2013 First Half Performance Review

As Week 9 roughly marks the half way point of the season, we thought it would be a good time to do a quick review of our performance: 

 Weightings

Thus far, we have been most heavily weighted on the Bucs, Patriots, Dolphins, Bills, and Browns. Though the performance of these teams has been mixed, they all share a common thread of underrated defensive play. Outside of the Patriots, these teams have seen value largely because they have been over-penalized for uncertainty or inconsistency at the QB position. 

Overall and Weekly Performance 

Overall, we have picked 22 games correctly and 21 incorrectly (51% with 2 pushes). Despite picking more games correctly than incorrectly, our total return (Return / Value at Risk minus 1) through 9 weeks is -2.7% due to the added "Juice." Our best teams, from strictly a cumulative performance perspective, have been the Chiefs, Panthers, and Jets, while our worst teams have been the Bucs, Texans, and Bears.

Looking at our performance on a weekly basis, our picks predictably struggled in the first quarter of the season while we gained a feel for each team, but rebounded in the second quarter, as we built a better feel for each team and capitalized on market inefficiencies created by public misconceptions. The chart below shows the average gain or loss per game per week. A perfect week would result in an average gain of 100 and a week with no wins would typically result in an average loss of -100 plus or minus the average "juice."

As a bonus, below is a node display of our performance by team. Team node size is relative to team pick weight (larger circles represent more frequently picked teams). Teams are connected to major nodes which represent the type of gain or loss that the pick resulted in. Teams connected to multiple major nodes (i.e the Bucs) are teams on which we have experience both gains and losses. Independent nodes represent less frequently experienced results such as -130.

Attribution 

Arguably the best way to look at performance is by attribution. Attribution measures added performance above a benchmark by measuring our team contribution to return versus the contribution to return gained if an equal weight is given to every team every week.  A team can add to performance in three major ways: picking a team that contributes above average return more frequently, picking a higher percentage correctly on a team than that team has experienced thus far, and a combination of picking a higher percentage on above average contributing teams. Our best teams from this perspective have been the Chiefs, Dolphins, and Jaguars, while our worst teams have been the Chargers, Bengals, and Bucs. Cumulatively, we have added approximately 1.84% of return above the benchmark.

Overall, our performance has been mediocre, but above average given the atypical unpredictability of the 2013 season. We look forward to boosting our performance over the second half of the season.