January 20, 2015

Good evening, blog!

I have began part one of my statistical analysis! The main focus for today's analysis was "what correlates with the number of points a team accumulates during the regular season?"

For this analysis, I inputted my data from an Excel spreadsheet on SPSS and computed the Pearson Correlation for twelve tests. For those of you 'non-math' lovers, the Pearson Correlation basically finds the relationship between if one thing happens, does another thing happen. For example, if a team has a high out-shooting percentage during the regular season, does that mean they will end the season with more points?

For the sake of this first analysis, I focused on the regular season only. Also, for the math lovers out there, my alpha significance was 0.01 (for those of you who are confused, that means I computed with a margin of error of less than 1%). 
 

Interesting Results

1) Pearson Correlation between number of points and number of wins in the regular season:  0.982.
This is a  very strong positive correlation. Naturally this makes sense because the number of wins you have correlates strongly with the number of points you recieve.

2) Pearson Correlation between the number of points and average number of goals per game in the regular season: 0.834.
This is a rather strong positive correlation. Therefore, the more goals you score, the better chance you have of winning (makes sense).

3)  Pearson Correlation between number of points and average number of goals against per game in the regular season-0.753.
This is a rather strong negative correlation. Therefore, more goals you let in has a negative effect on your success (obviously).

4)  Pearson Correlation between number of points and average out-shot percentage in the regular season:  0.888.
This was the most interesting statistic I found. What I got from it is that there's a strong correlation between being out-shot and earning points. So I guess teams are more successful with less scoring chances. I want to look into this statistic in particular more to see if it is credible or a fluke.

5) Pearson Correlation between number of points and average out-shooting percentage in the regular season:  0.821.
There is still a positive correlation when out-shooting your opponent; however, it does not seem as strong as when you are out-shot. 

6) Pearson Correlation between number of points and average win percentage when a team scores first in the regular season:  0.854.
This credits the 'first goal wins the game' theory since there is a strong positive correlation with winning and scoring first.

I found weak or no correlations when correlating number of points with face-off win percentages, trailing in the first period winning percentage, shots against average per game and shots per game.

Next Steps

I want to look into my findings further and support them with findings from past seasons as well as find other strong positive correlations which may not be as obvious.  
I will start doing this later this week!