The original blog entry brought up the question of sample size. And after some digging I found previous years data (2008 to 2011) to look at the effect of travel distance and back to back games on winning.

Do Teams That Travel Less Win More?

The correlation four all four years (2008 to 2010 and 2011 YTD) is poor (r-squared less then 2%) so there appears to be little relation between distance travelled and winning games. What can be said, is that statistically less then 2% of the variation in P% is explained by the distance a team travels. To be clear, a relationship may still exist and a more sophisticated test could separate long travel games from short travel games to see if there is winning difference between those two datasets but over the season it does not appear to overcome other variables (luck, skill or other). Note 2011 data is YTD.

Do Teams that play less back to back games win more?

The correlation four all four years (2008 to 2010 and 2011 YTD) is again poor (r-squared less then 2%) so there appears to be little relation to winning and the number of back to back games. What can be said, is that statistically less then 2% of the variation in the season P% is explained the number of back to back games played. Note 2011 data is YTD.

To be clear, a relationship may still exist and a more sophisticated test could separate the second back to back game played from the rested games to see if there is winning difference between those two datasets. That is a onerous task. Looking at the data in another way, the average # of back to back sets per year is 14.97 with a 2.66 standard deviation). That mean 18% of the schedule is back to back games and so the variability of those 14.97 games on the 82 season total may not show up in the regression of P%. That is, the variability of those 14.97 games (18% of the season) vs 66.03 games (82% of the season) may exist, but this regression does find the number of back to back games to be important. If a more sophisticated study exists I would like to see the results.