Friday, March 21, 2008

Back to the Tourney

Look, there's only a few things I can't stand, and one of them is giving me stupid reasons why you picked the teams in the first round that you did. I'm not saying that you shouldn't pick the upset if you want to because they are bound to happen. In fact, in 2001, 41% (13 teams) pulled the upset over the higher seed in the first round. This included a win by Hampton over 2 seed Iowa St., two 13 seeds winning (Kent St. and Indiana St.), and 10 more upsets. This was the most upsets in the first round in the last 8 years. Picking the upsets are more fun, but I also find that winning 100 dollars in my pool is even more fun. Here are some steps that could help you do just this:

1. "Well, three of the four 11 seeds won last year, so I'm gonna have to go with this 11 seed Central-Eastern Michigan-Lansing State University" or some variant of this has been used for years by average bracketpickers. I even heard this used by Mike Golic about ten times in a four minute period on the Mike and Mike show, today. My ears cringed.
Look, there is no correlation between certain lower seeds winning from year to year. Don't believe me? I looked it up and graphed it:

The graph looks confusing, but basically the gist is this: The previous years' upsets have no bearing on the next years' upsets. If anything, there is an opposite correlation. Basically, if there were lot of upsets last year, you better not pick any this year.

2. Want to really pick an upset? Pick Gonzaga. In the past 7 years, they have managed a total of 3 upsets in the first round. But even after that they've never gotten past the Sweet Sixteen. Looks like they lost in the first round this year also.

3. Want to really win your pool? Don't pick any upsets. Don't pick the school that you went to, your friend went to, are in now, is in your state, or most of the people in the pool picked. Pick 148 for your final tiebreaker score (the average composite score of the last 7 years). Pick 151 for the final score if you're a stat expert and don't like outliers when compiling your average (yeah, Maryland-Indiana was the worst championship game in the past 10 years). Know that correlation doesn't mean causation. That means, if you're seeing a pattern and you're playing off it, it will fail you in the end. You can make a correlation between anything...just like this:

Seems like the more profanity in music nowadays is strongly correlated to the number of US Military Deaths in Iraq! That must be what's causing it! No, not even close...

4. The only teams that can be counted on for points every year in the tournament are the teams that have consistent tournament experience, good coaching, and a complete and solid team.
Recently, this list has been dwindled down to: UNC, Duke, Florida (for some reason they came up short this year...), UConn, and UCLA. A couple more are second-tier teams that hang out in the top 25 every year, those being: Kentucky, Arizona, Memphis, Michigan State, Stanford, Texas, Georgetown, Kansas, and the Oklahoma public university system. These can be counted on to round out your sweet sixteen. Good luck picking between them.

5. Think of upsets in cost effective terms. Purely, if you were to pick that 14 over 3 upset, more than likely that 14 seed will lose in the next round. That means you're giving up about 6 points that the 3 seed would likely win for you(2 for the second round, and 4 for the third round) just to get that one point for the upset. That 3 seed has a higher probability of going further and getting more points for you and giving up those points to pick an upset are just not worth it.

No comments: