Commentary

Tournament Challenge: Inside the numbers

Updated: March 17, 2009, 8:07 PM ET
By Andrew Feldman | ESPN.com

This just might be my favorite column of the year. On most days you'll find me writing about pot-odds, calling raises in position or news from the poker industry. Not today. Today, it's all about college basketball.

The brackets in the ESPN Men's Tournament Challenge are open and since the selection committee released the field on Sunday night, users have been quite busy trying to predict the upsets while not remaining unduly loyal to their favorite teams … if they can. University of Michigan alumni, I'm looking at you.

So, how is your bracket coming along? There are some very tough decisions that need to be made, and I hope you'll be able to use the information below, one way or another, to decide what's best for your bracket this year. The data comes from you, the Tournament Challenge participants. Every bracket is taken into consideration and the statistics are accurate as of noon ET on Tuesday.

Not much has changed since 2008. ESPN.com users still love the Tar Heels.

During my column last year, I noted that more than 33 percent of users selected North Carolina to cut down the nets. This year, UNC has a little more competition, but still dominates the rest of the field when it comes to user predictions. Twenty-eight percent of users believe that North Carolina will be the 2009 national champion. The Heels' next closest competitor: Louisville at 17.7 percent, followed by Pittsburgh at 16.1 percent.

Unexpectedly, the fourth No. 1 seed isn't next. Last year's runner-up, Memphis, jumped Connecticut as users predicted the Huskies to win the title only 7.5 percent of the time versus the Tigers' 8.7 percent.

Faith in the No. 1 seeds?

For the first time in NCAA tournament history, all four No. 1 seeds made the Final Four in 2008. Would users feel that the unexpected could happen again in 2009? Well, somewhat. Users have predicted in 10.4 percent of all brackets that all four top seeds will be heading to Detroit. If you take Connecticut out of the mix, 24.9 percent of users predict that the other three No. 1 seeds will make the national semis. Looking at each team individually, Louisville is deemed to have the easiest path to the Final Four, predicted at 62.1 percent. UNC is at 60.9 percent, Pittsburgh at 57.2 percent and Connecticut at 40.1 percent.

This year No. 1 seeds are predicted to win the championship 69.5 percent of the time, significantly below the 77 percent in 2008. Why would users be less confident in the No. 1 seeds? Check out our college basketball section to find out.

Finding the 5-12 upset

Who doesn't love the 5-12 game? Almost every year, we wait with bated breath to see which No. 5 seed will fall "unexpectedly." Sure, "unexpectedly." Many of those filling out their brackets on ESPN.com have made sure to predict that the "unexpected" will occur. This year's chosen victim: Utah.

In the first three 5-12 matchups, users predicted Purdue over Northern Iowa (89.7 percent), Florida State over Wisconsin (79.4 percent) and Illinois over Western Kentucky (63.1 percent) to reach the second round relatively easily. What about the Utes?

53.6 percent.

Ouch. Hello, Arizona. One of the final two teams added into the field of 65 is predicted almost half the time to win its first game. Is this more name recognition than anything else? We'll see, but the confidence is high that all four No. 5 seeds won't make it to the second round, as users predict that all four will advance only 27.5 percent of the time. Just for fun, there are 21 entries that believe all four No. 5 seeds will be in the Final Four.

The 7-10 game

As the seedings get closer, we'd expect more upsets from the lower seeds … right? Not this year. The 6-versus-11 games are predicted to be smooth sailing for the higher seeds. According to the majority of user predictions, all four No. 6 seeds will move on, with UCLA coming in with the lowest confidence of advancement at 72.5 percent (West Virginia is the highest at 82.4 percent).

Then comes the 7-10 matchup.

California. Great state, entertaining governor and predicted more than 65 percent of the time to fall in the first round to Maryland. So much for being seen as the "favorite."

Users also are predicting half of the time that 10-seed USC will defeat Boston College. Users are more confident in the other two 7-seeds, as Clemson and Texas are predicted 62.9 and 74.4 percent of the time, respectively, to move on. Only 10.9 percent of users have faith in all of the 7-seeds moving on, compared with 40.9 percent who believe all 6-seeds will advance.

Loyalty

When filling out their Tournament Challenge brackets, users strive to make the most educated and researched selections in order to win the grand prize. However, some users can't let go of the teams they love most. It isn't unusual to see a couple people select the 16-seed to win just because they know nobody else has done so, but beyond that reasoning, why would people actually predict certain teams to have a chance to win when all signs point to a first-round elimination?

As I went through the list of predicted national champions (in order of user entries), I saw the typical 1-seeds, 2-seeds, etc. You get the point. After an appearance by fourth-seeded Gonzaga and UCLA, a No. 6 seed, jumping some 4s and 5s, there is one questionable result: the University of Michigan. As an alumnus -- "Go Blue!" -- but in all seriousness, the fact that 0.2 percent of brackets pick No. 10 seed Michigan to win the national championship is quite surprising.

Based on their seedings, other questionable teams include Maryland, Ohio State and Tennessee. Some higher (6 or better) seeds that aren't given a chance include Xavier, Illinois, Arizona State and Marquette. As mentioned earlier, it seems that No. 5 Utah is given the least respect out of all teams. Out of the field of 65, Utah has only enough bracket support to be ranked 37th on the list of teams that users believe will win.

Unlucky matchups

It's tough to be a 15th or 16th seed. You get the "privilege" of being invited into the tournament only to face one of the toughest teams in the country. Congrats on making the Big Dance, we hope that you enjoy your quick stay.

Here is a list of the teams that are picked least often to win their first-round games:

(16) Opening-round winner (Alabama State-Morehead State) -- 0.3 percent
(16) Radford -- 0.5
(16) Chattanooga -- 0.5
(16) East Tennessee State -- 0.5
(15) Cal State Northridge -- 0.6
(15) Morgan State -- 0.9
(15) Binghamton -- 1.0
(15) Robert Morris -- 1.0

The most surprising team in the bottom 20: No. 11 seed Dayton, expected to win just 17.6 percent of the time.

Quick hits

• 51.1 percent of users believe that all four No. 1 seeds will make the Elite Eight.

• 10.0 percent believe that the Elite Eight will contain all four No. 2 seeds. Only 0.2 percent of entries believe that we will have an all-2-seed Final Four.

• The most commonly selected final game (17.5 percent of entries) is North Carolina versus Louisville. UNC is predicted by most to win.

• Wake Forest is the most surprising team, in my mind, that is predicted to make the Final Four. Currently, 3.4 percent of users believe that Wake will come out of its bracket -- a higher percentage than fellow 3-seeds Villanova and Missouri are receiving.

• For the second year in a row, users do not like Xavier. I'm not sure why that is. Maybe it's the Musketeers' tough region or maybe it's because the school's name starts with an "X." Either way, Xavier is picked to make the Final Four 0.4 percent of the time, falling behind almost all of the Nos. 1 through 6 seeds. No. 5 seeds Utah and Illinois are also way below their expected placement given their seeding (0.1 and 0.2 percent, respectively).

• 10th-seeded Maryland is right behind Xavier in Final Four picks at 0.3 percent.

• Read Keith Lipscomb's bracket bits column for more interesting trends, of the historical variety.

Final thoughts

Users predict that 90 percent of the time, the Nos. 1-3 seeds will coast into the second round. For the most part, the 4-seeds fit into that category as well, except for Washington (predicted to win only 73.6 percent of the time). Not until the No. 5 seeds are users ready to take risks in the first round, and many of them do.

Best of luck to you and your selections. I hope that none of your brackets are busted after the noon games on Thursday.

You can read more from Andrew Feldman in the ESPN.com Poker section.

Andrew Feldman is ESPN.com's Poker Editor. He is the host of the Poker Edge Podcast and co-host of ESPN Inside Deal. Andrew has covered the poker industry for ESPN since 2004.