In the interest of burning the ridiculous amount of time between now and tipoff of the play-in games tomorrow, I thought that it might be fun to put together a post on the disparities between where certain teams are ranked in terms of efficiency vs. where they were seeded. I’m not really interested in the cases that the committee got "right" (who the hell is?), I’m more interested in panning the picks that they got wrong. Most of you have already probably pegged me as a huge fan of advanced stats, primarily because they provide some common ground to allow us to compare teams that are composed of completely different players, play in different conferences, and rarely have enough common opponents to allow us to draw salient conclusions from those results. I’m also a big fan of Ken Pomeroy, so I’ll be using comparing NCAA tournament seeds to his rankings to ferret out any discrepancies.
A few caveats before I get started. One of these is that, obviously, ranking teams from 1-347 is going to be an inexact science. I’ve tried to limit myself to teams that are seeded at least 3 lines lower/higher than they should be. I did this by dividing each team’s Kenpom rank by 4 and comparing that to the seed that they were given by the committee.
Secondly, after three or four years of poring over Kenpom’s ratings, I’ve come to the conclusion that he overrates turnover and rebounding rates at the extreme ends of the spectrum. For example, teams that turn the other team over a lot (VCU) have overrated defenses, whereas teams that never turn it over on offense have overrated offenses (Wisconsin). Obviously turnovers and rebounds matter, but I think that they way he weights turnover percentage and rebounding rate in his model leads to too much weight being given to rates of teams at the very high and very low ends of the spectrum. This comes into play for a few of the teams below, so I thought I’d bring it up now and just reference it later.
That’s enough introduction, on to the comparisons!
Florida (KP Rank: 1; Seed: 3)
This is the one that everyone’s been talking about, and not just after Selection Sunday. Florida’s sparkling computer profile has been the topic of much discussion, and is a result of a lot of things – high offensive shooting percentages and a low turnover rate, combined with a very good field goal percentage defense that still manages to turn opposing offenses over at an appreciably high rate (not easy to do). That’s the nice way of putting it. The other way of putting it is that Florida played a terrible non-conference schedule (sans 3 teams) and played in undoubtedly the worst major conference in the country this year. They took full advantage, completely blowing out any team that lacked the guard play to keep up with them. That led to a great efficiency profile, but calls into question its validity. There is a great debate between statisticians and traditional basketball fans as to what you can learn about a team when they play in "crunch time," and count me on the side of the traditional basketball fans in this case. There is something to the fact that every time Florida has played a good team in a close game, they have lost. I just don’t think they have a true point guard on the roster, and as a result they struggle in half court settings at the end of games. Additionally, they shoot more 3’s proportionally as a part of their offense than all but two teams in the tournament (Illinois and Iowa State), so they are going to be vulnerable to cold shooting nights. If they have one of those against a team that is good at rebounding, it might be hard for them to advance. Minnesota, anyone?
Pittsburgh (KP Rank: 7, Seed: 8)
This is another team whose computer profile is relatively great, but whose on-court performances and relative hype haven’t seemed to match up with that lofty rating. I think the problem with Pitt is partly one of timing – other than during their terrible non-conference schedule, they haven’t been able to win more than 4 games in a row all season, and therefore have never built up any hype. Toss in a loss to a terrible Rutgers team, and a home loss to Notre Dame in which they scored 42 points (42 points. At home. Against a team ranked 90th in defensive efficiency. This bears repeating), and you can start to see why you haven’t heard much about Pitt. That said, this is an extremely balanced team, and one that is more than capable of winning 2-3 games in the tournament. However, they really only have one quality ball-handler, so they could struggle against a deep team that can pressure. Still, they actually have a very good matchup against Gonzaga, who has guard issues of their own.
Wisconsin (KP Rank: 9, Seed: 5)
The poster child for the B1G’s style of play, Wisconsin is perennially overrated by Kenpom because of one of the factors I mentioned earlier. KP says that Wiscy’s offense is 52nd in the country, but take out the fact that they don’t turn it over because they don’t take any chances (15.6% TO rate, 5th best in the country) and the rest of their offense is terrible. They are like Oklahoma State without any of the ability to draw fouls (Wiscy is 316th in the country in FT rate, which tells you all you need to know about how they play on offense and what constitutes a "foul" in the B1G). Their defense, on the other hand, is truly great. They combine great field goal percentage defense with terrific defensive rebounding while (supposedly) not fouling. Their defense might even be underrated by Kenpom, as their only weakness is that they don't turn you over. However, this is not that big a deal, since when you play at a pace of 62 possessions per game, having a defensive turnover percentage of 18% instead of the median rate of 20% only equates to 1.25 fewer turnovers forced. Not that much of a price to pay for solid (or in this case, rock solid) fundamentals in all the other aspects of defense. I'm not sold on how the fouling numbers will hold up once they get in the NCAA tournament (it sure seems like they are pretty damn physical for a team that fouls so rarely), but their defense is definitely championship quality. Still, the fact that they play at such a slow pace makes them vulnerable to upsets, as it only takes an opponent having an out-of-body experience for 10 possessions or so for it to have a sizable impact on the outcome of a game. Marshall Henderson, anyone?
Creighton (KP Rank: 15, Seed: 7)
Creighton is like the mid-major version of Florida. A worse non-conference schedule than expected (remember when North Texas and St. Joes were supposed to be good this year? Yeah, I don’t either) was compounded by playing in a much worse MVC than was expected. Creighton took advantage in the early part of the year, completely blowing out most of their competition on the back of (in retrospect) unsustainable field goal percentages on offense. Although their offensive performance has come back down to earth of late, they are still the 6th most efficient offense in the country, and completely capable of blowing an offensively challenged team like Cincinnati out of the water. They are also completely capable of going toe-to-toe against Duke on offense, and they actually have a competent enough defense this year to be able to withstand a scoring drought or two. If you think Ryan Kelly can defend Doug McDermott one-on-one, allowing Duke to play straight up without compromising their D, then Duke is your pick. If you think that Duke will have to double McDermott, then I think you should pick Creighton. I’m a Ryan Kelly believer, so I’ll be picking Duke.
Missouri (KP Rank: 19, Seed:9)
I don’t know what to say in this space. Missouri has pretty much only lost either 1) close games or 2) blowouts to significantly better teams (Florida, Louisville….and Ole Miss. Which one of these teams doesn’t fit?). Their efficiency profile is that of a team with a much better overall record, which makes one wonder what is going on behind the scenes. My two cents: Phil Pressey thinks he is a hero. I think one of the worst things to happen to this Mizzou team was Pressey being labeled as SEC PoY in the preseason. In pretty much every close game that I watched involving Mizzou, Mizzou would play relatively unselfish, efficient offense until the final two minutes. In the final two minutes, Pressey would try and play hero, and usually fail miserably. I don’t think computers have any way to account for that. Additionally, Pressey is the only ball-handler on the team. They are completely reliant on him, which makes him going into hero mode all the more dangerous. I wouldn’t trust this team to beat anyone in a close game.
St. Mary’s (KP Rank 22, Seed: 11)
I said earlier that Creighton is the mid-major version of Florida, but St. Mary’s might be the West Coast version of Creighton. They have a lethal offense led by a guard I’m sure you’ve all heard of. Their offense may even be better than Creighton’s, or at least more well-rounded. Creighton shoots the lights out, whereas St. Mary’s shoots well while also sporting a very high offensive rebounding rate. This team is going to beat Middle Tennessee, and has a good chance of pushing Memphis in the next round. Memphis hasn’t played well in close games at all, and might struggle with a disciplined St. Mary’s offense. Memphis would also have an athletic advantage at every position on the floor, so run with that how you will.
Minnesota (KP Rank 23, Seed: 11)
Ahh Minnesota. The team that no one knows what to do with, other than accept their turnovers with grateful hands. This team pretty much shows you how far offensive rebounding can carry a team. They are the best offensive rebounding team in the country, rebounding 44.3% of their misses. That is an insane proportion. And it better be, because they are not good at shooting from the field or from the free throw line, and they are absolutely terrible at turning the ball over. Seriously. If you see Tubby Smith, give the man a hug. I can’t tell you the number of times I’ve watched a Minnesota game this year and seen him have to put his head in his hands after yet another ridiculous turnover. This is an example of how outliers in one of the four factors can lead to overrating one side of the ball as a whole – this is not a top-25 offense, but their offensive rebounding rate is driving up their rating. The reason this is important is because it is easier to gameplan for offenses like Minnesota’s, where there is one linchpin driving most of the offensive success. Interestingly enough, their prowess at offensive rebounding has not carried over to the defensive boards – this is a terrible defensive rebounding team (234th in the country). However, the two teams they would face in their first two games (UCLA and Florida) are not good offensive rebounding teams, and likely would not be able to make them pay for this rather sizable chink in their defense. And in the case of UCLA, they are also not a good defensive rebounding team and they struggle to turn over their opponents – making Minnesota a terrible matchup for them.
So I was going to continue this post to talk about which teams were greatly overseeded, but this endeavor has taken way longer than I anticipated and I kind of need to accomplish things early on this week if I’m going to do what I want to do on Thursday and Friday. Let me know in the comments if you like what I did, and if I hear good things I’ll try to bust out the counterpart section before Thursday.
Note: I also edited this and fleshed out the Wiscy section a bit to make the title a little less gratuitous and make it more pertinent to those fearing a reunion with Buckyball.