Thursday, February 14, 2013

Boilermetrics: Centers of Attention

There haven't been a lot of good things to say about this season, as you may have noticed recently, but there are a couple of guys who are having noteworthy seasons. In particular, a young man from Carmel has made Boiler opponents sit up and take notice.

I don't think so.
(Photo by Al Goldis/AP)
It's possible that the one thing separating the Boilers from having a Penn State-caliber season is A.J. Hammons. Despite the fact that Purdue can go several games without being a threat from the perimeter, Hammons has been a solid inside presence in nearly every game. He's scored in double figures in all but three conference games, rebound rates at both ends of the floor that rank him nationally, and an outstanding block percentage: 8.3%, 63rd in the country, and the majority of the guys in front of him are either eating up weak conferences or feasting on weaker teams. (Someone explain to me how a guy with the name Shaquille Duncan isn't leading the country?) He's drawing fouls (4.8 per 40, basically 3 fouls drawn per game given his minutes), hitting free throws, and not committing many fouls: only 3.2 per 40. (And that's playing in a conference where even at this moment, Ed Hightower is calling a foul on someone.)

As tempting as it is to project him as the next Lew Alcindor, it would be unfair and more than a little silly. (How many people even remember Kareem's old name anyway?) But we can take a look at other Purdue centers who had quite an impact in their day, and maybe look a little more closely at what Hammons might do if he stays. (He's not listed below, but I do want to mention Terry Dischinger, a two-time consensus All-American who played against guys like Jerry Lucas and John Havlicek. I'm also leaving out JJ because as good as he was when he played center, he never really had the bulk for the position. If Purdue had had a real center alongside him, I think Johnson would have been almost unstoppable. Finally, Brad Miller did a lot of things well, but wasn't outstanding at any one thing: I'm looking for exceptional more than simply effective here.)

"Do that dunk thing again.
That worked pretty well last time."
Joe Barry Carroll came off the bench as a freshman: Fred Schaus chose to start co-captain Tom Scheffler (Steve's older brother) ahead of the new guy. (After the 1977-78 team went 16-11 despite having Carroll and Jerry Sichting, Schaus was replaced by Lee Rose. Coincidence?) Carroll played just 20 minutes a game, but still averaged 7.4 rebounds and nearly 3 blocks per game: his 206 rebounds and his block average are still freshman records, his block average is second overall, and his block total (82) is third.

Starting full-time as a sophomore, Carroll still got only 31.7 minutes ... but posted 105 blocks for a whopping 3.9 blocks per game, both school records that stand today.

Out went Schaus (and Walter Jordan departed for the NBA); in came Lee Rose. Carroll got 35 minutes a game, and was a dominant force on the team that lost the NIT final to Indiana. (In these days, the NCAA tournament was significantly smaller, and people still cared about the NIT.) Carroll's 352 rebounds are a school record, and his 70 blocks (7th) and 798 points (3rd) are still in Purdue's all-time top 10. Carroll posted a career-high 42 points against Alabama in the NIT semifinals; he would not top that his senior year, but would play on a slightly bigger stage, slightly closer to home.

His senior-year numbers tailed off slightly: 736 points (still over 22 per game) and 302 rebounds (tied for 7th). He did block another 92 shots, second on the list, finishing his career as Purdue's all-time leader in blocks (349; only JJ is within 100) and rebounds (1148; only Dischinger is within 200, although Terry played just three seasons).

Carroll was instrumental in Purdue's Final Four run that season, the last time they made it to that level: 33 points against La Salle in the first round (at West Lafayette, no less), and 36 in an upset of 3rd-seed St. John's. He was held to 11 points, but the Boilers knocked off 2nd-seed Indiana in Lexington, 76-69, as sweet revenge for the NIT loss, despite 30 from Isiah Thomas. Carroll then poured in 26 over Mike Gminski as Duke fell in the regional final.

In Market Square Arena, the Boilers' quest for a first title would end at the hands of Cheatin' Larry Brown, as ineligible Kiki Vandeweghe scored 24 and held Carroll to 17 in a 67-62 loss. Carroll would finish his career in style in front of the semi-home crowd, though, with 35 points, 12 rebounds, and 4 blocks in a 75-58 win over Iowa for third place. (They don't do consolation games any more.) Carroll's 35 is the fifth-best total for a Boiler in NCAA tournament action; the 12 boards are tied for ninth-best, and the 4 blocks are tied for second. He finished his career as a consensus All-American; whether or not his reputation as an NBA player is deserved is a topic for another time, but in college, there was no question as to his greatness.

Purdue fans despairing after Carroll's departure had to wait exactly one game for the next force in the post. The Boilers opened the 1980-81 season against Colorado State with freshman Russell Cross at center; Cross would miss exactly one game in three years at Purdue and would come off the bench in only three others. (The man who made this wise decision was this Keady guy ... turns out he was pretty good at coaching.) Cross averaged at least 35 minutes in all three of his seasons at Purdue – he left after his junior year, and considering that he was the #6 pick overall, I'd say it was a wise move on his part. (Cross actually was Carroll's teammate at Golden State in 1983-84.)

This is not the 1974 trophy.
Do you really want to see it?
No, you don't.
With Carroll's departure, there were points to be had, and Cross scored them: his 16.9 average and 540 total points are still freshman records ('Twaun finished with 437, but just 12.9 per game; Walter Jordan's 14.1 is second in average). Cross blocked 60 shots, second to Carroll's freshman-record 82 and tenth all-time, and pulled down 202 rebounds, third behind Carroll and Jordan's 205. For a third straight season, Purdue made a Final Four ... but for the second time in three seasons, it was the New York version. Cross' team fell to Syracuse in the semis and defeated West Virginia in OT to win the third-place game.

Cross' production tailed off somewhat during his sophomore year, as he was unable to post any numbers that remain on the top-10 lists in West Lafayette. His teammates struggled as well, finishing with just a 14-13 regular-season record, but that and their previous tournament success was enough to earn another NIT bid ... which became yet another Final Four appearance. Again, the Boilers failed to take the title, losing to Bradley in the NIT final.

In 1982-83, there would be no NIT appearance. Cross scored 514 points, blocked 68 shots (tied for 8th), and pulled down 216 rebounds to get Purdue back to the tournament they really wanted to play in. It wasn't much different than what most Purdue fans would have guessed, though. In Tampa, Cross was held to 14 points as plucky Robert Morris nearly pulled off the upset, losing to 5th-seed Purdue 55-53. Cross scored 22 points and pulled down 9 rebounds, but the combination of Alvin Robertson and Darrell Walker was just too much, as 6th-ranked Arkansas knocked out Purdue 78-68. The game would be Cross' last with his departure for the NBA; he finished 3rd in blocks with 175, 20th in points with 1529, and 19th in rebounds with 622.

A.J. Hammons is the man under the microscope now. We'll assume that Purdue has just seven games left and that Hammons' production will be roughly the same the rest of the way. This is almost certainly not going to be the case – witness the way Illinois held him to 10 points – but you've got to start somewhere.

Hammons is averaging 6.7 boards per game, with 168 through Purdue's first 25 games. That would give him about 215 for the season, which would be a Purdue freshman record; the 6.7 would be third behind Carroll and Jordan. Carroll's 7.4 rebounds is within reach, but Hammons would have to average nearly 10 boards a game to catch him. (Fortunately, Purdue's shooting will give him lots of opportunities for offensive rebounds.) A tournament matchup with Northwestern would be nice: they are 11th in conference play in OReb% and dead last, with no one even close to them, in defensive OReb% (41.0%; 2 of 5 opportunities become offensive rebounds).

With 52 blocks, Hammons is already third on the list. He'll finish with about 66, which would be a distant second to Carroll's 82. Hammons would have to average over 4 blocks per game to catch Carroll, basically double his current rate.

For our final shot at the crystal ball, we'll take a look at Similarity Scores. Short version: looking at advanced metrics for other players with the same eligibility and calculating a number between 0 and 1000, with 0 meaning you're looking at a cat and 1000 meaning you're looking at the exact same player, fool. As Pomeroy says in the linked post, "anything over 900 is a great match[, and a]nything above 850 is still useful."

Hammons doesn't have any matches above 900, but he does have three guys at or above 890. (WARNING: wild speculation beyond this point. Two cars heading in the same direction at similar speeds do not necessarily have destinations anywhere close to each other.)

  • Melsahn Basabe, Iowa, 893. You probably already know this guy: he had 12 points and 8 rebounds in the loss to the Boilers this year. This wouldn't be a bad outcome for Hammons, in that Basabe has three seasons of solid rebound and block numbers ... except that Basabe actually dropped off a bit from his freshman numbers. He's also getting significantly fewer minutes now, which means either a) he doesn't fit McCaffery's style well or b) he's doing (or not doing) behind-the-scenes stuff that's costing him playing time. Either way, that won't be a problem here; Hammons would likely come out better if that's his comparison point.
  • Omar Samhan, St. Mary's, 890. You probably know this guy if you're a big NCAA fan or if you watch the tournament. He's the guy who got the Gaels into the Sweet 16 his senior year, knocking off Richmond and Villanova before being crushed by Baylor. That St. Mary's team was 20th in offensive efficiency, 7th in eFG%, 6th in FT% ... yes, please. Samhan had an ORtg of 116.6 despite getting 32.5% of the Gaels' shots. He rebounded (13.7 OReb%, 24.9 DReb%), blocked (8.0 Blk%), drew fouls (6.2/40), and hit free throws pretty well (.727). The downside to this projection is that Samhan had a similar junior year, and of course was playing in the West Coast Conference. If Hammons improves on the same track, his 2014-15 season will not be in Old Gold and Black; it'll probably be his second NBA season.
  • Alec Brown, Wisconsin-Green Bay, 890. Who? A guy from that school that indirectly brought us Ryanball (ugh). Brown's freshman year saw him post an ORtg better than Hammons', along with a 9.6(!) Blk%. He's currently a junior for the Phoenix, and he's had a bit of a dropoff this season. Again, that won't be a problem, because two solid seasons from Hammons will likely be all that he plays in West Lafayette.
If you could do this, would you stay?
Liar, liar, pants ... on fire.
So ... if Hammons develops like Samhan, he'll have two seasons to make a significant dent in the rebounding and blocked-shot record books, and then some lucky team will get a strong post player with a bit of developing to do. (Or Purdue will be one player away from having a shot at a long NCAA run, and Hammons will decide to stay, because there are few things better than being an outstanding college basketball player in Indiana.) Even if he has a Brown-like career, he'll have had more of an impact inside than a lot of players who preceded him.

I'll be watching the rest of the season from more of a distance, but I'll be paying attention to Hammons. Records are made to be broken, you know ...


Andrew Johnson said...

@zlionsfan - LOVED THE ARTICLE. Have really been wanting to learn more about the advanced stats behind basketball and with the way this season has gone I think it would be a welcome distraction to break down games with this type of material. Learning the Sabrmetric principles in baseball really opened up a whole other perspective for me and I think it could do the same for a lot of Boiler basketball fans. Perfect example was last night when Tim Doyle says that Byrd is the best shooter in the conference (kudos to J Money for calling out this BS in his game wrap up). Saw a few minutes later on Twitter that he doesn't even rank in the top 15 percentage wise. This type of 'analysis' being spouted out on a regular basis from the major media outlets of BTN and EsPN it can make you want to pull your hair out. If there are any advanced stats articles or sites you could recommend as introduction to the subject I would love to do some more reading. I realize this is the Boilermaker Enginerd coming out in me but I can’t help it.

Ever Faithful, Ever True!

zlionsfan said...

Thanks, Andrew! The CS major in me would have loved to see this kind of stuff 25 years ago. (Of course it would have been in USENET groups and such, seeing as how there wasn't exactly an internet or external email ...)

Starting points: has an overall ratings page and a blog that explains the theory behind a lot of what the main guy, Ken Pomeroy, does. For $20 a year, you get detailed stats at the conference, team, and player level, going back to 2002-03. I subscribe because I love this stuff; you might want to read some of his posts before you commit money to it.

It's an NBA site, but the principles are similar: has a lay guide to advanced stats. has good data on NBA players, and their college site has similar data on NCAA players (although not nearly as far back). Basically, I'd recommend all of the sports-reference sites if you're into this kind of thing.

Finally, has some interesting data from play-by-play information. Suggestion: do not look at Purdue's shooting percentage on two-point jump shots.

Just remember that, like with baseball, this is just another perspective ... and like football and hockey, there's a lot of team play in basketball that can't really be quantified properly.

Plang said...

Nice write up. I don't know enough about the kid to assume he will stay at Purdue all 4 years, but I hope he does.

Also - the Hightower remark made me do a spit-take.

Andrew Johnson said...

@zlionsfan - the knickerblogger link was exactly what I had been looking for. I had looked at the kenpom's site but never really got much out if since I ran into the paywall frequently. Thanks for the references, hope to see some more of your articles on the site in the future! Not that the regulars don't do a bang up job I am kidding obviously.

The narrative for this years team that pops up everywhere is that their lack of chemistry, lack of effort, lack of leadership has had a large part in their demise. The effort can be seen plain as day at times but I was curious just to see what type of additional insight some of these stats might have in pointing out the weaknesses of the team. Unfortunately I guess the question this year is not just what the weaknesses are since its pretty much.....everything, more like how bad is their bad compared to the rest of the league or all of D1 CBB.

Thanks again.

zlionsfan said...

Thanks, Plang. And you know we treat long-time Big Ten refs with all the respect they deserve here.

Andrew, yeah, the offense is pretty much bad in all respect except offensive rebounding. 37.4% is 33rd in the country, and in-conference 35.4% is 3rd in the Big Ten.

Defensively, the only significant problem is that they're not forcing turnovers: 17.3 TO% (308th) and 7.1 Steal% (329th). The only guard who has any kind of steal percentage is Anthrop, and that's in limited minutes even when he was healthy. (They're allowing more offensive rebounds in conference play, but that has a lot to do with caliber of opponent.)

The shooting itself is bad compared to all DI teams ... like, bad. From 3, 30.6% is 294th. From 2, 44.8% is 280th. From the line, 63.3% is 320th. There are 347 teams in DI this year. (Amazingly, Purdue isn't even last in the conference. Wisconsin, thanks to Ryan Evans-Prime, is shooting 61.8% at the line.) Offensive efficiency is above average, though, primarily due to the offensive rebounds and a lower turnover rate (19.1%, 117th overall).

That's about as far as the stats go ... from there, they really work better as supplements to what you see on the court. For example, the lack of chemistry and effort: on offense, you can see a lack of spacing, people not moving without the ball, poor communication, that kind of thing ... and a lack of leadership when these things happen over and over again and no one is getting the other players together and getting them to see what's going on. Then you look at the stats and see yeah, they're not getting a lot of assists ... shouldn't a motion offense be generating assists?