This is the first half of the bracket breakdown with log5 projections. For those not familiar with the log5 method, I did a similar analysis for every conference tournament. You can look at the very first preview to get caught up on what it is I do.
The folks that were here from the beginning of Championship Week might be wondering whether this method reasonably estimates the chances of a team's advancement. It worked out well for the conference tournaments. The most unlikely thing that happened was Georgia winning the SEC Tournament, which was given a 0.3% chance of happening. There were 133 events given a chance between 0.1 and 0.9%, and we would have expected one of them to occur. Indeed, the Georgia run was the only event in this range to happen.
Overall, a few more low-percentage events happened than would be expected, but it's something I'm comfortable with considering Friday and Saturday seemed to be more chaotic than usual. On the other end, there were 31 events given a chance of at least 90% of occurring and every one occurred. We would have expected two of them to not happen. (For the record, the most likely event to not happen was South Alabama reaching the Sun Belt finals at 89.0%.)
It should be noted that the underlying model used for these calculations--the Pomeroy Ratings--ranks teams relative to their conference mates better than it ranks teams relative to teams from other conferences simply because most games over the past couple of months have been between conference teams. That said, this exercise will at the very least explain why your brackets are about to get wrecked and why anyone that expresses certainty over his or her predictions is bluffing. The odds are against all but three teams in the field to get to the Elite Eight. The chance of the Doomsday Scenario--all four one-seeds making it to San Antonio--is 3.5%. Keep in mind, that's the most likely Final Four combination. It's worth noting that there's a 4.6% chance that a 16 beats a one-seed this season. Neither has ever happened, but year after year, they're both about as unlikely. That will not stop a lot of people from saying this is the year for all of the one-seeds to make it. It may be, but it's just as likely it's the year for one of them to lose their first game.
With the statistical mumbo-jumbo out of the way, let's take a look at the Midwest and East brackets today. All figures listed are the chances of a team (in percent) getting to a particular round based on log5 computations, with the rightmost column indicating its chances of winning the tournament.
2ndRd Swt16 Elite8 Final4 Final Champ
1 Kansas 98.55 93.61 82.05 61.47 50.16 36.81
3 Wisconsin 96.04 71.11 48.94 19.61 13.36 7.68
2 Georgetown 96.35 70.51 31.26 9.25 5.17 2.31
5 Clemson 77.09 60.96 12.60 4.93 2.36 0.88
11 Kansas St. 56.41 17.24 8.17 1.88 0.87 0.31
6 USC 43.59 11.23 4.64 0.90 0.36 0.11
7 Gonzaga 55.84 17.38 4.41 0.72 0.25 0.06
4 Vanderbilt 77.03 23.78 2.21 0.40 0.09 0.02
10 Davidson 44.16 11.77 2.52 0.35 0.10 0.02
8 UNLV 56.62 3.76 1.26 0.21 0.05 0.007
12 Villanova 22.91 12.40 1.08 0.18 0.04 0.006
9 Kent St. 43.38 2.32 0.66 0.09 0.02 0.002
13 Siena 22.97 2.86 0.10 0.01 0.00 0.00003
14 Cal St. Full. 3.96 0.42 0.04 0.00 0.00 0.000009
16 Portland St. 1.45 0.32 0.04 0.00 0.00 0.000006
15 UMBC 3.65 0.34 0.01 0.00 0.00 0.0000004
Kansas is the most likely team to come out of this region, but there are reasons to pick against them, too. Just three weeks ago, they lost to a team that didn't make the field in Oklahoma State. Yes, it was in a hostile environment, which is a situation the Jayhawks won't have to worry about the rest of the season. The ingredients in the loss were familiar, though. Combine a sluggish offensive night with an opponent that is uncannily accurate on threes. Of course, those two things rarely happen together. The Jayhawks played in a very difficult conference, had a few worthy non-conference challenges and yet lost just three games; 26 of their 31 victories were by double digits. Nonetheless, opponents do tend to shoot a bunch of threes against them, so the possibility of an early exit is out there.
Wisconsin has the best shot to derail Kansas' Alamodome plans. Their defense has been consistently nasty all season, preventing opponents from making shots from the field or the free-throw line. Their style never has won over basketball people, but statistically they have been a top-five team all season.
Georgetown has a potentially scary second-round game against Davidson on the horizon. The Wildcats' Final Four chances are Georgia-esque; their ability is probably underrepresented in my system by 10-15 spots. Whenever they lose--and it could be anywhere from the first round to much later--I'd be surprised if the game is decided before the final minute. They play big-conference-style basketball.
The only other team with a realistic Final Four chance is Clemson, which has about a 1-in-20 chance of seeing San Antonio. Like Tennessee, the Tigers are heavily dependent on their pressure defense forcing turnovers. That's a bad formula to use to beat Kansas or Wisconsin.
I'd like to give a special shout-out to Portland State. One of the secret screw jobs by the committee was not giving them a 15 seed.
2ndRd Swt16 Elite8 Final4 Final Champ
1 UNC 98.06 70.99 46.43 29.45 9.98 4.77
3 Louisville 94.66 72.82 45.07 23.21 6.60 2.72
4 Washington St. 91.16 61.12 28.45 15.35 4.02 1.54
2 Tennessee 95.66 68.56 35.57 16.29 3.95 1.42
8 Indiana 64.62 21.26 10.07 4.56 0.92 0.28
5 Notre Dame 81.77 34.29 12.01 5.04 0.92 0.26
7 Butler 71.24 25.35 9.11 2.73 0.39 0.09
6 Oklahoma 58.80 16.95 6.28 1.73 0.22 0.05
9 Arkansas 35.38 7.65 2.57 0.82 0.10 0.02
11 Saint Joe's 41.20 9.37 2.77 0.59 0.06 0.009
10 South Alabama 28.76 5.59 1.09 0.17 0.01 0.001
12 George Mason 18.23 2.91 0.33 0.05 0.00 0.0002
13 Winthrop 8.84 1.68 0.14 0.02 0.00 0.00004
14 Boise St. 5.34 0.86 0.08 0.01 0.00 0.000005
15 American 4.34 0.49 0.03 0.00 0.00 0.0000005
16 Play in 1.94 0.10 0.00 0.00 0.00 0.00000004
The East could be the most likely region to wreck your brackets. We'll see tomorrow that the South also has a claim as well. Louisville has the best chance to reach the second weekend, and yet there's better than a one-in-four chance that they won't. This has widely been described as the toughest region, though these kinds of claims are much like figuring out which conference is the toughest. It doesn't matter whether the East is the toughest, but it is the most competitive.
The Cardinals edge UNC for Sweet 16 probability honors because of the perceived strength of Indiana, a potential second-round foe for the Tar Heels. I'll vouch for Indiana's physical ability to test Carolina, but their mental strength appears to be shot under Dan Dakich. However, they're playing an Arkansas team in the first round whose coach spent the better part of the first half against Tennessee last Saturday having a profanity-laced conversation with the Vols' JaJuan Smith. So I'm not sure which team will be the least prepared to advance.
Back to Carolina...they beat Virginia Tech by 39 without Ty Lawson and won by two with him. Figure that one out. Louisville would be the worst regional final matchup for them by virtue of its outstanding halfcourt defense, especially down low.
I'm one of the few that isn't high on Tennessee, because they've had a bunch of close games since knocking off Memphis. Does Georgia winning the SEC Tournament make the SEC stronger or weaker? I'd argue the latter. The SEC is a distant sixth in the power conference hierarchy. There's a potentially interesting second-rounder with Butler, a team that takes care of the ball religiously. However, they've never faced a turnover-forcing squad like Tennessee. A UT/Louisville Sweet 16 game would be fun. One team likes to press for turnovers, the other likes to press to run clock.
Notre Dame and Washington State are other reasonable Final Four picks. Both teams appear to be better than last season's counterparts, ones that suffered premature exits.
Ken Pomeroy is an author of Basketball Prospectus.
You can contact Ken by clicking here or click here to see Ken's other articles.