## Tuesday, May 12, 2009

### Probability Taken Personally

There is a theory in the blogospehere that pucks hitting goalposts are just unlucky. Everything in the game of hockey can be explained by mathematics and statistical analysis, but goalposts, well they're just the hockey gods spittin' in your eye. Some of the arguments for this theory can be found here and here. I don't buy it. Basically, what I understand from these articles, is that teams success or lack thereof correlates well to the number of posts the team hits. Those teams that hit a tonne of posts are unlucky - you can expect them to turn things around. Those that have a lot of posts hit behind they're goalie - well, don't bet the farm on them. It's a ridiculous argument and some of the sharpest writers on the net subscribe to it wholeheartedly. L. Ron Hubbard must be involved somehow. Can you take one element of the game and contribute it to luck? Yet ignore that every goal for or against in every game would not be suject to luck as well? How many goals are scored on tipped shots? How many go in off defenders sticks or skates? Might there be a few that don't really end up where the shooter was intending? They count anyway right? Has a goalie ever had a shot hit him that he didn't see? Has a linesman ever gotten involved in a play? Take 1000 darts and throw them at triple 20. Are the ones that bounce off the wire the only shots where luck was invloved? How about 1000 three foot puts? The ones that miss the hole are misses, but the ones that rim out are bad luck? Or could it be that rimming out means you missed the cup by hair wider than the radius of the ball? A hair you say....wow, that's some shitty luck.
There's a very good reason why posts aren't counted as shots on goal - they missed the net. There is litterally a hair width difference between a puck nicking the post or going wide. That same hair could mean the difference between the puck hitting the post and going in, or hitting both posts and staying out. How does one have more impact on a game than the other?
The NHL records goalposts in it's game logs and that has allowed data to be mined out and tabulated. I think it's unfortunate. I'd like to see how many teams miss the net by millimeters. Is there a significant difference in the amount of luck involved relative to the shot that hits the post? How about lucky goals? Has a team ever hit the post and scored a lucky goal in the same game, and do they negate each other?
Anyone see Chuck Kobasew's goal against Cam Ward last night? Patrice Bergeron's pass missed the defenders skate by an inch, Kobasew flubbed what looked like an attempt to tip the puck over Ward, the puck hit the heel of his stick and slid in along the ice nicking the post on the way in.....goal - no luck accounted for. That puck hits the post and all of a sudden, there's a luck event, let's tabulate it. Ridiculous.