Fran-Graphs, Indiana

Horace E Cow

Iowa's fantastic defense keeps them close to the #1 team, but turnovers and bad offense prevent them from getting over the hump.

Frangraphs_indiana_medium

[UPDATE BELOW -- I messed up on something further down in the piece and there are corrections at the relavent places.]

You could certainly find flaws in Iowa's play in this game, but when you consider the opponent, the location, and the fact that Iowa was missing their starting point guard, I thought they played pretty well. Consider the following:

  • Indiana has the best offense in the country according to Ken Pomeroy's adjusted offensive efficiency ratings at 1.25 points per possession; Iowa held them to 1.04 points per possession, and .79 points per possession in the first half.
  • Only two other teams have held Indiana to a lower offensive efficiency: Wisconsin and ... Iowa, the last time they played.
  • Iowa overcame a hideous first half on offense and put up 46 points on 37 possessions in the second half.
  • Iowa committed 18 turnovers, 27 fouls, made one three pointer, had 18 fewer free throw attempts, and still only lost by 13.
There are no moral victories, and losing this game probably eliminated the Hawks from the NCAA tournament (at least as an at-large team), but Iowa played very courageously, all the same. It would have been very easy to give up in the second half and concede the game, but Iowa fought to the bitter end and, at least on one end of the court, looked like Indiana's equal.

They certainly missed Mike Gesell, though. As the turnovers mounted in the first half, Iowa could have used the steady influence of another experienced ball-handler. Indiana really turned the intensity of their defense up to 11 to start the game, and the Hawks had a difficult time fighting through. Their strategy seemed to be to expend all their energy up front and build an insurmountable lead, and to their credit it worked. Their defense was noticeably more slack in the second half, but the 12-point lead they built in the first half held up and provided them almost exactly the final margin of the game.

Indiana's defense in the first half, though, was very impressive. In fact, it was a lot like a turbo-charged version of Iowa's defense. The Hoosiers pressured on the perimeter, trapped pick and rolls and dared Iowa to fight through the pressure. A few teams have managed to shred Iowa's harassing perimeter defense -- Penn State and Nebraska most recently -- and the way they did it was by having a big, strong point guard fight his way around or through the traps, thus exposing the tender underbelly of the defense. Iowa didn't really have a player with the combination of size, strength and handles to match up against the terror of an Oladipo-Zeller trap -- few teams do -- but the opportunities were there. Indiana was like a really athletic football team that blitzes on every down and dares you to throw on them. If you can get the throw off, there will be opportunities to score, but they're betting you'll never get the chance.

Indiana is a fascinating, weird team, especially for a top-rated squad. They have tremendous offensive talent -- I think unquestionably the best in the country -- with NBA-ready players in Oladipo and Zeller and quality shooters at every other position, but from a defensive perspective, they are eminently beatable. They are small. Not short, per se, although Hulls and Ferrell are both sub-six feet, but just not very strong. Adam Woodbury did a marvelous job defending Cody Zeller, and he did it by just standing in front of him and absorbing his ineffectual body blows (and if you want a preview of Cody Zeller in the NBA, keep that picture in mind). They are an F-1 car: lightweight, built for speed, but as flimsy as a balsa-wood plane. I wouldn't be surprised if they made a ton of threes and blitzed their way to an NCAA championship, but I also wouldn't be surprised to see them get bounced by a good rebounding team like St. Mary's or Colorado State. They have the feel of certain recent Duke teams: they take advantage of their tremendous home court to play aggressive, harassing (fouling) man defense, but as soon as they get to a neutral court, they look like a team with a lot of skill and only so-so athleticism. We shall see, but I'm skeptical of their chances to go all the way.

But back to Iowa. The Hawks showed once again that they are an elite defensive team, but they can't score to save their lives. It's been really startling to see a team that was so bad defensively last year transform into one that is so stingy this year. It would be easy to attribute the change solely to the addition of a certain 7'1" big man, but Woodbury is only averaging about 13 minutes a game in conference play. As cliché as it is to say, it's really been a team effort. Gesell and Clemmons are a huge upgrade at the point over Bryce Cartwright, Eric May and Melsahn Basabe are both healthy and energetic, Aaron White and Gabe Olaseni have both made big strides defensively, and Zach McCabe has played well, too.

To get an insight into just who is doing what for the Hawks, check out this website run by the stat guy Daniel M. He has calculated a stat called adjusted plus-minus for all of college basketball. Adjusted plus-minus is like reguar plus-minus (how much the lead goes up/down when a player is in the game), but it controls for the fact that the nine other players on the court differ from player to player and minute to minute. It can be a pretty noisy stat, even in the NBA where there are more games, but with that huge grain of salt, here's what APM says about how the various Hawkeye players impact the game:

[UPDATE/CLARIFICATION 9:23 pm

I misread what Daniel M. was doing and made a pretty big mistake. He's doing adjusted statistical plus-minus (ASPM), not adjusted plus-minus, and those are two very different kinds of stats. ASPM is basically a linear weights stat; that is, you multiply the various box score stats (points, rebounds, turnovers, etc.) by various weights, add them all together and get a big, comprehensive number. How do you get those weights? In this case, by running a regression on adjusted plus/minus (what I described above) for a large sample of players (in this case, the NBA). Adjusted plus-minus, although it has the simpler name, is actually much more complicated. You need detailed on-off lineup data for every game so that you can know who is on the court with a given player at every possible moment. What Daniel M. is doing is creating a sort of estimate or guess of what a player's APM will be just based on simple counting stats. In situations where detailed on-off data is unavailable or there's not enough overlap in the pool of opponents (as in college basketball), this is a handy, if uncertain, workaround.

My impression was that Daniel had done the Herculean task of collecting all that on-off data himself and creating a true APM for the NCAA. That would be an amazing achievement and well worth sharing. As it is, it's still very interesting, but is not all that different from some other linear weights stats you may have heard of (wins produced or wins shares). The important thing to remember about these kinds of stats is that they can only tell you about things that the box score measures. Things that are harder to measure, like good defense or setting quality screens, won't get picked up. Sorry for the error.]
  1. Aaron White: +7.37 (+4.92 offense, -2.46 defense)
  2. Eric May: +5.73 (+2.47 offense, -3.26 defense)
  3. Gabe Olaseni: +5.67 (+1.54 offense, -4.13 defense)
  4. Melsahn Basabe: +4.62 (+1.38 offense,-3.24 defense)
  5. Dev Marble: +4.59 (+3.41 offense, -1.18 defense)
  6. Zach McCabe: +2.82 (+.85 offense, -3.74 defense)
  7. Mike Gesell: +3.88 (+1.09 offense, -2.49 defense)
  8. Adam Woodbury: +1.81 (-.72 offense, -2.98 defense)
  9. Anthony Clemmons: +1.84 (+.34 offense, -1.61 defense)
  10. Josh Oglesby: +.75 (-.83 offense, -.43 defense)
  11. Pat Ingram: +.02 (-2.61 offense, -.46 defense)
Daniel M. doesn't list the error terms on these numbers, but they are likely pretty large given the sample is just 29 games. [this would be a problem with APM; with ASPM, it doesn't really apply]. With that said, the broad strokes picture it paints of the team seems about right. Aaron White is the best offensive player and the best player overall, Eric May and Melsahn Basabe have both been very good, Dev Marble has been a good offensive player and a so-so defensive player, Adam Woodbury has been a minus on offense. That's all expected. There are some surprising insights, too, though. For one, Gabe Olaseni looks like he makes a big impact in his limited minutes (although we have to be wary of any number that is derived from such a limited sample), Zach McCabe is better on defense than he is on offense, and Josh Oglesby is worse on offense than he is on defense. That last one is perhaps the most startling for me. Oglesby is on the roster purely for offense, and he shows up by these numbers as the worst offensive player in Iowa's rotation. That's... not good. When you look at his raw numbers, though, it makes sense. When a player shoots 26% from three and shoots almost entirely threes, he is not an efficient offensive player.

The Sloan Sports Analytics conference was this past weekend, and there was some interesting new research presented by Kirk Goldsberry and Eric Weiss that is sort of related to these APM numbers. In their paper "The Dwight Effect", (PDF here, summary here), they use spatial data from the SportVU system* to estimate the impact various players have on opposing field goal possession when the ball is in their area. Certain players, like Dwight Howard, Larry Sanders and Roy Hibbert, dramatically lower the percentage the opposition shoots when the ball is near them, while other players, like David Lee, make that percentage skyrocket. This all makes intuitive sense, but it helps explain how certain players have an impact that far exceeds their box score numbers. Gabe Olaseni is a perfect example. He averages just 2.8 points and 2.8 rebounds a game, but has a huge impact on the defense's performance while he's in the game. [ASPM can't really tell us this -- it relies solely on what's in the box score, not on the way the lead changes when a player is in the game.] Basabe and May are nearly as good on defense by the APM numbers. The story the numbers tell is sort of Iowa's season in a nutshell: lots of talented defensive big men + too few good shooters = a good defense and a bad offense.
X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Black Heart Gold Pants

You must be a member of Black Heart Gold Pants to participate.

We have our own Community Guidelines at Black Heart Gold Pants. You should read them.

Join Black Heart Gold Pants

You must be a member of Black Heart Gold Pants to participate.

We have our own Community Guidelines at Black Heart Gold Pants. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9347_tracker