Week #4 Thoughts

LSU seems like a Jenga tower ready to tip over. And no one has done it yet. They almost lost to Washington State. Now they almost lost to Mississippi State. Who will finally knock them off?

Cal was demolished on the road, despite being called out for their road-woes. Terrible coaching.

Michigan has now won two games that could go either way, so I'm glad to see Rich Rodriguez getting a little bit of luck on his side.

#6 Cal and #9 Miami were beaten thoroughly. Shows the value of preseason rankings. And interesting thing to study is to figure out what goes into pre-season rankings and what determines them? Specifically, how does a coach from Purdue figure out if Cal should be ranked 9th or 19th? Or a coach from UCLA determining the relative merit of a Big 10 team versus an SEC team who played no common opponents the last five years?

Its a huge crap shoot. And unfortunately, pre-season rankings actually carry a lot of weight.

I'd also like to study how and why teams with high early season rankings lose, how they lose, and what factors (likely recurring) lead to some teams being over-hyped. As there is no repercussion for ranking a team high that loses, and I'm sure that coaches tend to over-value teams they know and under-value teams they aren't familiar with or that get little to no national media exposure.

Lastly, I'm sure there is a significant amount of "clumping" or simply coaches asking other coaches what they are ranking other teams, and simply doing the same thing. For example, a coach from Navy won't rank a team #5 in the country that someone else has unranked. Or vice versa. So there is less extremes and less desire to have independent views or to "stick ones neck out" and appear wrong and out of touch. Its better to be wrong with everyone else than be right alone. Because when the reverse happens, being wrong alone and everyone else is easily right, you look like an idiot.

This is a great post and a topic for further exploration. The concept of recurring, systematic errors in preseason rankings that are then found to be fallible later in the season, and that certain factors and consistently over-valued. And the basis for this recurring error is there is no punishment for being wrong AND a desire by each coach to have his rankings resemble those of everyone else's.

There are certain unwritten rules of how far a team drops after a close loss, a major loss, and upset, a loss to a higher ranked team, etc. Those rules are followed. I bet that those are pretty predictable, in fact. Another interesting topic. I wonder if I could make a formula to predict a teams rank the following week based on whether they won or lost, at home or on the road, and the margin of victory. Awesome topic.

And lastly, this of course would represent an investment opportunity if certain trends were discovered in the ranking system that were exploitable.