In my blog post this week, I will show my findings from a regression analysis done on the correlation between Rushing Yards in a given game vs Defensive Rushing Yards Allowed. Contrary to popular belief, Fantasy Football is not a game of complicated stats, but a game of simple choices made in your initial draft and the subsequent lack of “tinkering” with your line-up throughout the season.  It is not about playing the lesser player vs the weaker defense, or benching the better player vs the stronger defense.  It is simply about always playing the better player in all but extreme cases. Using data pulled from 2011 on Running Back (RB) rushing yards in a given game and average Defensive Rush Yards Allowed, an initial simple regression analysis indicates that there is little correlat

The combination of a low R-squared and sub-0.005 F and p-values suggests that basing your decision simply upon Defensive Rush Yards Allowed is unlikely to result in a successful play. This is simply just for predicting rush yards, though, and not a predictor of points by any other means (Receiving Yards, Touchdowns, etc). As of right now, the data suggests that you’re best option will always be to play your best player in all but extreme cases, and not a lesser player facing a weaker defense. So if you’re thinking about starting Matt Asiata vs ATL (27 yards/game vs 124 yards/game allowed), Donald Brown vs JAX (27 yards/game vs 160 yards/game allowed), or Stevan Ridley vs KC (58 yards/game vs 130 yards/game allowed), DON’T DO IT. Though the matchups look sweet, these players have a low Yards Per Carry and have not proven to be reliable RBs. If you’re afraid to start guys like Demarco Murray vs NO (128 yards/game vs 101 yards/game allowed), or Arian Foster (if healthy) vs BUF (120 yards/game vs 83 yards/game allowed), don’t be. Though their matchups look tough, the best players tend to find running room in any game. If anyone would like further information on where the data was pulled from, or requests for the raw data itself, please let me know. In my next blog post, I’ll look to measure incr