Project Defensive Score Sheet: Kansas@Iowa State

Here’s Game 2 of my stab at charting individual defensive stats.  Check Game 1 for an introduction and explanation, otherwise you’ll have no clue what the alphabet soup is trying to tell you.

Before I get to the numbers, I have to complain about the TV coverage last night.  There were multiple times where I missed a turnover or shot attempt because the producers decided to show close ups of bench players celebrating a dunk, or something along those lines.  And then the announcers failed to clue me in on what had just happened (“Great play there by Iowa State” isn’t very informative, guys).  So, my stats didn’t quite line up with the box score at the end of the game, and I lumped the ~5 plays I was missing into the “Team” lines.  Ugh.

More notes and comments are after the break, but first, the tables (click for larger view):

For NBCSports.com: Purdue’s Life After Hummel

In what to me is an awesome piece of news, I’ve been invited by Mike Miller to write a few articles for NBCSports.com this season.  The first one went live yesterday, and takes a statistics-oriented look into the question of how Purdue has managed to play like an elite team this year without Robbie Hummel, after struggling mightily without him at the end of last year.

As a teaser, here’s one of the tables from the article, showing that E’Twaun Moore was forced to carry a ridiculous load after Hummel went down last season, and struggled to do it.  But he’s been able to dial his Shot% back down to normal-for-him this year, and has seen his offense numbers bounce most of the way back to their 2010 pre-Hummel-injury levels:

To read the rest of the piece, head over to NBCSports.com.

Project Defensive Score Sheet: Kansas@Michigan

As I mentioned a couple weeks ago, I just read (most of) Dean Oliver’s Basketball On Paper for the first time.  A few of the chapters focus on defense, and there’s one specifically about individual defensive stats.  Oliver laments the lack of useful defensive data, and talks about a WNBA pet project of his: Project Defensive Score Sheet.  This was (is?) an effort to keep track of individual defensive credit or blame for all of the shots and turnovers a team allows.  After reading the chapter, I rushed to the internet in hopes of finding the results of his efforts, but was disappointed to find very little information available, and very few people attempting to continue or expand on the project.  [See the end of this post for a list of what I did find.]

I thought it would be fun, and possibly illuminating, to keep track of Oliver’s individual defensive stats for some college basketball games.  In a happy little coincidence, the top defensive team in the country happens to also be my favorite team, so I decided to try to chart as many Kansas games as time and TV coverage would allow, starting with this past Sunday’s game against Michigan.

It’s amazing how different the game-watching experience was when I was forced to pay close attention to defensive players.  I would never have otherwise noticed how active – and cagey – Jordan Morgan is.  More on that later.  Anyway, at this point, I don’t have any brilliant ideas on what to do with the numbers, so I’ll simply start by posting the results of each game.

Here’s the table (click for larger view); the explanation is after the break.

Gone, Gaddy, Gone

[EDIT: Ken Pomeroy posted a treatise on plus-minus which pointed out there’s a great deal of error inherent in the use of +/- data, something which has also been pointed out to me by John Ezekowitz.  At this point, I’d advise you to take everything I wrote below with a huge grain of salt.  And I doubt you’ll be seeing any more +/- analysis from me.]

About a month ago, I predicted that Duke would be fine without Kyrie Irving.  My reasoning was that, in the games while Irving was still healthy, Duke had actually performed better with Irving on the bench than they had while he was on the floor – about 7 net points per 40 minutes better.  Well, since the toe became The Toe, Duke has actually played ever so slightly better, with a post-Irving-injury adjusted efficiency margin of 39, compared to 37 with Irving:

Obviously that difference is insignificant, but the main point is that, so far, they’ve been fine.

I fear Washington’s prognosis for life without Abdul Gaddy is not so optimistic.  The same +/- data that soothed Blue Devil worries paints an uglier picture for the Huskies (click to enlarge):

Whereas Duke was 7 points per 40 minutes better with Irving on the bench, Washington has been almost 7 points per 40 minutes better with Gaddy in the lineup.  Some of the Twitter buzz hoped that the injury would give freshman CJ Wilcox a spot in the limelight.  That’d be great, as he has a team-high 131.8 ORtg in his limited time.  Taking a quick look at his +/- data, though, Washington has struggled with him in the game, despite his good offensive rating.  A safe bet here would be that Wilcox’s defense is a step down from Gaddy’s, though the KenPom individual stats don’t paint a big defensive difference between the two, other than Wilcox’s more frequent fouling. Whatever the reason for the struggles, somebody will have to step up if Washington hopes to retain its lofty rating.  If I’m Seth Davis, I’m changing my Washington grade from BUY to HOLD.

[On a side note, it’ll be a shame if this injury does end up impacting Washington, as I expect it to.  The Huskies were looking like a great ratings litmus test, as they were in the top 5 in both Pomeroy and Sagarin’s PREDICTOR, yet #23 and unranked in the two major polls.  Now watch, instead of tearing through the Pac-10 like their namesake through the Delaware, they’ll decline due to the impact of this injury, and the “you either win or lose” crowd will get a chance to trumpet their demise as a black mark against predictive ratings.]

All I Want For Christmas Is An Offensive Rebounder

In the spirit of the season, I thought I’d dole out some gifts to 5 teams that have some glaring weaknesses.

PITTSBURGH

Pitt has the #1 offense in the nation (all ranks throughout the article are from the Pomeroy ratings), but ranks only 43rd in defense, which is a pretty big red flag, in terms of contending for a national title.  Their main problems are a lack of turnovers (234th) and steals (246th), 3pt defense (223rd), and few blocks (114th).

Santa’s Bringing Them – Ayron Hardy, Jacksonville.  The  6’5” senior was named A-Sun Defensive Player of the Year in 2008-09.  He averages a steal every 20 opponent possessions (13th), blocks 4.7% of opponent shots (242nd), and helps anchor a Jacksonville defense that holds opponents to only 29.1% from deep (41st).  The fact that he’s an extremely efficient role player on offense (130.2 ORtg, 16.7 %Poss) is a sweet bonus.

Mini-Preview: UNLV-Kansas State

[EDIT: After writing this, Jacob Pullen and Curtis Kelly were ruled ineligible fro Kansas State.  So, umm, never mind.  KSU is no longer a “good” offense, so UNLV should be fine.]

[This post is really just an excuse to use a spreadsheet I whipped up, which takes a year and a school as input, and automatically creates a chart like the one below.]

UNLV tips off against Kansas State in Kansas City a mere hour from now.  The Rebels started off the year with a bit of hype, and lived up to it at first, topping Pomeroy ratings #10 Wisconsin at home by 3, and blowing out some scrubs.  But they’ve stumbled over the last 10 days, following up a very understandable loss at #13 Louisville with a much less forgivable home loss to #100 UC Santa Barbara.  In truth, their problems started to show up 2 games previous to the Louisville loss, when they beat a poor Nevada (#202) team by only 12, but sometimes a W can have a lipstick-on-pig effect.

It’s clear that a 12-point win over #202 isn’t fantastic (Pomeroy had predicted UNLV by 18), but sometimes it’s hard to get a feel for exactly how good/bad a performance is.  We can use opponent ratings to shed some light on the issue by using Pomeroy’s efficiency prediction formula:

Predicted Offensive Efficiency = ([Team Adj Off] + HFA) * ([Opp Adj Def] + HFA) / [Lg Avg Eff]

For each game that a team has played, we can replace the predicted efficiency with the team’s actual raw efficiency in that game, plug in their opponent’s rating, the league average rating, and the appropriate home field advantage (+/- 1.4% for each team, in a normal H/A situation) , and solve for [Team Adj Off].  That gives us the team’s single game adjusted offensive efficiency rating – essentially, this is how efficient a team would have been if they played exactly the same, but were facing an average opponent on a neutral court.  We can do the same for defense, and from those two numbers we can calculate the efficiency margin (which I find more intuitive to use) or Pythagorean rating (which Pomeroy uses to rank teams).  As a last step, we can take the single game Pythagorean rating, pretend that’s how the team has played the whole year, and see where they would rank in the Pomeroy ratings.  That allows us to say, for example, UNLV played like the #232 team in their loss to UC Santa Barbara.  The chart below shows this “played like” rank for each of UNLV’s games so far. (It also shows, from left to right, the location, opponent, opponent ratings, raw game efficiencies, and adjusted game efficiency ratings.)

 

You can see that UNLV played great through the first 7 games, and then has really struggled over the last 5.  They’re still rated #22 by Pomeroy, but taking a close look at the individual game adjusted ratings reveals something interesting.  Their overall defensive rating so far is 89.7, but that’s largely because of ridiculous defense in a few games against poor offensive teams.  Here are the teams they’ve managed to post a sub-90 defensive rating against so far, along with those teams’ offensive ratings (keep in mind, average is 100):

• UC Riverside (90.6)
• SE Louisiana (91.7)
• Illinois State (97.1)
• Southern Utah (90.9)

That’s it.  Their 30th ranked defensive rating comes in large part from really cranking the screws on the little guys.  If they want to compete in the Mountain West this year, that’s going to have to change.  And if they want to win their semi-road game tonight against Kansas State (offensive rating of 106.7), that’s going to have to change.  I’m not saying it definitely won’t, but I’d say KSU has a better chance than UNLV of bettering their Pomeroy prediction (KSU by 1) tonight.

Bruce Pearl Before Swine

As I’m sure anybody reading these words must know, Tennessee managed to achieve “fuhreal” status on Saturday, and fritter it away by Tuesday.  It was an impressive about face, as they played like the #1 team in the country at Pittsburgh, then like UC-Irvine at home against Oakland:

A few messages insinuating that the loss showed Bruce Pearl’s true colors flashed across my Twitter feed last night, and it made me wonder if this kind of massive quality swing was really as rare as people were making it out to be.  So I checked.

I found every instance since the 2002-03 season where a team:

1. Won on the road vs. a top 10 opponent (by Pomeroy rating)
2. In their next game, lost to a team ranked 50 or lower

I figured that would catch only very impressive wins.  And I couldn’t really make the loss section much more strict, because Oakland is a pretty decent team (#71 as I write this).  With the above criteria, there were 9 cases from the last 8 and a half years.

And, wow.  A third of them are Bruce Pearl’s Tennessee teams.  So for non-Tennessee teams, the chances of it happening any given year are around a quarter of a percent.  For one of Pearl’s Tennessee teams, the chances are around 50%. Apparently there’s something to the idea that his teams are prone to a big letdown after a huge win.  Of course, the fact they even have 3 road wins over top 10 teams is something to be proud of.  I wish I would have kept track of how many teams didn’t lose their next game after a top-10 road win, so I could account for opportunities.  But at this point I’m not going to go back through again.  My eyes are already bleeding.