Wednesday, May 30, 2007

Not Convinced

I know the Suns have been out of the playoffs for a while now, but I was out of town. Plus, I'm still bitter. Given the sleazy way the Spurs played I must find a way to say that they are not a better basketball team - and I have.

Sure the Spurs won the series 4:2, but did they significantly outscore the Suns? Let's do a test of significance.

We are trying to show whether the Spurs score minus the Suns score is statistically different than zero. A simple paired t-test will suffice. This test will show whether the difference in scores is due to random game to game fluctuation, or whether the Spurs statistically outscored the Suns during the series and are therefore a better team. If the Spurs score was statistically different (i.e. 'significant'), the P-value will be less than 0.05.

Lets look at some data:
Format (Spurs Score, Suns Score, difference)
Game 1: (111, 106, 5)
Game 2: (81, 101, -20)
Game 3: (108, 101, 7)
Game 4: (98, 104, -6)
Game 5: (88, 85, 3)
Game 6: (114, 106, 8)
Average: (100, 100.5, -.5)
Std Dev: (13.3, 7.9, 10.8)

And finally:
t-value: 0.1136
p-value: 0.9140
Degrees of Freedom: 5

Wow, not even close to significant. In fact, the Suns averaged more points than the Spurs in the series, thanks to their 20 point victory in game 2. Even if you don't include game 2, the P-value is barely less than 0.25.

In review, I must recommend the Suns-Spurs series be extended a few more games.


Anonymous Owen said...

Hear, hear!

9:08 PM  

Post a Comment

<< Home