It's been a while, hasn't it? I promised John Gasaway I wouldn't write anything this season unless I disagreed with something he wrote, and it took until yesterday for that to happen.
I have to confess to having cringed a little when John mentioned the poor performance of log5 in conference tournaments. Sure, the method didn't do a good job of predicting who was going to win the conference tourneys, but that's not what it's supposed to do. When log5 tells you that Kansas has the best chance of winning the Big 12 tourney, it doesn't mean it's telling you they should win it. In fact, log5 explicitly said there was about a 70% chance they wouldn't.
We're dealing in probabilities here, which is why I can't end this piece by telling you who's going to win it all. Given the wild outcomes in the three- and four-game tournaments last week, one can imagine what's possible in a six-game event. It's the kind of uncertainty where one can agree with John that Memphis is the "favorite", but also say with 80% confidence that your champion is going to be somebody else.
On that note, let's dive into the second half of the log5 analysis of the NCAA field and take a look at the West and South regions.
Seed 2ndRd Swt16 Elite8 Final4 Final Champ
2 Memphis 96.09 83.91 63.41 45.21 32.42 22.54
1 Connecticut 97.89 70.35 47.38 22.01 13.05 7.39
3 Missouri 90.36 60.52 21.93 11.20 5.77 2.82
5 Purdue 82.53 47.13 19.71 6.60 2.98 1.26
4 Washington 76.23 41.79 17.12 5.61 2.48 1.03
8 BYU 71.29 24.36 12.36 3.85 1.63 0.64
6 Marquette 71.68 30.69 8.37 3.33 1.34 0.50
7 California 62.17 10.71 3.87 1.21 0.38 0.11
9 Texas A&M 28.71 5.17 1.55 0.26 0.06 0.01
10 Maryland 37.83 4.50 1.18 0.26 0.06 0.01
11 Utah St. 28.32 6.94 1.00 0.21 0.05 0.01
13 Mississippi St. 23.77 6.97 1.30 0.19 0.04 0.01
12 Northern Iowa 17.47 4.11 0.58 0.06 0.01 0.00
14 Cornell 9.64 1.84 0.14 0.01 0.00 0.00
15 Cal St. Northridge 3.91 0.88 0.11 0.01 0.00 0.00
16 Chattanooga 2.11 0.12 0.01 0.00 0.00 0.00
Yes, Memphis' chances are absurdly high and in fact, the highest of any team in the field by a factor of two. Their lofty status is the result of late-season mediocrity among the one-seeds and the string of microscopic defensive efficiencies put up by the Tigers in March. This created a rather large gap between Memphis's adjusted defensive efficiency and its nearest competitor. For good reason.
Over the three games of the Conference USA Tournament, Memphis opponents scored about 0.7 points per possession. In John's endorsement of the Tigers, he noted they allowed 0.86 points per possession during regular-season conference play, which was easily lower than any other contender. Yes, yes, it was against the weaker competition found in CUSA, but you can imagine how far from the norm 0.7 is, and it was against the upper half of the league. In defense of the conference, two of Memphis's victims on this run, Tulsa and Houston, are roughly the equivalent of NIT-worthy ACC and Big East teams. Even at home, the total shutdown of teams who were thinking about stealing an NCAA bid is startling.
The difficulty of assessing Memphis's chance going forward is translating dominating performances against these teams to potential match-ups with stronger opposition such as California, Missouri, Marquette, Purdue, Washington or Connecticut. To my knowledge, little research has been done in this area. For that reason alone, I'm looking forward to seeing how the Tigers do. Put me in the camp of people that think it's possible for a team to be great without having great wins in its portfolio.
An interesting experiment is to make Memphis' strength equal to UConn's, which may be a more realistic model for predicting the tournament. This results in the following chances for a national title...
John Calipari got a tremendous gift from the committee--the "insult" of being a two-seed with a degree of difficulty no worse than if he were a one.
Seed 2ndRd Swt16 Elite8 Final4 Final Champ
1 North Carolina 96.95 79.61 46.89 31.60 18.90 9.62
4 Gonzaga 91.42 68.67 37.95 25.08 14.66 7.26
6 Arizona St. 73.55 42.06 24.49 10.22 4.80 1.84
3 Syracuse 85.78 46.40 26.01 10.31 4.62 1.68
2 Oklahoma 91.85 53.65 26.50 9.89 4.19 1.43
7 Clemson 68.28 34.70 16.22 5.66 2.26 0.72
5 Illinois 83.12 27.37 9.50 4.18 1.53 0.45
9 Butler 50.62 10.22 2.66 0.85 0.22 0.04
8 LSU 49.38 9.79 2.50 0.78 0.20 0.04
11 Temple 26.45 9.06 3.07 0.66 0.16 0.03
10 Michigan 31.72 10.70 3.16 0.65 0.16 0.03
14 Stephen F. Austin 14.22 2.49 0.45 0.05 0.01 0.00
13 Akron 8.58 2.14 0.27 0.04 0.01 0.00
12 Western Kentucky 16.88 1.82 0.20 0.03 0.00 0.00
15 Morgan St. 8.15 0.96 0.08 0.00 0.00 0.00
16 Radford 3.05 0.38 0.02 0.00 0.00 0.00
Second on the list of apparent log5 absurdities is the chance of Gonzaga winning the title. With the Zags' last marquee game being a humiliating loss in Spokane to Memphis, there's some well-founded skepticism that their chances of escaping the South region could approach a healthy UNC's chances. A banged-up UNC is a different story, of course, so let's consider this the Ty Lawson-with-a-bad-toe scenario. (In addition, this analysis doesn't include a home-court boost to Carolina for the first two rounds.) In that case, I can sign off on Gonzaga as the second choice from the region. Note also the three-way photo finish to get to the regional final from the other side of the bracket. Arizona State, Syracuse and Oklahoma all have roughly equal chances.
More than a few people have noticed that my ratings this season are much more compressed than they were in either of the last two seasons. They're comparable to 2006 and, for those with short memories, the seeds of the Final Four teams that season were 2, 3, 4, and 11. This year, our log5 analysis indicates that no single team that has at least a 50% chance of reaching Detroit. The probability of all four one-seeds making it there together is less than one percent. The chance of none of them getting there is a hefty 23%. And, oh yeah, the probability of at least one of the six seeds playing in April is a whopping 42% (ignoring the effects of a potential UCLA road game in the second round).
What does this all mean? A week from now, some predictions made this week are going to look bad. Two weeks from now, a lot of those predictions are going to look bad. Three weeks from now, when "One Shining Moment" fades out, many analysts are going to provide simple explanations for the seemingly shocking events that have taken place: Memphis's offense was too dependent on Tyreke Evans. UConn couldn't survive without Jerome Dyson. UNC didn't play enough defense. We should have known that Matt Kingsley could put the Lumberjacks on his back for six straight games. Et cetera.
OK, that last one is a stretch, but you get the idea. (Keep in mind that Stephen F. Austin's defense ranks significantly better than the Tar Heels' does right now.) Depending on the outcomes, there will be some truth to each of those lines. It will seem so obvious then. Just like it was clear after the 2006 event that Joakim Noah and Al Horford were rapidly developing interior forces with an exceptional supporting cast.
However, nobody was saying those things about the Gators this week three years ago. Not in the context of winning a national title anyway. Likewise, about the only thing that seems certain now is that there isn't a team in the field that doesn't have significant questions surrounding its path to Detroit. That should make for an entertaining tournament--and plenty of messy brackets by this time next week.
Ken Pomeroy is an author of Basketball Prospectus.
You can contact Ken by clicking here or click here to see Ken's other articles.