I’ve been participating in simple fantasy football pools for a while, although the closest I’ve come to winning was a tie for second place in the first year I took part. Picking winners and losers isn’t too hard — except when lots of upsets occur (see, for example, opening weekend this year) — but picking against the spread is much tougher. If you’re not doing as well as you’d like in your fantasy football picks, Gregg Easterbrook has a formula you might be interested in trying:
The Isaacson-Tarbell Postulate, proposed by TMQ readers Eric Isaacson of Indiana University and Catey Tarbell of Kirkland, Wash., holds: Best Record Wins Unless Records Equal, Then Home Team Wins. If you’d picked 2009 NFL games using this simple algorithm, you would have picked 167 of 267 contests correctly, a better performance record than many professional touts. The Isaacson-Tarbell Postulate is a huge time-saver because you don’t have to think about picks — you don’t even need to know who’s playing! Eric and Catey’s system works well except in the first week of the season, when the formula simply chooses the home team, and in Week 17, when many winning teams have locked in their highest seed and are resting starters. So last year I tested a Transformed Isaacson-Tarbell Postulate, which uses the formula except in Week 1 and Week 17; those I forecast the old-fashioned way, by thinking. Using the Transformed approach, I called 174 games correctly. That beat nearly all NFL writers and television personalities; the only one I saw who did better was Pete O’Brien of USA Today, who got 176 right.
This year TMQ will test a new auto-predictor, proposed by Tom Davis of Las Vegas. Davis writes, “A while back you noted in a column, ‘Studies have shown that the prediction most likely to be correct is simply forecasting that existing trends will continue.’ Why not apply this to the NFL? If a team won the previous week, predict them to win the following week; if they lost, predict them to lose again. If you have two teams who won last week or two teams who lost last week playing each other, default to the generic and choose the home team. It’s basically a tit-for-tat strategy, but I’d assume it would be as accurate as any of the ‘experts’ out there. And it does not require time-consuming thinking.”
The Davis Postulate holds: Existing Trends Continue; If Trends Same, Home Team Wins. I will track games using this predictor and also using a Transformed Davis Postulate in which I predict the first and final weeks of the season the old-fashioned way, by thinking.
I went 9-7 on the opening weekend (could have gotten 10, but I always take the Vikings to win). I may try the Isaacson-Tarbell Postulate this week, to make up lost ground.
