"Regression to the Mean" refers to the tendency of a variable to move away from extreme values towards the average value as more data is added to the set. Our specific data set is the career Win Loss record of our beloved Alexander Douglas Smith and the question before us is - Will ADS's 2012 performance reflect his average performance over his career or will it mirror last season's 13-3 record?
Lets regress -
2005: 7 GS, 2-5, 1 TD, 11 INT, 50.9 Cmp%. Rookie year, new head coach, dramatically different system than his college team. Performance in line with expectations.
2006: 16 GS, 7-9, 16 TD, 16 INT, 58.1 Cmp%. Second year under Mike Nolan. Modest improvement.
2007: 7 GS, 2-5, 2 TD, 4 INT, 48.7 Cmp%. Norv leaves to coach the Chargers, leaving Smith with his third OC in three years
2008: Smith was placed on IR in September with a broken bone in his shoulder.
11 GS, 5-5, 18 TD, 12 INT, 60.5 Cmp%. Nolan out, former interim coach Mike Singletary is in.
11 GS, 3-8, 14 TD, 10 INT, 59.6 Cmp%. New OC. Same ASD.
16 GS, 13-3, 17 TD, 5 INT, 61.3 Cmp%, New OC. New Head Coach. New ASD.
So over his career, ASD has started 66 games and posted a 32-34-0 record. His Cmp% has averaged 56%.
So what will we see in 2012? The new improved ASD? Or a regression to the mean performance in the years prior to 2011 (something around 5 wins statistically)?
What do you think?