"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Wednesday, December 28, 2011

Blogger says Vandy with Ezeli his top pick to stay within single digits or win; MU must leverage defensive and bench superiority

A look at the match-ups tonight shows that with the return of 6-foot-11 Festus Ezeli against Marquette, the starting squads are virtually even offensively and Vanderbilt is either 2- or 3-inches taller at every position. That combination has one blogger saying his surest college hoops pick is Vanderbilt staying within single digits or winning.

Here is a match-up player-by-player of the Value Add (% of points added to offense), Defensive Subtract (% of opponent points stopped) and the player’s total value as a combination of these two. I go through the starters first, from tallest to smallest, then the reserves in the same order.



No.NameHt.OffDefTotNameHt.OffDefTot
3Festus Ezeli ***6-f-113.60.23.8Davante Gardner6-f-83.51.24.7
5Lance Goulbourne6-f-81.01.92.9Jae Crowder6-f-66.23.29.4
44Jeffery Taylor6-f-74.01.05.0Vander Blue6-f-41.92.13.9
23John Jenkins6-f-45.20.05.2Darius J-O6-f-24.21.05.1
1Brad Tinsley6-f-32.30.02.3Junior Cadougan6-f-11.30.41.7
 Starters Total 16.13.219.2MU total 17.07.824.9
33Steve Tchiengang6-f-90.40.30.7Chris Otule **6-f-110.02.32.3
34Shelby Moates6-f-80.00.00.0Jamil Wilson6-f-71.20.61.8
45Rod Odom6-f-90.00.00.0Jamail Jones6-f-60.00.60.6
2Kedren Johnson6-f-40.00.00.0Juan Anderson6-f-60.00.00.0
24Dai-Jon Parker6-f-30.00.00.0Todd Mayo6-f-31.90.42.3
11Kyle Fuller6-f-10.00.10.1Derrick Wilson6-f-00.00.30.3
 Reserves Total 0.40.40.8MU total 3.14.17.2

*** Last year’s figures for Ezeli and projected totals for Otule over 12 games; some numbers appear 0.1 off due to rounding

In the preseason projections, Vanderbilt actually had the highest rated offense in the country with five returning starters from the 13th best offense in the country last year (116.5 points per 100 trips). Even though they haven’t quite lived up to expectations, if you add the 3.6% Ezeli was adding from last year, then the starting five are adding 16.1% this year (a 116.1 per 100 trips). Combined with the height advantage, the team’s excellent shooting (particularly on two-point shots), and the fact that this team took Xavier (before the meltdown) and Louisville (on the road) into overtime WITHOUT Ezeli, Vanderbilt does look like a big threat.

However, there is good news when you look at the rest of the table.
First of all, Marquette’s starting five actually has put up slightly better offensive numbers, even with Ezeli on the court, at 17.0% for their starting five. While John Jenkins has been one of the best offensive players in the country (5.2% Value Add), and Jeffrey Taylor has been almost as good (4.0%) they are slightly behind Marquette’s top two offensive players to date (Crowder 6.2% and DJO 4.2%). Ezeli’s was almost exactly as valuable last year as Gardner has been this year (3.6% to 3.5%). Neither Goulbourne or Blue have had a big impact offensively (1.0% to 1.9%), and Brad Tinsley has been only a little better than Cadougan (2.3% to 1.3%), though if Tinsley returns to last year’s offensive form he would give Vandy the overall edge offensively.

MU must leverage defense and bench to win

So if Tinsley heats up, perhaps Vandy has a better offense overall. However, when we look at the defense and the bench, Marquette has a huge edge that must be leveraged to pull out a win if it gets close (or preferably to give MU a little breathing room).

Marquette has had three fantastic defenders this year, though obviously we assume one of them will still be on the bench at least for a while. When Otule has been on the court he erases an extra 1.7% of opponents’ points that the average shotblocker as part of his overall 2.3% of opponents’ points erased. When Blue is on the court he erases 1.8% more of opponents’ points via the steal than the average ball stealer. And finally Crowder is among leaders in blocked shots (0.3% of points erased beyond average), steals ( 1.3% erased) and defensive rebounding (1.8% erased) to be the most effective overall defender with 3.2% of opponents erased in all of Marquette games (the first three totals are while he is on the court, while the cumulative is for the entire game, thus lower than the total).

As covered in the column on deriving defensive value this summer, there are limitations beyond that. Clearly Derrick Wilson is a much better defender than Davante Gardner because he forces many more non-steal turnovers and pressures opponents out of good spots, but since the NCAA doesn’t keep track of those two stats, we have to give all five players on the court equal 0.2 stops, or buckets allowed, when there is no steal or block. Therefore Davante’s statistical rating is a little better than Derrick’s just based on the fact that he actually is one of the top ball stealers in the country so far (2.9% of opponents’ trips Davante has stolen the ball compared to 2.1% for Derrick, even though his overall defense obviously is not.

But overall, you can see that every Marquette player off the bench adds value – Anderson just doesn’t have enough minutes yet to register – while Vanderbilt’s subs have produced virtually nothing to date. The defense has been much better throughout all 11 players since last year, and Todd Mayo and Jamil Wilson have produced a ton of additional offensive (1.9% and 1.2%) despite playing only 50% and 41% of the minutes respectively.

Certainly Vanderbilt comes in with a chance to show they are an elite team now that they are back at full strength with a huge road win in the Bradley Center. Unfortunately Marquette has little to gain and a lot to lose with a home loss, so they need to use superior defense and depth to hold serve and get a win.

This does not mean that Marquette necessarily needs a low-scoring game to win, only that the defense has to contain the explosive Vandy offense to create baskets at the other end.

As a reference for those interested who didn’t see the initial pieces, here are the links to how the Offensive Stat (Value Add) and Defensive Stat (Defense Sub) are calculated.

2 comments:

cato said...

Wow, yet another lengthy and detailed statistical analysis that turns out to be utterly innacurate when bumped up against the reality of, you know, PLAYING THE GAMES.

This Marquette grad will be at the 'nova game this Sunday not really caring to calculate Maalik Wayns' Offensive Value-Add differential vs. Junior Cadougan. More from Japan, please--I get enough of numbers doing my taxes.

Unknown said...

Another embarrassment on national TV. I would love to see statistical analysis of how poorly MU plays when on ESPN, ESPN 2, ESPNU, or CBS. I just don't get it.