Monday, November 21, 2011

Complain , but Stats dont see it

Since the NBA is on a lock out, ESPN has decided to go into hibernation, but I wont let that stop PatchStats. It has been a wild month, working on publications, and assisting the Saberphere on matters. However, we are back to the grind giving you a new look and reason to things.

It has been a long held belief that in sports, playing back to back games leaves a huge disadvantage for the team playing the back-to-back game. This past month I watched the Maple Leafs and Capitals both drop games on back-to-back days. The theory carries heavy in the NBA, where teams typically travel very far for one game, then off to another city for the next game. A lot of the scheduling has to do with revenue. Teams would love to get Friday,Saturday and Sunday games since they produce the highest attendance, aka Cash Money. But players cant do this, because for now, they are just humans. Disgruntled fans, and some coaches, have complained when they get a schedule with back-to-back games, as it is almost assumed the second game is a loss. However, thoughts are not numbers, and on this blog we use numbers to make conclusions.

In 2010-2011, 9 teams had a higher winning percentage on back to back games. [Canucks,Senators,Bruins,Blue Jackets,Penguins,Blackhawks,Rangers and Blues], of those 9 teams, only two of them did not make the play-offs. So most teams, had a worse winning record when playing back-to-back. However the true question is, how big is this difference, is one game, or 10, these things matter in sports. The thing is, the difference between regular season winning percentage and back-to-back games was on average only .08. Most teams only play about 1 out of every 5 games on a back-to-back schedule, the team with the highest percentage was Buffalo who played 27% of their games after a prior game.

This is your Statistician Speaking : Here is the testing done to show it has little effect

Ho : The Difference between Back-to-Back games and Regular Season Record  = 0
Ha : There is some difference
T = .09/.07
T = 1.37
Thus, we conclude that there is no difference in the result of the game due to playing a game the day before.

There is some evidence, but not very strong evidence. Still, people will complain, and every loss will be because "We played the day before"

For now, keep those skates moving, because points are points.


Loopaleep said...


V for ..... said...

did you test for differences within the subset of all playoff teams and all non-playoff teams? that might give an interesting predictor into post-season success