Yet Another Site that Rates HS MN Hockey Teams!
Moderators: Mitch Hawker, east hockey, karl(east)
-
- Posts: 443
- Joined: Tue Sep 06, 2005 1:45 pm
Yet Another Site that Rates HS MN Hockey Teams!
Sorry for the self promotion, but I gotta get my hockey ratings site out there somehow.
I may be a bit premature in posting and opening this up to everyone, so be aware that my whole project still a work in progress. But hopefully there's some good stuff that you'll enjoy, or maybe want to argue to me about.
https://sites.google.com/site/mnhockeystats/
Remember to take them with a grain of salt, as the season goes on, they become more accurate.
I may be a bit premature in posting and opening this up to everyone, so be aware that my whole project still a work in progress. But hopefully there's some good stuff that you'll enjoy, or maybe want to argue to me about.
https://sites.google.com/site/mnhockeystats/
Remember to take them with a grain of salt, as the season goes on, they become more accurate.
Last edited by NumberCruncher on Fri Jan 07, 2011 5:15 pm, edited 2 times in total.
That is all,
NumberCruncher
NumberCruncher
-
- Posts: 4422
- Joined: Wed Oct 10, 2007 12:22 pm
- Location: Typical homeboy from the O-Town
You're right, you need a bigger data set. In time those games will be played.
Pomeroy figures in every possession and his true measure is how much teams score independent of pace.
Delano is so high because of the same reason a statistical evaluation of the top college QBs ever said 4 of the top 6 were from the WAC. Sure, Dan Fouts and Jim McMahon were good, but their stats were so gaudy because of who they played.
A lot of teams would have Delano's record with Delano's schedule. Separation of A and AA would address this. If you want them together it will take more time with more games played between A and AA teams.
Pomeroy figures in every possession and his true measure is how much teams score independent of pace.
Delano is so high because of the same reason a statistical evaluation of the top college QBs ever said 4 of the top 6 were from the WAC. Sure, Dan Fouts and Jim McMahon were good, but their stats were so gaudy because of who they played.
A lot of teams would have Delano's record with Delano's schedule. Separation of A and AA would address this. If you want them together it will take more time with more games played between A and AA teams.
Be kind. Rewind.
-
- Posts: 6480
- Joined: Tue Jul 17, 2007 9:03 pm
- Contact:
-
- Posts: 6848
- Joined: Thu Aug 31, 2006 10:21 pm
I don't know the best way to explain it, nor do I understand it completely, but it seems to be more of a rating system (as the site says) than a ranking system.
"relative to the teams they play, this is how they do"
Not sure exactly how to use it practically though. If teams only played teams in their conference, for example, it seems it would be great to judge match ups, and probably rank.
"relative to the teams they play, this is how they do"
Not sure exactly how to use it practically though. If teams only played teams in their conference, for example, it seems it would be great to judge match ups, and probably rank.
-
- Posts: 4422
- Joined: Wed Oct 10, 2007 12:22 pm
- Location: Typical homeboy from the O-Town
Delano lost. I read your entry on the blog. It was only 3-2, so I don't know if they'll drop.
It seems your system doesn't factor the value of a win enough. Agree that 3-2 is a whole lot different than 1-2? I'm not sure you capture that.
More important than your score relative to how many you usually give up and how many you usually score is I think your goals scored in that game relative to how many they scored in that game.
Sagarin's process makes sense.
Pomeroy's work is predictive because it actually measures team strength in terms of points scored and points allowed per possession. (His blog is great BTW.) I don't see yours as predictive at all.
Thanks for providing the back years. An A team that only plays A teams seems to be way overrated. It works well when they play 10 AA teams, though.
It seems your system doesn't factor the value of a win enough. Agree that 3-2 is a whole lot different than 1-2? I'm not sure you capture that.
More important than your score relative to how many you usually give up and how many you usually score is I think your goals scored in that game relative to how many they scored in that game.
Sagarin's process makes sense.
Pomeroy's work is predictive because it actually measures team strength in terms of points scored and points allowed per possession. (His blog is great BTW.) I don't see yours as predictive at all.
Thanks for providing the back years. An A team that only plays A teams seems to be way overrated. It works well when they play 10 AA teams, though.
Be kind. Rewind.
-
- Posts: 25
- Joined: Tue Nov 30, 2010 10:47 am
-
- Posts: 464
- Joined: Sun Mar 18, 2007 12:13 am
Re: Yet Another Site that Rates HS MN Hockey Teams!
I think the one thing you could add to make these rankings more accurate is some kind of a ranking system for the overall strength of the conference and teams with in. Then use that as a multipler. For example, Lake would be one of the strongest conference/teams would have the highest ranking/highest multipler. Wright County conference/teams would be much lower.NumberCruncher wrote:Sorry for the self promotion, but I gotta get my hockey ratings site out there somehow.
I may be a bit premature in posting and opening this up to everyone, so be aware that my whole project still a work in progress. But hopefully there's some good stuff that you'll enjoy, or maybe want to argue to me about.
https://sites.google.com/site/mnhockeystats/
Remember to take them with a grain of salt, as the season goes on, they become more accurate.
Not sure how one would handle the independents though..
-
- Posts: 6848
- Joined: Thu Aug 31, 2006 10:21 pm
OK, probably the worst rating system I've ever seen. Enough said.HShockeywatcher wrote:It is not a ranking system, it is a rating system, which is different.CB00 wrote:Probably the worst ranking I've ever seen. Litchfield, Delano, Providence![]()
![]()
Too many others. Why even do this? Brutal.
I think if it also had a conference page, that would help.


-
- Posts: 283
- Joined: Sun Aug 13, 2006 2:28 am
-
- Posts: 5140
- Joined: Fri Jan 02, 2004 3:28 am
- Location: Minnesota