck story?

Jake

Jake describes my views accurately, although I think the quotes should apply to the word stories. I don’t doubt that teams improve during the season. Every coaching staff in America is obsessed with making their team better, and nearly all of them are successful over the course of a season. But yeah, I’m pretty skeptical of the “stories” where the light bulb goes on, and a team suddenly realizes it needs to play better defense or a player decides to take over games. This game is a little more complicated than that.

I do believe that teams like Texas and Louisville should improve more relative to the rest of the D1 nation, just because they rely on freshmen so much, and freshmen should have a steeper learning curve than upper classmen. Outside of that kind of situation and obvious personnel changes, I need to see a better reason than a team meeting or some other form of soul-searching to explain a team’s newfound winning ways.

However, the stories will continue to get written. With a few exceptions, that’s what sportswriters do. They are trained to write stories rather than do analysis. Otherwise, we wouldn’t have had Oregon, Oklahoma State, and Wichita State thought of as potential Final Four teams at various times this season. It doesn’t make for good copy to say that a team’s great record, even one that includes a few quality wins, is not a good indicator of its performance in some cases. Nor does it sell in Poughkeepsie to say that teams that experience four-game losing streaks, like Michigan State, Georgia Tech, and Duke, may be the victims of a tough schedule and bad karma, and could actually still be capable of great things.

Speaking of Michigan State, check out the graphs our pal David Hess produced. Real analysis included!

Thursday, February 22, 2007

The Foul That Wasn’t

These aren’t the games I would have chosen to run, but I’m at the mercy of SID’s and webmasters from around the country on acquiring the necessary data to create boxes. And unfortunately, the SEC guards substitution data like it’s gold.

Penn State/Ohio State

It’s not often you see a cellar-dwelling team lose a starter in the first minute and still hang with the #2 team in the nation in their building for 35 minutes.

Georgetown/Cincinnati

Georgetown’s Jeremiah Rivers was credited with a foul when he wasn’t playing, early in the second half. (I had to remove this for my program to work properly.) The foul should have been credited to Roy Hibbert. It didn’t matter much - that would have been Hibbert’s second foul, and when Hibbert eventually picked up what was officially his fourth foul with 4:29 to play, he was removed from the game by John Thompson III and didn’t return.

Wednesday, February 21, 2007

Key to the Game

We always talk about “shooting” as one of the four factors, and easily the most important of the four. But it isn’t really shooting, more accurately it’s “making shots.” And in my mind, making shots has two components - shooting and shot selection. You don’t have to be a good shooting team to shoot a high percentage if you’re getting mostly dunks and layups. Likewise, it’s every defense’s dream to have their opponent take contested 15 to 19 foot shots all game.

Yet, we just roll shot selection into shooting percentage and say a team shot well or poorly based on that stat alone. Well, the HD box score has come to your rescue to add some clarity to this matter.

The fine folks who score NCAA ball games take great care, or at least some care, to distinguish shots between dunks, layups, and jumpers. And the cause du jour around here is now how 2-point jumpers are often the key to the game. Every time you hear an analyst state that free throws were the key to the game, just think in your mind that 2-point jumpers were really more important, because chances are, they were.

In the admittedly very limited sample I have looked at, 2-point jumpers are about as unlikely to be made as 3-pointers. Without the one-point bonus in your offensive account, this would seem to be an area to avoid.

Indeed, in Marquette’s 80-67 win over Villanova on Monday night, the Wildcats took 19 of their shots from no-man’s land, compared to 12 for Marquette. Both teams were horrible on those attempts - a combined 7 for 31, but the seven “worse” shots Villanova took could be thought of as the difference offensively.

Or check out last Saturday’s Air Force/Colorado State game. At Air Force, it’s nearly mandatory to avoid the 2-point jumper. The teams were a combined 9 of 25 from this zone, but Air Force only took eight of those shots. The shots they didn’t take from there were shot from beyond the arc. The Falcons went 10 of 24 on 3’s while CSU was 5 of 13. Air Force would win by 9.

This concept doesn’t work out so nicely for every game, but normally 2-point percentage is heavily dependent on the types of shots that are taken.

(Oh yeah, if you look at the second box, you can tell that my possession counter is not quite rock solid yet. CSU should have a one-possession advantage for the game, but is shown with a 58-55 edge. Nonetheless, I’d like to point out that I have CSU center Jason Smith with 51 offensive possessions and just 41 defensive possessions, which has to be some kind of record for possession imbalance. Maybe the difference was really only eight, but still, whoever is in charge of substitutions for the Rams deserves a raise.)