2013-14 KRACH RANKINGS - NOW ON MGHCA WEBSITE

Discussion of Minnesota Girls High School Hockey

Moderators: Mitch Hawker, east hockey, karl(east)

ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

End-of-Sections ranking coming soon - hopefully yet tonight

Post by ghshockeyfan »

End-of-Sections ranking coming soon - hopefully yet tonight...
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

Updated 2/18 10PM 23490 (End of Sections)

Post by ghshockeyfan »

Updated 2/18 10PM 23490 (End of Sections)
Last edited by ghshockeyfan on Sun Feb 23, 2014 10:14 am, edited 1 time in total.
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

Re: Updated 2/18 10PM 23490 (End of Sections)

Post by ghshockeyfan »

KRACH post-sections (even after some "upsets" said the following should advance):

AA SECTION - ACTUAL (section rank - KRACH if different)
1 - Lakeville North
2 - Eden Prairie
3 - Burnsville
4 - Hill-Murray
5 - Mounds View (3) - Blaine
6 - Benilde-St. Margaret's
7 - Andover
8 - Roseau

A SECTION - ACTUAL (section rank - KRACH if different)
1 - Red Wing
2 - Litchfield/Dassel-Cokato (2) - Mankato East
3 - New Ulm
4 - St Paul United (2) - South St Paul
5 - Blake
6 - Alexandria
7 - Proctor/Hermantown
8 - East Grand Forks (2) - Thief River Falls


SEEDING COMPARISON:

AA (KRACH SEED) - TEAM
1 (1) - Benilde-St. Margaret's
2 (2) - Hill-Murray
3 (3) - Eden Prairie
4 (4) - Lakeville North
5 (7) - Mounds View
6 (6) - Andover
7 (5) - Burnsville
8 (8 ) - Roseau

A (KRACH SEED) - TEAM
1 (1) - Thief River Falls
2 (2) - Blake
3 (3) - Red Wing
4 (5) - Proctor/Hermantown
5 (4) - St Paul United

6 (6) - New Ulm
7 (7) - Alexandria
8 (8 ) - Litchfield/Dassel-Cokato
MNHockeyFan
Posts: 7260
Joined: Tue Nov 07, 2006 10:28 pm

Post by MNHockeyFan »

Would like to publicly thank ghshockeyfan for managing and posting the KRACH rankings once again this year. He's obviously a very busy guy, so taking the time out to do this is very much appreciated, especially when you consider he works full time and is coaching on the boys' side.
OsMetroDad
Posts: 69
Joined: Tue Sep 17, 2013 1:42 am

Post by OsMetroDad »

Not sure if this is even possible in your version of KRACH, but you might want to try using a parameter to down-weight games that you feel that need to be de-emphasized ("for social reasons") instead of corrupting your data set by changing actual scores. That might solve your SOS issue and maintain the overall numerical integrity of your data set and its results. Another way of saying this is it might keep the relative positions of the teams intact based on your full solution.
Bulldog3489
Posts: 97
Joined: Mon Jan 28, 2013 3:52 pm

Post by Bulldog3489 »

OS - good first step in using words like integrity and corruption. Now apply.
MN_Bowhunter
Posts: 220
Joined: Wed Jul 04, 2012 10:08 am

Post by MN_Bowhunter »

OsMetroDad wrote: ("for social reasons")
http://www.youtube.com/watch?v=BNsrK6P9QvI

It's quite clear most if not all members of the forum are ready to let this go.

http://www.youtube.com/watch?v=Jj4nJ1YEAp4
OsMetroDad
Posts: 69
Joined: Tue Sep 17, 2013 1:42 am

Post by OsMetroDad »

Bulldog, You can be humerous in a good way when you want to, its a good first start. For the purposes of testing commerical software and reporting its results, I'll just agree to disagree with you with regard to the content of the data set, numerically if there was a game played with two teams, a couple of refs on the ice and a signed score sheet, its a valid numerical entry regardless of the perceived social implication or desires of some who spend way too much time on this forum. just kidding ;-) :wink: The computer that I use as an email server doesn't have sound on it so I'll have re-play that Star Trek clip on one with sound. A big fan of Cpt. Picard.
Nevertoomuchhockey
Posts: 1138
Joined: Thu Oct 03, 2013 2:59 pm

Post by Nevertoomuchhockey »

Ok here we go, Ill bite. Stats mean nothing if you can't use them to motivate your team or positively influence development in the W column. Having said that, OS... How's that working out for you? Unless you are so out of touch that you think games forfeited for cheating are the equivalent of "social reasons," you can recognize that OS needs more than computer models. Quit rationalizing how your computer expertise may be compromised by the TRUTH. Frankly, we are bored.
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

Updated 2/23 3AM 24600 (2013-14 FINAL)

Post by ghshockeyfan »

Updated 2/23 3AM 24600 (2013-14 FINAL)

Most Recent Ranking Links:
==========================================
OVERALL:
http://www.bgoski.com/KRACH_OA.htm
CLASS AA:
http://www.bgoski.com/KRACH_AA.htm
CLASS A:
http://www.bgoski.com/KRACH_A.htm
SECTION:
http://www.bgoski.com/KRACH_SEC.htm
SOS (Strength-Of-Schedule) is at:
http://www.bgoski.com/KRACH_SOS.htm
==========================================

Past GIRLS KRACH rankings on USHSHO thread are at:
http://www.ushsho.com/forums/viewtopic. ... sc&start=0
Main GIRLS KRACH website: http://www.bgoski.com/krach
Added KRACH info./details (MN Hockey Hub) at: http://www.bgoski.com
Main BOYS KRACH website: http://www.bgoski.com/b
USHSHO BOYS KRACH thread: http://www.ushsho.com/forums/viewtopic.php?t=29709
Last edited by ghshockeyfan on Sun Feb 23, 2014 11:16 am, edited 1 time in total.
MNHockeyFan
Posts: 7260
Joined: Tue Nov 07, 2006 10:28 pm

Re: Updated 2/23 3AM 24600 (End of State Tournament)

Post by MNHockeyFan »

ghshockeyfan wrote:Updated 2/23 3AM 24600 (End of State Tournament)

OVERALL:
http://www.bgoski.com/KRACH_OA.htm
CLASS AA:
http://www.bgoski.com/KRACH_AA.htm
CLASS A:
http://www.bgoski.com/KRACH_A.htm
I guess I'm a little surprised that EGF finished higher than Blake, and that BSM finished higher than Hill-Murray, but I do realize that KRACH bases its rankings on each team's whole body of work throughout the entire season, so it's completely understandable.
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

Re: Updated 2/23 3AM 24600 (2013-14 FINAL)

Post by ghshockeyfan »

MNHockeyFan wrote:
ghshockeyfan wrote:Updated 2/23 3AM 24600 (2013-14 FINAL)

OVERALL:
http://www.bgoski.com/KRACH_OA.htm
CLASS AA:
http://www.bgoski.com/KRACH_AA.htm
CLASS A:
http://www.bgoski.com/KRACH_A.htm
I guess I'm a little surprised that EGF finished higher than Blake, and that BSM finished higher than Hill-Murray, but I do realize that KRACH bases its rankings on each team's whole body of work throughout the entire season, so it's completely understandable.
Me too - I'm trying to think if this happened much over the last 10 years + that I've been running KRACH and I think it's pretty rare but my memory isn't what it once was either!
OsMetroDad
Posts: 69
Joined: Tue Sep 17, 2013 1:42 am

Post by OsMetroDad »

MN & GS Hockey Fan...
I guess I'm a little surprised that EGF finished higher than Blake, and that BSM finished higher than Hill-Murray
You need to adjust your weighting parameters in the KRACH software, if they exist? Regardless, the plus/minus standard deviation of the overall solution is large enough that there ther is no statistical difference between any one contigous group of 4 or 5 teams sequentially. What this is saying that although the KRACH software spits out an ordered solutuion, if it did the proper post-adjustment test mean two means, i.e. where is the dividing line where one team is truely a statiscally better team, it would report a tie between BSM & HM as far as ratings go, which is what you are essentially seeing, but it is just not being reported because of its design limitations.
MNHockeyFan
Posts: 7260
Joined: Tue Nov 07, 2006 10:28 pm

Post by MNHockeyFan »

OsMetroDad wrote:You need to adjust your weighting parameters in the KRACH software, if they exist?
I suppose you can adjust your weighting parameters any way you wish to produce a desired outcome, when that outcome is already known. But how do you know which weightings will produce the most accurate predictions looking forward, rather than adjusting them retroactively? With regard to KRACH, as far as I know it's a uniform (fixed) system that's remained the same for years, for everyone that's used it, including to this day at the D1 level.

http://www.uscho.com/rankings/krach/d-i-women/
OsMetroDad
Posts: 69
Joined: Tue Sep 17, 2013 1:42 am

Post by OsMetroDad »

The best accurraccy of a solution is usually determined by minimizing the function to achieve the smallest sum of the squares of the observation residuals. In the LSQRANK case, I designed in flexibility to adjust the importance of the weighting components. To find the best solution, you should run multiple scenarios and then plot your minimum function score against the parameter(s). If you are dealing with more than one parameter, you will need to use a "response surface" approach to determine the best fitting values that minimizes the function. The best referance that I know of that might help you determine this "Empircal Model-Building and Response Surfaces" by George E.P. Box & Norman R. Draper.
Sparlimb
Posts: 2491
Joined: Tue Feb 05, 2002 7:11 am

Re: Updated 2/23 3AM 24600 (2013-14 FINAL)

Post by Sparlimb »

ghshockeyfan wrote:
MNHockeyFan wrote:
ghshockeyfan wrote:Updated 2/23 3AM 24600 (2013-14 FINAL)

OVERALL:
http://www.bgoski.com/KRACH_OA.htm
CLASS AA:
http://www.bgoski.com/KRACH_AA.htm
CLASS A:
http://www.bgoski.com/KRACH_A.htm
I guess I'm a little surprised that EGF finished higher than Blake, and that BSM finished higher than Hill-Murray, but I do realize that KRACH bases its rankings on each team's whole body of work throughout the entire season, so it's completely understandable.
Me too - I'm trying to think if this happened much over the last 10 years + that I've been running KRACH and I think it's pretty rare but my memory isn't what it once was either!
Didn't we have this discussion before where you said HM gets subtractions in the rankings for playing Class A teams that happen to be in their conference so they can't avoid? I'd like to see the KRACH ratings after assuming all teams are the same class (since being in A or AA doesn't change how good you are). That said, you may have changed that years ago and I'm behind...
Bleedinred
Posts: 22
Joined: Wed Oct 05, 2011 9:41 pm

Post by Bleedinred »

OS...the goal is not to generate more precise data. Any data will do to get the peanut gallery chattering...and that is exactly the point of this forum. The good news is that the game is played on the ice and not in a statistical modeling software. Enjoy the game and the chatter.
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

Post by ghshockeyfan »

QRF ranking did some changing to how class a vs aa opponents were handled some time back. KRACH has never introduced any other parameters to weight aspects of the game differently. Although, I'd guess the conference wins vs class a opponents don't help hm as much as they would if they played in a class aa stronger conf. That's probably true regardless of algorithm unless the algorithm intentionally addresses this through some weighting parameter. Bottom line is even I don't get too concerned about this anymore as it truly is played on the ice and congrats to hm for taking first there where it really counts!
ghshockeyfan
Posts: 6132
Joined: Thu Jun 05, 2003 2:33 pm
Location: Inver Grove Heights, MN
Contact:

KRACH RANKINGS NOW ON MGHCA WEBSITE

Post by ghshockeyfan »

The KRACH rankings for MN Girls' HS Hockey are now available on the MGHCA website.

http://www.bgoski.com (forward link)

http://www.mghca.com/page/show/427504-krach-rankings (direct link)

The final rankings for the past 12 seasons will also be archived there in addition to the most current season's rankings, once ready for release (later this month).
ghshockeyfan wrote:Main website: http://www.bgoski.com/krach
See MN Hockey Hub for rankings and more info.:
http://www.mghca.com/page/show/427504-krach-rankings
1) This only counts games vs. MSHSL varsity teams.
2) This only uses scores entered on the Girls' Hockey Hub/MGHCA website:
http://www.mghca.com
To enter a missing score contact your team rep. Each team should have a person responsible for updating this information (coach, parent, manager, etc.) - you can also email: scores@tstmedia.com - see: http://www.mngirlshockeyhub.com/page/sh ... mit-scores (they really want a scoresheet image though)
3) For past years' ranking and further explanation:
http://www.ushsho.com/forums/viewtopic.php?t=29597
http://www.ushsho.com/forums/viewtopic.php?t=27266
http://www.ushsho.com/forums/viewtopic.php?t=24859
http://www.ushsho.com/forums/viewtopic.php?t=21598
http://www.ushsho.com/forums/viewtopic.php?t=18448
4) Most Recent Ranking Links:
==========================================
OVERALL:
http://www.bgoski.com/KRACH_OA.htm
CLASS AA:
http://www.bgoski.com/KRACH_AA.htm
CLASS A:
http://www.bgoski.com/KRACH_A.htm
SECTION:
http://www.bgoski.com/KRACH_SEC.htm
SOS (Strength-Of-Schedule) is at:
http://www.bgoski.com/KRACH_SOS.htm
==========================================
5) Past Ranking Links for 2013-14:
==========================================
OVERALL:
http://www.bgoski.com/KRACH_OA_20140223.htm (2013-14 FINAL)
http://www.bgoski.com/KRACH_OA_20130302.htm (2012-13 FINAL)
http://www.bgoski.com/KRACH_OA_20120226.htm (2011-12 FINAL)
http://www.bgoski.com/KRACH_OA_20110227.htm (2010-11 FINAL)
http://www.bgoski.com/KRACH_OA_20100228.htm (2009-10 FINAL)
http://www.bgoski.com/KRACH_OA_2008-9_FINAL.htm
***2007-8_FINAL (No KRACH)
http://www.bgoski.com/KRACH_OA_C_S_2006-7_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2005-6_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2004-5_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2003-4_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2002-3_FINAL.htm
pre-KRACH OA rankings:
http://www.minnhock.com/ps2003g.htm (2002-3 FINAL)
http://www.minnhock.com/ps2002g.htm (2001-2 FINAL)
http://www.minnhock.com/ps2001g.htm (2000-1 FINAL)
http://www.minnhock.com/ps2000g.htm (1999-2000 FINAL)
http://www.minnhock.com/ps1999g.htm (1998-9 FINAL)
http://www.minnhock.com/archives.htm (General Pre-KRACH Archives)
CLASS AA:
http://www.bgoski.com/KRACH_AA_20140223.htm (2013-14 FINAL)
http://www.bgoski.com/KRACH_AA_20130302.htm (2012-13 FINAL)
http://www.bgoski.com/KRACH_AA_20120226.htm (2011-12 FINAL)
http://www.bgoski.com/KRACH_AA_20110227.htm (2010-11 FINAL)
http://www.bgoski.com/KRACH_AA_20100228.htm (2009-10 FINAL)
http://www.bgoski.com/KRACH_AA_2008-9_FINAL.htm
***2007-8_FINAL (No KRACH) - instead see http://www.minnesota-scores.net/classqr ... 10&class=3
http://www.bgoski.com/KRACH_OA_C_S_2006-7_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2005-6_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2004-5_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2003-4_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2002-3_FINAL.htm
***No pre-KRACH Class rankings
CLASS A:
http://www.bgoski.com/KRACH_A_20140223.htm (2013-14 FINAL)
http://www.bgoski.com/KRACH_A_20130302.htm (2012-13 FINAL)
http://www.bgoski.com/KRACH_A_20120226.htm (2011-12 FINAL)
http://www.bgoski.com/KRACH_A_20110227.htm (2010-11 FINAL)
http://www.bgoski.com/KRACH_A_20100228.htm (2009-10 FINAL)
http://www.bgoski.com/KRACH_A_2008-9_FINAL.htm
***2007-8_FINAL (No KRACH) - instead see http://www.minnesota-scores.net/classqr ... 10&class=2
http://www.bgoski.com/KRACH_OA_C_S_2006-7_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2005-6_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2004-5_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_2003-4_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2002-3_FINAL.htm
***No pre-KRACH Class rankings
SECTION:
http://www.bgoski.com/KRACH_SEC_20140223.htm (2013-14 FINAL)
http://www.bgoski.com/KRACH_SEC_20130302.htm (2012-13 FINAL)
http://www.bgoski.com/KRACH_SEC_20120226.htm (2011-12 FINAL)
http://www.bgoski.com/KRACH_SEC_20110227.htm (2010-11 FINAL)
http://www.bgoski.com/KRACH_SEC_20100228.htm (2009-10 FINAL)
http://www.bgoski.com/KRACH_SEC_2008-9_FINAL.htm
***2007-8_FINAL (No KRACH)
http://www.bgoski.com/KRACH_OA_C_S_2006-7_FINAL.htm
http://www.bgoski.com/KRACH_OA_C_S_2005-6_FINAL.htm
***2004-5_FINAL (No KRACH by Section)
***2003-4_FINAL (No KRACH by Section)
http://www.bgoski.com/KRACH_OA_C_S_2002-3_FINAL.htm
***No pre-KRACH Section rankings
SOS (Strength-Of-Schedule) is at:
http://www.bgoski.com/KRACH_SOS_20140223.htm (2013-14 FINAL)
http://www.bgoski.com/KRACH_SOS_20130302.htm (2012-13 FINAL)
http://www.bgoski.com/KRACH_SOS_20120226.htm (2011-12 FINAL)
http://www.bgoski.com/KRACH_SOS_20110227.htm (2010-11 FINAL)
http://www.bgoski.com/KRACH_SOS_20100228.htm (2009-10 FINAL)
http://www.bgoski.com/KRACH_SOS_2008-9_FINAL.htm
***2007-8_FINAL (No KRACH)
***2006-7_FINAL (No KRACH SOS)
http://www.bgoski.com/KRACH_SOS_2005-6_FINAL.htm
***2004-5_FINAL (No KRACH SOS)
***2003-4_FINAL (No KRACH SOS)
***2002-3_FINAL (No KRACH SOS)
***No pre-KRACH SOS rankings
==========================================
Post Reply