Hockey vs Penn State Open Thread

Submitted by GoBluenoser on
Going for the sweep tonight at Yost. Already up 1 zip on a tally from Dancs
MGoBoard
Hockey

stephenrjking

January 20th, 2018 at 10:57 PM ^

We're both wrong here. You are wrong because PWR doesn't add winning percentages; they total up the record against opponents held in common and express that as a percentage. There is no such thing as a winning percentage of 1.167 or 1.500. The current PWR comparison gives BG the common opponents comparison due to a winning ratio of .7500 over Ferris State and MSU. Michigan's ratio against those same opponents is currently .5833. Thus, BG gets the comparison.

I am wrong because a sweep of MSU will bring our total record against common opponents to 5-2, a ratio of .7143. This still loses the comparison. We would need to sweep MSU in a hypothetical playoff series to take the comparison unless BG loses to Ferris State in a hypothetical WCHA playoff game. 

With a few exceptions like the BG comparison (which is a bit of an anchor and is unlikely to change unless the described scenario occurs) advancing in the PWR is all about advancing in RPI. Winning will help that, obviously. 

 

Joined: 03/15/2012

MGoPoints: 20143524

mfan_in_ohio

January 20th, 2018 at 11:18 PM ^

Per USCHO.com, "Starting in the 2011-12 season, the common opponent calculation compares the sum of the winning percentages against each opponent." They provide a specific example where team A wins the common opponent comparison even though team B has the better overall record against common opponents, because adding together the winning percentages means that if one team goes 1-0 against an opponent, and another goes 5-0, those records count the same. Link: http://www.uscho.com/faq/pairwise-rankings-explanation/

Joined: 01/26/2009

MGoPoints: 18189

stephenrjking

January 20th, 2018 at 11:29 PM ^

Ok, I'm sort of wrong here. Here's collegehockeynews:

Record vs. Common Opponents: As of 2012-13, two teams' records vs. common opponents is not a straight win-loss percentage. Instead, you get a win-loss percentage against each individual common opponent, then average all those percentages together. This helps smooth out situations, for instance, where a team can beat up on the same opponent four times, while the other team in the comparison only was 1-0 against that opponent. 4-0 vs. 1-0 was a big difference. But under the new method, both go down as just 1.000.

Net effect is the same, but the comparison is still properly expressed as a sub-1.0 percentage as an average of the percentages is taken (in Michigan's case, the relevant number is a COP of .5833 against Ferris and MSU) . A change I hadn't caught (or had, and then forgot about) when it was implemented, alas. Michigan loses this comparison unless Ferris happens to play and beat BG in the WCHA playoffs.

Joined: 03/15/2012

MGoPoints: 20143524

stephenrjking

January 20th, 2018 at 10:01 PM ^

Soundly outplayed and won anyway. Kinda like last year, sort of, except this is on the heels of five really good performances, the last three of which have been wins. 

My expectation for the team remains the same: A finish outside of the tournament, some promise and some frustrations. But this is some of the promise, a hot streak at the perfect moment of the season. 

Things are rolling now. A split at OSU would be huuuuuge.

Joined: 03/15/2012

MGoPoints: 20143524

stephenrjking

January 20th, 2018 at 11:42 PM ^

Here you go:

I'll tell you exactly what happened with your embeds: The musical note emojis broke the html when you tried to embed. I deleted them in plain text mode and it works fine.

We figured this out last year when it took several tries to embed the video of Turco, Morrison, and Legg going legend in Score-O.

Joined: 03/15/2012

MGoPoints: 20143524