A couple of articles on (home) recording bias and NHL Stats, including blocked shots now that OLG has offered this prop pretty steadily since after the Olympic Break
http://www.leftwinglock.com/articles/in ... -category/
http://www.arcticicehockey.com/2011/4/5 ... rding-bias
http://www.hockeyprospectus.com/puck/ar ... icleid=351
Home recording bias
-
ProlinePlayer
- Site Admin
- Posts: 2076
- Joined: Sun Mar 31, 2013 6:33 am
Re: Home recording bias
I think this is a very interesting area which needs a lot more study.
At present I do not account for a scoring bias in any of my models but I am planning to start trying to include this.
The links to the varying articles you've posted on this do not completely convince me that it is real in all cases. For example in the first link the writer has a graph showing the bias by stadium for blocked shots. We're supposed to accept that in half the stadiums the stats keeper shows a bias towards the home team (understandable) but in the other half of stadiums the stats keeper shows a bias towards the visit team (why?).
I don't know but the first thing that comes to my mind is not that the graph is a proof of bias but rather that the graph shows the random nature of the stat and that in the end all we have is a graph of random noise.
But on the other hand the results for shots on goal in the 3rd link are much more compelling as proof of recording bias.
One of the problems in analyzing this would be in seperating the different forms of bias. Is the stats recorder favouring the home team? Or is it just a case of more/less of a event? eg: more sogs for both sides or just for the home team?
PLP
At present I do not account for a scoring bias in any of my models but I am planning to start trying to include this.
The links to the varying articles you've posted on this do not completely convince me that it is real in all cases. For example in the first link the writer has a graph showing the bias by stadium for blocked shots. We're supposed to accept that in half the stadiums the stats keeper shows a bias towards the home team (understandable) but in the other half of stadiums the stats keeper shows a bias towards the visit team (why?).
I don't know but the first thing that comes to my mind is not that the graph is a proof of bias but rather that the graph shows the random nature of the stat and that in the end all we have is a graph of random noise.
But on the other hand the results for shots on goal in the 3rd link are much more compelling as proof of recording bias.
One of the problems in analyzing this would be in seperating the different forms of bias. Is the stats recorder favouring the home team? Or is it just a case of more/less of a event? eg: more sogs for both sides or just for the home team?
PLP