Week 16: Spread Picks, Results and Analysis

by thesanction1

Towards the end of the season the machine has come alive! Is it self aware? Are we in danger? Another killer week at 10 of 16 and the statistical weight of the results is starting to accumulate… Let’s get into it.

Overall performance – week: 10 of 16 (62%), season: 45 of 86 (52%)

Here is the performance if we ignored the absolute size of the Cover.Index (SCI), which is a proxy for the confidence of the pick, and simply picked every game:

Performance_OverallPicks_Season_2013_Week_16

SCI greater than 1.0 – week: 9 of 13 (69%), season: 39 of 71 (55%)

Here is the performance if we restrict our selections to those games in which the SCI is greater than 1.0:

Performance_HighConf1_Season_2013_Week_16

SCI greater than 2.0 – week: 6 of 9 (67%), season: 24 of 47 (51%)

Here is the performance if we restrict our selections to those games in which the SCI is greater than 2.0:

Performance_HighConf2_Season_2013_Week_16

SCI greater than 3.0 – week: 5 of 8 (62%), season: 19 of 33 (58%)

Here is the performance if we restrict our selections to those games in which the SCI is greater than 3.0:

Performance_HighConf3_Season_2013_Week_16

SCI greater than 4.0 – week: 4 of 7 (57%), season: 12 of 21 (57%)

Here is the performance if we restrict our selections to those games in which the SCI is greater than 4.0:

Performance_HighConf4_Season_2013_Week_16

SCI greater than 5.0 – week: 4 of 6 (67%), season: 11 of 16 (69%)

Here is the performance if we restrict our selections to those games in which the SCI is greater than 5.0:

Performance_HighConf5_Season_2013_Week_16

Obviously these results are encouraging.  I don’t want to break down the statistics too much until the season has ended, but from an anecdotal perspective the algorithm is doing better than I could have expected.

Week margin of victory statistics:

Performance_WeekWinMarginGrid_Season_2013_Week_16

This performance is confirmed at all threshold levels by the margin of victory stats as well.  Not only did the algorithm pick 62% of games correctly against the spread – the games it picked correctly were won, on average, by 11.2 more points than the games it picked incorrectly were lost by.  The most confident pick of the week, PHI over CHI, ended up an absolute blowout which no one (well, no person I suppose) expected – PHI covered the spread by an astounding 40 points.  There were bad beats as well. IND walking all over KC was certainly a shocker; KC failed to cover by 22.5 points.  But on average we are losing more close games, and picking more blowouts, indirect evidence that the algorithm is on to something.

Season margin of victory statistics:

Performance_SeasonWinMarginGrid_Season_2013_Week_16

The margin of victory trend for this week is just an extension of the season long trend.  One thing to note is that the only category which had previously had a negative Margin.Difference, the highest SCI.Threshold picks, was turned positive this week.  Undoubtedly the aforementioned PHI game tipped the scales.

Alright – that’s all for now.  It was an early Merry Christmas for those who bet the picks, more to come next week!