Saturday, December 06, 2014
Corsi v Tango: Soccer Version
?A Straight Arrow reader pointed me to this post from Mark Taylor who has a soccer blog. While he did not (initially) run it like I did, I asked him:
Because you are correlating to the seasonal differential, you are, as you stated, not doing exactly the same thing. As a result, as you approach the end of season, you are in effect correlating x to itself.
Could you repeat the same exercise I did, but correlating current data to future data? So, first two games, to next 36 games. First three games to next 35 games, and so on? Presumably, you'll reach a peak when you correlate 18 to 18 games. And it would be interesting to see the pattern for the various shot results.
And he came through, at least with the split-half correlation:
Splitting the season in half, the goal difference in the first half of the season has a r of 0.72 to goal difference in second half.
If you include shot differential for shots that were saved and shots that missed the target, r between these inputs and future goal differential increases to 0.78.
Goal Difference in 2nd half of season=0.46*GD first half+0.021*wide shots differential first half +0.137*saved shots differential first half.
Just to be clear, in my terms which I put relative to goals, that would be:
- 1.00 Goal Differential
- 0.30 Save Differential
- 0.05 Wide Differential
Which... well... is in the ballpark of what we see in NHL. And in his case, it was only 18-game to 18-game correlations, and ALREADY the goal differential was able to shine through.
Recent comments
Older comments
Page 1 of 151 pages 1 2 3 > Last ›Complete Archive – By Category
Complete Archive – By Date
FORUM TOPICS
Jul 12 15:22 MarcelsApr 16 14:31 Pitch Count Estimators
Mar 12 16:30 Appendix to THE BOOK - THE GORY DETAILS
Jan 29 09:41 NFL Overtime Idea
Jan 22 14:48 Weighting Years for NFL Player Projections
Jan 21 09:18 positional runs in pythagenpat
Oct 20 15:57 DRS: FG vs. BB-Ref
Apr 12 09:43 What if baseball was like survivor? You are eliminated ...
Nov 24 09:57 Win Attribution to offense, pitching, and fielding at the game level (prototype method)
Jul 13 10:20 How to watch great past games without spoilers