I did a very basic preliminary Analysis, just for the hell of it.
Bold indicates a game in which the outhitting team ending up losing the hockey game.
G1: W3-2 Outhit by SJ 14-12
G2: W5-3 Outhit by PHIL 24-14
G3: L4-2 Outhit by DET 19-13
G4: L2-0 Outhit by MIN 10-4
G5: L 5-2 Outhit by VAN 34-17
G6: L 4-1 Outhit VAN 14-9
G7: W4-2 Outhit PHX 18-17
G8: L 4-1 Outhit by CGY 22-9
G9: L 4-2 Outhit COL 20-4
G10: W 5-4 Outhit MIN 15-11
G11: L 4-1 Outhit by LA 21-17
G12: W 3-2 Outhit by ANA 15-11
G13: L 2-1 TIED DET 13-13
G14: L 4-1 Outhit NSH 10-8
G15: L 5-2 Outhit MIN 10-9
G16: L 4-3 Outhit COL 12-6
G17: W 4-2 Outhit by CAL 25-12
G18: W 1-0 Outhit by VAN 15-13
G19: L 4-2 Outhit by MIN 13-12
G20: L 3-1 Outhit CAL 20-16
G21: W 5-4 TIED VAN 10-10
G22: L 3-2 Outhit Col 9-8
G23: W 3-2 Outhit by CHI 13-9
G24: W 3-1 TIED CLB 12-12
G25: L 4-2 Outhit COL 14-11
G26: W5-1 Outhit by ANA 22-8
G27: W4-0 Outhit by ANA 17-8
End result: 16 games in which the outhitting team lost the hockey game. 3 games resulted in ties in the hits battle, so should be filtered out if anyone chooses to take this data into excel. 8 games in which the outhitting team won the hockey game.
Very preliminary (and quite possibly ignoring certain details), but indicates to some extent, hits aren't a good predictor of winning hockey games. Perhaps even a negative correlation relationship exists. Sample size is closing near 30, drawing out the luck element. I will update this at the end of the year, and perhaps look at more teams.
Next time the masses argue that losing Matt Greene leaves the Edmonton Oilers soft on the back end, perhaps the counter-argument should then be, so what?
Posts
2 comments:
The problem with hits + giveaways + takeaways is the same as "baseball errors" - they are reliant on human judgment.
Yes, within the same rink the judges are the same but often I read the final game stats and think... "that's not right."
From your first analysis, the answer is probably no - "out hitting does not win hockey" games.
The next question to ask is - what does? (ignoring the obvious one - scoring more goals than the other team).
That's a good point regarding human judgement - but as long as the stats keepers aren't homers, I think if you gather data from all teams, and all 82 games, the sample size would be large enough to eliminate bias/error/luck.
As for what does, that's another good question. I'm puzzled.
Post a Comment