From my biased view, it was another successful year for PHSports’s bracket projections. While readership was down, comments increased. When the brackets were released, PHSports finished an admirable 17th out of 83 brackets.
Like many, I was left scratching my forehead when A-10 regular season and postseason champions Temple were given a #5 seed and a matchup with a senior-laden Cornell squad. Clement and I still can’t tell how Notre Dame got a #6 seed, but they did earn a date with an undesirable opponent in ODU. Also, we have no clue how Duke’s resume was better than that of Syracuse, or West Virginia and Ohio State (WITH Evan Turner, which is how they should have been evaluated).
I was appalled by the laziness of the committee in putting two Big East in the same sub-region TWICE.
Back to our performance this year. What did I do differently?
First, I watched less basketball, in comparison with the last 3 years. Yes, I didn’t stutter. In 2006 (and a few years prior), before PHSports was on the World Wide Web, I did bracket projections for my own edification. The catch was that I was on international travel for 2 weeks during January, was completing graduate school, and even went out of town during Selection Sunday weekend. Had I posted my bracket, it would have finished first. By a lot. Why? Because I probably watched as much basketball as the members on the committee evaluating teams, meaning I did not peel as many layers of the onion (i.e. examine all available data) as many of my colleagues in the bracketology community.
Second, I avoided the horserace journalism mentality of updating brackets every week during the conference season. This year, mostly due to time constraints and life choices, I chose to update brackets once every two weeks during the regular season. I cannot tell you the number of times that I saw sites – great and awful – overseeding teams who had a 2-win week and then crushed them for a letdown the following week.
Third, I factored in whether a team either won a share of a conference or won outright. This is why I was one of the few who gave Maryland a #4 seed. This is why UTEP, Utah State, and Cal (#8 seed, my behind) were all in my field with zero hesitation.
Fourth, to gather data, I populate team-specific index cards (yes, I’m old school) including school name, team record, conference record, injuries / suspensions, RPI, Pomeroy rating, key wins and losses. This year, I noted whether a team won their conference, as well as their record against the RPI Top 50, and record against the RPI Top 100. Late Saturday night / early Sunday morning, I recognized that Tennessee was 2-5 (wins against Kansas, Kentucky at home) against the Top 50, and moved them from the 4 to the 5 line.
Fifth, I conducted a more thorough final examination of teams from the one-bid conferences, usually residing along seed line #s 13 through 16. In 2009, when you plow through the data, PHSports’s final projections were as good as nearly everyone else (if not better) for the first 12 seed lines. When it came to those last four seed lines, I was among the worst. This year, PHSports did not reach world-beater status, but because we addressed this concern, the site finished ahead of some of the big guns.
So what will we not change? Of course, PHSports spots relevant statistical trends almost as fast as Usain Bolt runs, and we’ll continue to do that. We’ll also continue to encourage our readers to study the data, and not just point out the elements within the data that support an assertion (yes, I’m talking to you, UAB fan, and to several Minnesota haters).
Until next year, stay classy San Diego.