NCAA Tournament Selection: The Manifesto -- Version 1.0

For every great piece of thought or revelation, there is a systematic element recorded in writing that explains laws, rules, and how decisions are to be made. The first version of The NCAA Tournament Selection Manifesto lays out the approach in which I predict the participating field. So, without further adieu, enjoy The Manifesto -- Version 1.0.

Q: How do you select representatives from one-bid conferences, in addition to conference champions?
A: Before I go into the questions, let me first state I am a major advocate for Ken Pomeroy’s ratings. Prior to conference play, I usually select the conference winner based on an average of the ratings percentage index (RPI) ranking and the Pomeroy ranking. Once conference games begin, I almost always pick the team with the best record in the conference, with RPI ranking as the tie-breaker. While this exposes some weaknesses in the system, the problems resolve themselves come mid-February.

Q: What are your major criteria (in no order)?
A: First, let me state that the polls play a small role in my predictions. Preseason predictions taint so many minds. Second, here are my criteria:

1. Conference RPI. That means how your conference fares against the others. I’m generally weary of outliers that skew the overall conference rankings.
2. An average of RPI and Pomeroy rankings. RPI measures strength of schedule and how a team does against that schedule. The basic formula is 25% team winning percentage (WP), 50% opponents' average winning percentage (OWP), and 25% opponents' opponents' average winning percentage (OOWP). The Pomeroy ratings, a more complete rating system, take into account the metrics of the game in addition to strength of schedule.
3. Conference Record. Ultimately, this depends on what conference you’re in. If you have a winning record in a top 3 RPI ranking conference, you should be in. Last year, I didn’t include Florida State. This is a rarity due to a very poor non-conference schedule. Greater attention must be shown to this by bracket predictors, as larger conferences tend to use unbalanced schedules.
4. Big Wins. I look for signature victories, with a higher road/neutral weighting. Date of game is a factor, but I don’t completely devalue an impressive win just because a win was in November or early December.
5. The Eye Test. This is very subjective, especially when considering teams who I do not have an opportunity to watch. Basically, does the team look good when I watch them? Last year, Air Force failed the Eye Test.
6. Bad Losses. In addition to losses against bad teams (especially against RPI sub-150 teams), I take into account margin of defeat. If a team consistently loses big when it succumbs to defeat, I will pay less attention to the RPI ranking.
7. Injuries/Suspensions. A great example of this is the suspension of Sean Williams. With him, Boston College is a sure-fire tournament team. Without him, they’re currently either among the last four in or on the outside looking in.

Q: How do you rank the one-bid conferences?
A: The two major factors are the RPI/Pomeroy average plays a huge role and the Conference RPI ranking of the one-bid school. I also look for impressive road and neutral court victories.

Q: Do you allot bids by conference?
A: I try my best to include the best 34 at-large teams in the pool. With that said, there are biases that exist. If your team is in a RPI top 4 conference, then you should have no less than 4 bids. That’s more of a guideline than a rule.

Q: What is your biggest weakness in selecting teams?
A: I would say it’s selecting the next four out (AKA teams 70-73).

If you have any further questions that you think should be answered in the manifesto, feel free to e-mail me at phashemi@gmail.com.