Lindy's Five Essential Websites (Non-Major Media) for 2013
[+] Team Summaries

Wednesday, August 6, 2008

The People's Poll: Everyone Else's Top 25

Now that all the "pundits" of the blogosphere are taking their potshots at the Coach's top 25--some well-informed and others full of cliche's and misinformation--I now present the season's first People's Poll so we can examine our own choices.

The People's Poll is a representation of voting data from prediction markets (see here). I'm a big believer in the power of prediction markets as a tool for aggregating a massive amount of data into probabilities. I've used these probabilities to generate a ranking of the top 25.

And here's the beautiful part--if you disagree, you can cast your own vote at any major prediction market with college football options (e.g. Tradesports). Just buy the teams that should move up and sell short the teams that should drop down. And if you're right, you can make you, or your favorite charity, a little money.

This early in the season, the trading is still relatively low so plenty of winnings are available to the smart investor at the margins. You might, for example, think Notre Dame should not be at 18, that Kansas deserves better than 21, or that BYU belongs in the top 25. If so, you can place your reputation or petty cash on the line.

Good Luck.


  1. Great idea! Have you done this over any of the past football seasons, and if so has it panned out better than the "official" polls?

  2. Wish I had tried it in past seasons. The idea just occurred to me this morning and I ran with it. I'm excited to see how it turns out.

  3. Let me get this straight. A prediction market list is one where people are placing a monetary amount beside their prediction of what will happen in the future, right? And I think that at one time, the Pentegon was suggesting using this as a way of predicting future terrorist events...

    All irony aside, can a prediction market on a season that is so heavily predicated upon predictions for the final results do very well? Seems like it almost has two variables that affect it, rather than one. Or maybe that's the beer talking. Esplain, please...

  4. Your example of the Department of Defense (DOD) using prediction markets for predicting terrorist events is a good example of prediction markets gone wrong because the outcome of the market would be used to try to alter the outcome predicted by the market. Consequently, no intelligence could be generated from such a market because the price also reflects the DOD's capacity to alter the outcome using the information. Therefore, the price reflects two variables--the likelihood of a terrorist plot and the likelihood that said plot can be averted using market information--and the effect of each variable cannot be differentiated.

    I do not believe the market for college football is analogous. I would disagree that the outcome of a season is too heavily predicated on predictions as your comment suggests, because a final poll reflects a 16 stage adjustment (one for each week) to reflect real outcomes. This is sufficient to compensate for any initial bias.

    It is also important to note that, unlike the DOD case, Tradesports does not have a heavy influence on coaches or the media (at least, not yet), and therefore the system does not erroneously generate self-consistent results.

    But that is a very thought provoking question and one I will try to address more rigorously in future posts on the People's Poll. I'm also not completely sure if I have actually addressed your question, so let me know if I missed the point.