Jammin’ on the James
by Ken Pomeroy on Wednesday, February 28, 2007
Here’s the CAA log5 breakdown, as contributed by many. Surging ODU has the best odds out of the 2-seed.
Qtrs Semis Finals Champ 2 ODU 100 89.2 67.1 42.5 1 VCU 100 88.8 54.9 27.6 4 Drexel 100 85.4 40.4 17.5 6 GMU 91.3 47.6 14.7 5.6 3 Hofstra 100 51.6 15.1 5.4 7 Towson 68.9 8.9 2.8 0.6 5 N'eastern 79.7 13.6 2.4 0.4 8 Bill & Mary 68.0 9.0 2.0 0.3 10 UNCW 31.1 2.0 0.3 0.0 9 Georgia St. 32.0 2.2 0.3 0.0 12 Delaware 20.3 1.0 0.1 0.0 11 JMU 8.7 0.8 0.0 0.0
It’s That Time of Year
by Ken Pomeroy on Tuesday, February 27, 2007
It’s time to calculate conference tournament odds based on log5 principles. The Valley Ledger has the data for Arch Madness.
This is strictly a volunteer effort. If you go through the trouble of computing the odds for a conference tournament, I’ll post them. For those interested, 11.5 is the exponent of choice this season.
The Value of Consistency
by Ken Pomeroy on Monday, February 26, 2007
One of the big issues in my mailbox is how consistency impacts a team’s prospects in The Tournament. More specifically: how can we expect Southern Illinois (Pythagorean rating: 34, Consistency: 1) and Nevada (Pythag: 63, Consistency: 2), teams that have dominated their respective conferences without piling up big margins of victory, to perform in The Tournament? Keep in mind, with just three prior years of data, we don’t have a whole lot to go on statistically. But here are the teams that had some things in common with the canines hoping for a deep run this season:
2004 Northern Iowa (Pythag: 71, Consistency: 11) Result: First round loss to eventual Final Four participant Georgia Tech. Close game that wasn’t decided until the final minute. (Note: UNI didn’t win the MVC that season, but was the closest thing fitting my criteria in 2004.)
2005 Gonzaga (Pythag: 32, Consistency: 12) Result:…
Trying to Improve
by Ken Pomeroy on Friday, February 23, 2007
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 Foul That Wasn’t
by Ken Pomeroy on Thursday, February 22, 2007
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.
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’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…
Key to the Game
by Ken Pomeroy on Wednesday, February 21, 2007
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…
Plus/Minus Redux
by Ken Pomeroy on Monday, February 19, 2007
Raw plus/minus is nice, and I think interesting, if not statistically significant, for an individual game. But step two is to calculate an on/off rating a la the work Roland Beech does on NBA players.
This kind of analysis is problematic for star college players, because many of them play nearly all of their team’s non-garbage time minutes. We need a pile of data in the off category to make the results meaningful.
It just so happened that when the Pittsburgh Post Gazette’s Ray Fittipaldo wanted to step outside the box and do a story on the plus/minus concept, it gave me an opportunity to carry out this kind of analysis on a team where the star player does get some meaningful time on the bench. Aaron Gray is widely thought of as All-American/Big East POY material, and yet he averages 29 minutes a game. Not surprisingly, on/off data…
The First Poll is the Best Poll
by Ken Pomeroy on Friday, February 16, 2007
I’ve been kind of distracted by work on the HD box score. Here’s Wednesday’s Ole Miss/LSU game, a 71-70 Rebel win in 60 (counted!) possessions. A decidedly bizarre game where Ole Miss (now leading the SEC West, kudos to Andy Kennedy) turned it over just four times but saw that advantage offset by LSU’s slightly better shooting and much better rebounding. I really preferred to run through Radford’s 114-107 win over VMI, but the Radford scorer’s table apparently gave up on tracking substitutions at some point in the second half. I don’t blame them.
I hardly pay attention to the opinion polls, but one poll I do think has merit is the very first one of the season. It’s when the games start being played that the polls go to crap. LSU is headed to the rare fate of being a preseason top 10 team, only to…
HD Box Scores
by Ken Pomeroy on Wednesday, February 14, 2007
A lot of folks send in great tips on some cool box scores that appear during the season. I wish I was dutiful enough to print them all. I did get a lot of mail about Butler’s offensive assault on Cleveland State last week. A historic game indeed - Butler’s 92 points on 53 possessions was the most efficient game (1.73 PPP) in the modern era*, knocking off DePaul’s annihilation of Syracuse last season (1.65).
It seems like a good time to reveal my latest project - the HD box score. Much like HDTV, it’s not necessary to view a box score in HD to understand what went on during a game, but it does add some resolution to your experience. I’ve got some more work to do on this, but the player data is pretty much set. The essence of the HD box score is that opportunities are…
Gray Matter
by Ken Pomeroy on Tuesday, February 13, 2007
[Note: I have no ftp service today, so there will be no updates to the stats/ratings pages.]
Q: How much is a star player actually worth to his team?
A: To be revealed later this week. But based on the data I have, I found it interesting when Jamie Dixon went to the coaches playbook and pulled Aaron Gray after he picked up his 4th foul against Louisville last night. Sure, Pitt was down by 15 to Louisville, and the game was slipping away anyway. But there was just over 14 minutes remaining, and at that moment, Jamie Dixon conceded a loss.
Subject: Georgetown
Ken,
This team is on a run…..just watched them embarrass WVU…..what’s your take on these boys running the table in the Big East? They host Pittsburgh on the 24th I believe to avenge an earlier loss….I just don’t see anyone in the nation stopping their offense,…