Clemson: The Numbers Comment Count

Brian

op_has_a_posse_whitebg_small  clemson-basketball

"OP has a posse" from Clemson b-ball blog the OP.

What Say Tempo Free Stats?

Tempo free stats say we gon' die. Clemson's ACC performance on a possession-by-possession basis was better than anyone not named Duke or North Carolina:

                                     Opp.
                     Pace    PPP     PPP      EM   
1.  North Carolina   74.7    1.16    1.01   +0.15
2.  Duke             69.2    1.08    0.98   +0.10    
3.  Clemson          70.5    1.09    1.01   +0.08
4.  Wake Forest      74.3    1.07    1.01   +0.06
5.  Florida St.      69.1    1.00    0.98   +0.02

Meanwhile, Michigan was actually in the red in the Big Ten:

5.  Ohio St.         60.8    1.07    1.05   +0.02
6.  Minnesota        62.2    0.98    0.98    0.00
7.  Michigan         61.0    1.01    1.03   -0.02
8.  Penn St.         59.7    1.00    1.04   -0.04
9.  Northwestern     59.6    1.03    1.07   -0.04

If you hold that the ACC and Big Ten are basically equivalent this year—generous—this bodes very unwell. I guess there's an argument that Clemson got way up there by whipping some teams badly and their week-in week-out performance is too erratic to take that number seriously, but, man, that's a big gap.

Kenpom also says this is a crappy draw for Michigan. It has Clemson the #22 team in all the land, which corresponds to a solid six seed in Tempo Free Ability Land, not the seven they got in reality. (This is not an attempt to imply the committee was wrong to make Clemson a 7; sometimes teams under- or over-perform their overall ability in the actual wins and losses. For example: Michigan, as you'll see.)

Michigan, meanwhile, is #49. If we took the top 65 teams by their ratings and broke them into seeds, this matchup would be a 6 versus a 13. Kenpom says we have a 32% chance of victory. This is considerably lower than the other ten seeds: Maryland has a 38% shot at Cal, Minnesota a 42% shot at Texas, and USC is actually a tempo-free favorite (65%) over BC.

Of course, these are just simple averages that take every possession in every game, adjust them for difficulty, and project future games based on that and a home/road adjustment. They don't take matchups into account.

So: matchups.

Clemson Defense vs Michigan Offense

Turnovers. Clemson's major distinguishing characteristic is a 40-minutes-of-hell full press that sees Clemson force a butt-ton of turnovers: 24.1 percent of opponent possessions end without a shot. That's 16th nationally. You can see the costs of this strategy in a couple spots: a generous 2PT FG percentage allowed (46.1%; 89th) and horrible defensive rebounding.

Three-pointers. Bad news for Michigan: Clemson does a good job of protecting the three-point line. Only 28% of opponent's shots are threes, which is 26th nationally. They do allow a good percentage when opponents get a look.

Inside the arc. As mentioned, Clemson's two-point percentage allowed isn't good despite (yet another) huge block percentage: 14%, 14th nationally. They offset that with a healthy rate of fouls.

Specific People Who Are A Specific Height. The good news for Michigan is that Clemson's main lineup isn't Illinois huge or anything: the starting lineup has a 6'7" guy and a 6'9" guy but no 7-foot menace, despite the shotblocking prowess. Sims defied earlier predictions here about an inability to combat seriously large posts when he tore up Minnesota, but he tore up Iowa even worse and was then reduced to a wide array of outside shots against very large Illinois, none of which went down. I'm still much more comfortable when Sims is dealing with an opponent of approximately the same size.

The big question. Grady? Kelvin Grady, once Michigan's starting point guard, struggled badly in Michigan's ugly stretch midway through the season and got benched in favor of CJ Lee and David Merritt. Grady's limited but has one grade-A skill: handle. Attempting to press him is a futile waste of time. Merritt and Lee… eh, not so much. Beilein's not likely to start Grady or anything, but if Clemson forces a couple of ugly turnovers early he might get exhumed.

Michigan Defense vs Clemson Offense

How much of the Tiger offense is turnover based?

Opponent turnovers don't translate directly into offense unless it's a steal, which can lead to a fast break opportunity. This took forever and came out with bupkis, but it took forever so look at the graph you bastards:

scatterplotimage

That's a scatter plot of Clemson's steals against their points per possession. Correlation: eh, eyeballing it… just about zero. Where the pressing will tell is in Michigan's offense, it appears.

Their turnovers. Eh… they're mediocre, turning the ball over on exactly 20 percent of their possessions. That's about the national average. Michigan forces opponents into 21% turnovers.

terrence-oglsebyThree-pointers. Clemson's Terrence Oglesby (the near-albino guy at right) is an excellent shooter hovering around 40% on his long range attempts; this is a major component of Clemson's excellent three-point percentage: 38%, which is 35th nationally. They take a slightly above-average number of them.

Michigan's been surprisingly effective at shutting down the three point line, though: 43rd in opponent makes and opponents get off a below-average number of triples. That might have something to do with…

Two pointers. Michigan is horrible at defending them, allowing opponents to shoot over 50%. Clemson is very good at getting them, making 51.4% of their attempts. Clemson's offense is mostly just good all around, with a high eFG and a very high offensive rebound percentage offsetting a crappy ability to get to the line and meh turnovers.

Miscellaneous Factors

Past performance, future results. I kind of hate the "X-factor" as a concept/cliche, but there's an X-factor here in how the teams react to styles they haven't seen before. Michigan hasn't been pressed consistently all year; Clemson hasn't seen the Beilein offense or the 1-3-1. A lot of people are banking on Beilein's tendency to overperform in the NCAA tournament to see Michigan through, and that's usually attributed to his unusual style.

beilein-ncaa

via bracketscience and mvictors 

Here's some Debbie Downer: it could just be chance. Beilein's only made it a few times and has done well, but he's pulled a couple fortunate matchups, with Northwestern State blowing up Iowa before the Mountaineers had a chance; WVU also missed the three-seed in '05 and had an opportunity against an 11 in '98.

The other side of the coin: Clemson coach Oliver Purnell is 0-fer in five trips to the tourney:

  • 1992: 15 seed Old Dominion, a 15-15 team is clunked by #2 Kentucky.
  • 2000: 11 seed Dayton loses to 6 seed Purdue by one point.
  • 2003: 4 seed Dayton loses to 13 seed Tulsa by 13.
  • 2008: 5 seed Clemson loses to 12 seed Villanova by 6.

The 1992 and 2000 trips don't raise eyebrows, but Purnell's suffered two straight upsets at the hands of inferior competition.

There's a case the coaches' historical performance is more than a fluke.

Sliding. Though Michigan hasn't exactly burned up the nets of late, they are 3-2 in their last five with wins over Purdue and Minnesota. Clemson, meanwhile, has been in free fall: 1-4 to finish the season with the win over horrible Virginia and the final loss an opening-round ACC tournament matchup against last place Georgia Tech.

Common opponents. These do not bode well, but mostly because they encompass Clemson's best performances of the year:

  • Duke: Michigan split a neutral/home pair; Clemson obliterated the Blue Devils by 27 at home.
  • Maryland: Michigan lost a road game; Clemson obliterated the Terrapins by 29 at home.
  • Illinois: Michigan went 1-2, with a ten point win at home and ten point losses on the road and in Indy. Clemson scraped a two point road win over the Illini.
  • Savannah State: Clemson clubbed them; Michigan won in overtime, barely escaping a tourney-killing loss.

Well?

This will be a big test of the Beilein-as-impossible-to-prepare-for meme, because Michigan drew a tough, tough seven seed. Fellow 10 seed USC is staring at an ACC team that managed to lose to Harvard (which who does that, really) and finished 7th in conference when it comes to efficiency margin. In contrast, Michigan draws a Clemson team that, while sliding a bit, beat some good teams absolutely raw and is a solid third in conference, above 4 seed Wake Forest.

Meanwhile, Michigan is the team that finished 7th in efficiency margin in its conference, and you can tack on some uncomfortably close victories in the nonconference to that. I think Kenpom is about right here: Michigan is a decided underdog.

Elsewhere:

Comments

Erik_in_Dayton

March 18th, 2009 at 12:40 PM ^

Michigan can put a lot of decent ball handlers on the court at once. They usually have four guards plus Sims on the floor. They share the ball well and know where to find each other...This leads me to think that they'll break Clemson's press often and get a bunch of open shots...More, Clemson's offensive numbers as compared to Michigan's are overrated, as an ACC team does not go up against the defense (and officiating) you do in the Big Ten.

nymumich

March 18th, 2009 at 1:17 PM ^

Why does every single one of your basketball arguments revolve around kenpom and goofy charts and graphs? And beyond that, why should anyone take your b-ball comments (which have always, always been imbued with negative undertones this year) seriously? You try to argue both sides, but everything you write about this team this year has tilted toward pessimism and doubt. Beilien's teams' performances have probably been more luck than anything, right? Maybe his teams' numbers never positively correlate with kenpom's numbers, but the man knows how to prepare his teams in crunch time, and that is undisputed. Let's look at the real factors in this game. Clemson has been a mess the last few weeks, lacking any consistency and fighting among each other. Terrible sign. Purnell is 0-4 in the tournament, and is notorious as a coach for watching his team play sub-par in the 2nd halves of seasons. And the team lacks height, which you point out. I'm not saying we win, but I'm going to use something other than kenpom to make a judgment (and this isn't bias), and say we are going to be in this game without a doubt and could very well win. But hey, keep up your negativity, it worked well for us this year.

Chrisgocomment

March 18th, 2009 at 1:29 PM ^

Statistics tell us something about a team that generalized observations do not. What they tell us is that this team is a bubble team. Which they were. Face it, they aren't very good, they barely got into the Tourney and are lucky to be there. Am I happy they are in the Tourney? Hell yes, but logically the odds of them advancing aren't very good, just like the odds of them making the final 65 weren't very good to start with. Brian is giving you his opinion based on empirical facts. That's more than I can say for you and your "real factors".

dex

March 18th, 2009 at 1:22 PM ^

BOOOOOOOOOOOOOOOOOOOOOOO to "numbers" that "represent" the "number of times things happen" and show "how often teams do something" and "track performance" in "basketball"!!!!!!!!!!!! Especially in a post titled "NUMBERS"!!!! BOOOOO I SAY!

Jeff

March 18th, 2009 at 3:45 PM ^

CBS also says we are an "Automatic bid. (Southland Conference tournament champions)." That's nice that we've got that going for us. It's been 11 years since we won a conference tournament, about damn time. Aside from that mistake with cut & paste that is a pretty good writeup.

TomW09

March 18th, 2009 at 8:20 PM ^

Wow, that's one of the worst pre-game articles I've ever read. Terrible. It's like someone looked at box scores and said "This Manny Harris fellow puts up huge numbers, clearly if he does well Michigan wins and if he doesn't they'll lose. In a related note, Michigan sure did dominate that Southland Conference".

AC1997

March 18th, 2009 at 1:58 PM ^

Statistics can tell you a lot and are a powerful predictive tool. In baseball they can tell you just about everything. In basketball they are becoming more powerful, but they're not perfect yet. However you look at it, Michigan is a 10 seed. The likelihood of them advancing even one game is not high and more than that is almost zero. You don't need Brian's stats to show that. And at the end Brian does throw some X-factor things out there that aren't accurately depicted by stats. Here are some other thoughts that might work in Michigan's favor: -- Time to prepare. I'm surprised this didn't come up in the piece, but Michigan has been VERY good this year with 4+ days to prepare/rest. They've been significantly worse with less time to prepare. That could be fatigue, coaching, preparation, or scouting. But it has happened enough to suggest a trend. -- Officiating. The three people with the biggest effect on the outcome of a game don't play for either team. Remember how Manny dominated the non-conference season with his ability to get to the free throw line? If the officials give him the Dwayne Wade treatment in this game, he could win it by himself. If Hightower or Valentine are officiating, he could be sitting on the bench with four charging calls in the first two minutes. -- Expectations. Who knows if this plays a positive or negative role, but Michigan enters this game knowing their season is a success win or lose and will be out there having fun. Clemson has been on a slide and went from having national championship expectations to fighting to win games. They think they need this game to validate their season perhaps. -- Experience. We don't have any, they do. Probably not good for us, though their experience has been negative with their early exits. I think you'll see a very good game on Thursday that Michigan has a chance to win. I'm not sure they're good enough in this match-up to actually win, but I think they have a shot. I would give them no shot against Oklahoma though.

Michigan Arrogance

March 18th, 2009 at 2:33 PM ^

That's a scatter plot of Clemson's steals against their points per possession. Correlation: eh, eyeballing it… just about zero. ehhh... just about? that's as close to a non-correlation as you're gonna get. statistically, id guess it is zero.

jamiemac

March 18th, 2009 at 2:54 PM ^

Who cares if WVA ducked the 3-seed in the 2005 regional semifinals? They took out a top-10 ranked Wake team in the second round. Chris Paul. Sleeper final four team, by most accounts, that season. Tore them apart. Big time. And, he wins as a 14 seed at Richmond, so how can that be downgraded because he lost to an 11-seed in the second round?

jmblue

March 18th, 2009 at 3:42 PM ^

Adding to the Beilein-is-hard-to-prepare for argument: he's also 13-6 in the NIT, including the title in 2007. In addition, he's done pretty well in conference tournaments.

nymumich

March 18th, 2009 at 4:46 PM ^

I'm not saying statistics don't play a role, my point is that sometimes people put too much emphasis in a bunch of numbers that a computer cranks out. And beyond that, why does everything always come back to kenpom? Let's look at some of the teams that kenpom ranks in the top 40. Georgetown, Wisconsin, San Diego St., Notre Dame, Miami (Fl.), New Mexico, Washington St.? Riiiight. If you're going to draw an argument from any rankings, please diversify once and a while. There are plenty of sites and blogs out there with their own viewpoints and logarithms blah blah blah. Throw enough shit together and you might come up with something fair and coherent. And sorry I didn't have time to write you a blog post with all my factors, big guy, considering I'm at work and all that. In any case, I've been watching basketball long enough (and I'm sure you have) to realize that in the NCAA tournament especially, numbers only go so far, and emotion, momentum, and coaching are of major import. Breaking down a matchup by numbers would essentially call for a mostly chalk bracket, and we sure as hell know that isn't going to happen this year, or any year for that matter. I'll take a Beilein team in March (which has CONSISTENTLY reflected a statsitcal outlier through all of its success) against a team that is a ticking time bomb with a coach that is 0-4 in the tournmament.

dex

March 18th, 2009 at 4:52 PM ^

you aren't "wrong", but you are making this argument on a post that is specifically about the numbers involved. i don't usually go read a statistical analysis of something and then get pissed when it contains statistical analysis. just like if i read a mitch albom column and founda bunch of hocus pocus speculation about team juju and such i wouldn't be angry that it didn't contain statistical analysis.

nymumich

March 18th, 2009 at 5:26 PM ^

Hey Dex, it's not about being pissed because it's a statistical analysis. Leaving the substantive stuff aside, almost every mgoblog post re: basketball has been "yea, but," or "we're really not thattt good." Except, of course, for the post-win revelry. For a program that's been getting shit on nationally and locally, it'd be nice to be able to read more positive things without the overaching tone of pessimism. Maybe it's more subtle in this post - but it's there - and has been all too much in basketball discussions. Hate to tell you, but realism and wet-blanket are not synonyms. Either way, go blue.

bronxblue

March 18th, 2009 at 8:37 PM ^

While I do agree with Brian's general sentiment that UM is clearly the underdog in this match-up and Clemson is a decent team, I do agree with the sentiment that the numbers may be skewed a bit in terms of Clemson's "advantages". For starters, the team only beat one ranked team (at the time of the game) all season (given it was Duke, and by 27). Of their 9 wins in conference, the combined record of those teams was 144-107, and 55-73 in conference. And those numbers are a bit skewed, since Duke was 28-6 and 11-5 in conference. In fact, the only team they beat with a .500 or better in-conference record besides Duke was BC at 9-7. Compare that to UM, which beat 3 ranked teams (at the time), both in conference and outside. Neither team was great on the road in conference (Clemson was 4-4, UM was 3-6), but UM actually seemed to play pretty well on the road outside of conference (win over UCLA on a neutral court, close losses to UConn, Maryland, and Duke on the road/neutral), while Clemson's 2-point will in Champagne was the highlight of the OCC road schedule. So I guess my point is that while I agree with Brian's general belief that UM is the clear underdog in this game, the statistics he cited painted a more dire picture than is perhaps the reality. Will UM have to work hard for a win, and is it more than likely that UM is 1-and-done? Sure. But this Clemson team is not the same one that destroyed Duke and Maryland when it is not playing at home. On a neutral court, it is an average team that has a history of stumbling toward the end of seasons.

NJWolverine

March 19th, 2009 at 10:20 AM ^

Clemson is very simple. If they can't break you with the press, they are an extremely vulnerable team because they lack size. It's not altogether different from Michigan when the threes don't go down. But the key to the press is hustle. They have to work hard to create turnovers and rattle the other team. Otherwise, they can lose to anyone. I don't think they can rattle a well coached, disciplined Beilein team. While we don't have great ballhandlers, I'm sure there are set pieces to not only get the ball past halfcourt, but also to open up a three point shooter. Whether or not those threes go down will go a long way towards determining the outcome of the game. I also think the 1-3-1 favors us. Clemson is a free flowing team. It's almost like the controlled chaos theory where they like to run up and down but they do so out of pressure they create in the defense. Problem is, those driving lanes will be closed by the 1=3-1 and they don't have a big man. If Michigan can simply break their press, it will throw them off of everything they do. Teams most vulnerable to the press are undisciplined, young teams who needlessly turn the ball over a lot. That's certainly not our team. Finally, did you read that piece about how the team went to the Cheesecake Factory last night just to have a good time. That's a factor too. Michigan has NOTHING TO LOSE in this game. Nothing. They can lose by 30 and still the season will be a resounding success. That will allow them to play loose, which will unleash all of the disciplined Beilein principals that break the trap. Meanwhile, Clemson is fighting deamons. They once again faded down the stretch. They have to play with great intensity, but I don't think that'll matter because our guys won't be rattled. Clemson has to rattle you in order to win. That's where all their numbers come from. I don't think Michigan will be rattled.