StatCar Racing: Homestead

November 16, 2008 9:07am CST

Season Review – StatCar Performance

As we enter the season’s final week, it presents a good time to evaluate our performance compared to users overall—as well as other experts. For the year, our Yahoo! team is ranked in the top few percentile, despite missing one week altogether because I forgot to set my lineup.

In the competition against the Fanball Cheat Sheets a slight edge goes to the Cheats. Performance of both models was almost identical in terms of the percent of drivers who finished within 0,1,3 and 5 places of their forecasted positions.

Where the the Cheat Sheets had an advantage was forecasting drivers near the rear of the field, which resulted in them having a 1 point better average difference between pre-race ranking and driver finish position (9.6 compared to 10.6). At the front of the field, however, our StatCar model had a slight edge with our top ranked driver besting the Cheat Sheet’s numero uno two more times.

In the spirit of fair competition, we will go double or nothing in the final race.

Overall, using either model will land your team near the top of the fantasy standings. Using both will put you in an even better position.

A sincere thanks to Dan and the Fanball staff for letting me join this season.

This Week’s Predictions
Reviewing last year’s race at Homestead reveals the following variables that were significant predictors of driver finish position:

Variable, Impact Coefficient
Constant, 29.3
Total Races, -7.5
Pass Differential – Texas, -0.19
Laps led – Atlanta, -0.4
Average Finish – Homestead, 0.24

Applying these relationships to each driver’s current information yields the following rankings. Go Kenseth!

Rankings