Friday, December 7, 2007

The Air Must Be Really Thin

I’m sort of on the fence about looking at this year’s statistics. On the one hand, most major teams have only played 2-3 BCS opponents, so the stats could very well be unrepresentative of what we’ll see later. On the other hand, now that college football’s regular season is over, I might as well write about something.

I was thrilled to see Pomeroy also started looking at the numbers on Wednesday. He has a nice article that reintroduces his statistical metrics, and while I found that useful, I thought his first look inside this year’s stats was even more insightful.

For instance, who knew Chris Allen was such a shot hog? Now I really want to watch a few more of the Michigan St. games to see what’s behind this development. On the one hand, this could be a huge mistake by a selfish guard. After all, Neitzel, Morgan, and Suton are all off to incredibly efficient starts, and almost every possession should end with one of those three taking a shot. But on the other hand, maybe Allen’s large number of shots are a function of the team. After all, there are no other three point shooters besides Neitzel, and even though Allen is hitting a disappointing 32% of his threes to date, if teams pack the paint or double team Neitzel, Michigan St. is going to need someone to take three point shots. To me the key question is whether Allen is taking natural shots in the offense or forcing shots. That’s something to watch for.

I also didn’t know to watch out for Butler’s Matt Howard and his 126.5 ORtng. He’s obviously getting a lot of easy layups due to his all-star guards Graves and Green who just happen to have ORtngs of 124.4 and 125.2 respectively, but if he can keep defenses honest by providing an inside presence, this team will maintain its lofty poll status.

Of course, my jaw is still dropping over Malcolm Grant’s 128.1 ORtng and clutch baskets in Villanova’s comeback last night, but Grant is only taking 18.2% of Villanova’s possessions while Howard is taking 29.7% of Butlers.

I was also pleased that Pomeroy emphasized my biggest pet peeve in his column. You need to interpret all data within its context. Games are not played in a vacuum, and if you see a weird trend, it helps to watch the games to actually understand what is happening. For example, you need to know volume of shots to know whether efficiency is meaningful. A few weeks ago I flashed up the numbers for last year’s freshman All-Americans. As I said then, Scottie Reynolds efficiency numbers looked pedestrian compared to the others, but when you consider his importance in his team’s offense, the team needed him to take a lot of shots.

One place I differ is I think if you want to summarize the most information in the least amount of data, I’d show PPG and ORtng instead of % of Possessions and the ORtng. You can mostly back out % of Possessions if you have PPG & ORtng and since general college basketball fans can relate to PPG, I think that is a little more natural pair of numbers to examine. The key advantage of % of Possessions is that it isn’t influenced by pace, but whether or not you want to take pace out when summarizing a player is probably a personal decision.

Speaking of pace, I thought I’d take a few moments to look inside this year’s pace statistics (as found on Ken Pomeroy’s statistics pages).

Something is obviously happening in the water in Colorado. Not only does Air Force continue to have one of the slowest paced teams in the country, look what has happened to three of the other local teams:

Pace         Year  Raw Rnk  Adj Rnk
Colorado St. 2007 68.9  93 68.7  92
Colorado St. 2008 61.0 324 62.0 327
Denver       2007 68.0 109 67.3 136
Denver       2008 56.6 341 58.0 339
Colorado     2007 72.8  20 72.7  13
Colorado     2008 60.2 330 63.1 311

Well what’s happening is that two of these teams hired former Air Force coaches. Both Colorado’s Coach Jeff Bzdelik, and Denver’s Coach Joe Scott (at Air Force and more recently Princeton) have always been known for slow paced teams that work the shotclock and try to get a high percentage look. Colorado St. on the other hand is a big surprise. New coach Tim Miles was not known for a slow-it-down approach at North Dakota St. Perhaps he is trying to shorten the game to make things simpler for his young team (which rotates just one senior), or maybe there really is something in the water in Colorado. I would think someone would start a fast pace team in the area, just to attract the high school recruits with something different. Perhaps the VMI coach is available, although even he has toned back his ridiculously fast paced team from last season.

Pace         Year  Raw Rnk  Adj Rnk
VMI          2007 90.7   1 90.9   1
VMI          2008 80.0   3 78.4   7

Like Bzdelik, Todd Lickliter kept possessions down at Butler, and he’s trying to limit them at Iowa. And you can’t blame Lickliter. In the roaring 70 possession game against Wake Forest, Iowa had an offensive efficiency rating of 67.0 and lost by 9 Meanwhile, in the crawling 53 possession win over Northern Iowa, Iowa’s offense had an efficiency rate of 116.1 and Iowa won by 7.

Pace         Year  Raw Rnk  Adj Rnk
Iowa         2007 65.5 221 67.6 126
Iowa         2008 61.1 323 61.0 330

It can take time to change a culture of losing or winning, but it is nice to know that coaches can at least change some things within their first year. For teams that maintained their coach, dramatic change is less likely, but still possible. Both Notre Dame and Texas have played significantly slower this year.

Pace         Year  Raw Rnk  Adj Rnk
Texas        2007 69.7  74 68.5  97
Texas        2008 64.5 285 63.4 306

Notre Dame   2007 70.5  53 71.5  29
Notre Dame   2008 67.6 216 66.6 232

One reason may be that both teams have clamped down on defense. By allowing fewer easy looks, they’ve caused their opponents to hold the ball longer which has slowed down the game. From a defensive efficiency standpoint:

Def Eff      Year  Raw  Rnk  Adj Rnk
Texas        2007 100.5 131 94.6  62
Texas        2008  95.2 105 92.3  75

Notre Dame   2006 104.5 233 98.5 120
Notre Dame   2007  96.2  49 93.0  49
Notre Dame   2008  89.5  44 89.6  56

I was actually under the impression that Notre Dame’s defense was significantly worse last year than these numbers show, but that may have been because of 2006. It also may have had to do with the defensive efficiency rating of 131.8 against Georgetown in the Big East tournament last year. Either way, Notre Dame’s defense is off to a strong start this year including an impressive 81.7 in Monday’s win over Kansas St.

I’ve always been under the impression that strong defense and slow pace are highly correlated, and if you run the raw numbers you do see that trend, but to a lesser degree than I might have expected. In 2007 a 10 possession faster game only decreased the defensive efficiency rating of a team by 1.5 points. That said, most dominant defensive teams do play at a slightly slower pace. (The key exceptions are pressing teams like Tennessee which can cause lots of turnovers and ultra-talented teams like North Carolina.)

On the flip side, the biggest increase in unadjusted pace has occured at Navy and Creighton.

Pace         Year  Raw Rnk  Adj Rnk
Navy         2007 63.9 269 66.2 187
Navy         2008 77.7  12 77.9   8

Creighton    2007 62.4 301 63.6 289
Creighton    2008 70.4 118 74.1  43

Navy has historically been a faster paced team, so this may simply be returning to what Billy Lange does best. Creighton on the other hand, has historically been one of the slowest teams in the country, so perhaps Dana Altman has decided to reinvent himself after a controversial off-season. Here are a few other notable faster teams:

Pace         Year  Raw Rnk  Adj Rnk
W. Virginia  2007 63.6 277 64.1 273
W. Virginia  2008 72.0  75 69.9 133

Kansas St.   2007 67.6 130 66.5 174
Kansas St.   2008 76.7  14 76.3  19
Indiana      2007 64.0 268 65.4 217
Indiana      2008 72.0  77 71.0  93
Duke         2007 66.1 203 65.9 203
Duke         2008 73.0  56 75.0  31
Louisiana St 2007 65.0 240 63.9 279
Louisiana St 2008 70.8 103 73.4  50

Bob Huggins has made West Virginia faster, while Kansas St. has gotten even faster after he left. Indiana is riding Gordon’s skill to play at a faster pace, while Duke’s 8 McDonald’s All-American’s are pushing it at a different level than last season. Finally, LSU has sped up considerably this year now that it no longer has Big Baby Davis to feed in the paint. But after last night’s colossal collapse, LSU probably wishes they had kept the game about 3 possessions shorter.