What's new
By:
USCHO Fan Forum

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

  • The USCHO Fan Forum has migrated to a new plaform, xenForo. Most of the function of the forum should work in familiar ways. Please note that you can switch between light and dark modes by clicking on the gear icon in the upper right of the main menu bar. We are hoping that this new platform will prove to be faster and more reliable. Please feel free to explore its features.

WCHA 2014-15: Semi-colons make things seem important

Status
Not open for further replies.
Re: WCHA 2014-15: Semi-colons make things seem important

Shirtless Guy said:
Hey math geeks, explain the par to me. Sioux sports says that Michigan tech's rpi ignores the 4 wins over UAA and UAH but the 2 lowest ranked teams they've swept are UAH and LSSU. So why doesn't it exclude the wins over LSSU?
Click to expand...

My (un)educated guess is that LSSU's opponents record is better than UAA and UAH's opponents records, making the overall numbers less weak against MTU than UAA or UAH.

I can't remember, but does Sioux Sports still show the breakdown of how many comparisons each school has won/lost?
 
Last edited:
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
My (un)educated guess is that LSSU's opponents record is better than UAA and UAH's opponents records, making the overall numbers less weak against MTU than UAA or UAH.

I can't remember, but does Sioux Sports still show the breakdown of how many comparisons each school has won/lost?
Click to expand...

Does it take into account at all when in the season the game took place?
 
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
My (un)educated guess is that LSSU's opponents record is better than UAA and UAH's opponents records, making the overall numbers less weak against MTU than UAA or UAH.

I can't remember, but does Sioux Sports still show the breakdown of how many comparisons each school has won/lost?
Click to expand...

It looks like I could be wrong, BU's RPI excludes its first 2 games of the season. Im also quite confused at what the current RPI calculation is.
 
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
My (un)educated guess is that LSSU's opponents record is better than UAA and UAH's opponents records, making the overall numbers less weak against MTU than UAA or UAH.
Click to expand...

Correct. LSSU's Opp% is pretty good, which accounts for the difference.
 
Re: WCHA 2014-15: Semi-colons make things seem important

MaizeRage said:
Correct. LSSU's Opp% is pretty good, which accounts for the difference.
Click to expand...

So trying to recreate the info SiouxSports.com why can't you do an average of the Completed Contests "Win% vs others" column to get the unadjusted "Opp Win %" from the top? I get 0.5175 vs the 0.5073 listed. Having a hard time confirming whether LSSU games should be excluded if I can't even get over this first hump.
 
Re: WCHA 2014-15: Semi-colons make things seem important

As far as I know from tinkering with the PWR calculations last year and my experiences with the PWR/RPI for the last 13 years is: it doesn't matter when the teams played each other, the RPI is a pure "numbers" calculation. And thus being a pure "numbers" calculation with weighted bonuses, the LSSU wins improve Tech's RPI where UAA and UAH's wins decrease their RPI.

Up until a few weeks ago, NMU's win against UAH was excluded. But our sweep of Wisconsin has never been excluded.

It's numbers. Sioux Sports isn't out to screw Tech. USCHO's RPI calculation isn't screwing Tech. The NCAA didn't have it out for Tech when the fine tuning of the win bonuses were tweaked for the 2013-14 season. And I'd even go so far as to say that come 11pm on March 21st, the wins against UAA and UAH will help their RPI and this won't be an issue.


*edit* Christ, I'm the one defending the da** PWR/RPI and my team has been on the a<!-- * -->ss end of a few number calculations in the last decade and a half. 2003 NCAA tournament is what really irks me where Mankato and St. Cloud State got in but NMU didn't still p*sses me off.
 
Last edited:
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
As far as I know from tinkering with the PWR calculations last year and my experiences with the PWR/RPI for the last 13 years is: it doesn't matter when the teams played each other, the RPI is a pure "numbers" calculation. And thus being a pure "numbers" calculation with weighted bonuses, the LSSU wins improve Tech's RPI where UAA and UAH's wins decrease their RPI.

Up until a few weeks ago, NMU's win against UAH was excluded. But our sweep of Wisconsin has never been excluded.

It's numbers. Sioux Sports isn't out to screw Tech. USCHO's RPI calculation isn't screwing Tech. The NCAA didn't have it out for Tech when the fine tuning of the win bonuses were tweaked for the 2013-14 season. And I'd even go so far as to say that come 11pm on March 21st, the wins against UAA and UAH will help their RPI and this won't be an issue.


*edit* Christ, I'm the one defending the da** PWR/RPI and my team has been on the a<!-- * -->ss end of a few number calculations in the last decade and a half. 2003 NCAA tournament is what really irks me where Mankato and St. Cloud State got in but NMU didn't still p*sses me off.
Click to expand...

I never said that, after doing more research into other teams, I noticed BU had its first 2 games excluded. I assume everything is fine, but now I'm to the point of wanting to understand how to calculate to see what effect excluding a win over LSSU has. The fact that my previous understanding and SiouxSports details doesn't allow me to calculate it myself is frustrating. I can't even get close to their numbers for Opp W% or Opp Opp W%.

My initial thought was that they start at the beginning and add each game and see if it positively or negatively impacts the RPI and exclude the negatives, after further review its obvious they go through and remove each game and see what effect it has.
 
Re: WCHA 2014-15: Semi-colons make things seem important

Shirtless Guy said:
I never said that, after doing more research into other teams, I noticed BU had its first 2 games excluded. I assume everything is fine, but now I'm to the point of wanting to understand how to calculate to see what effect excluding a win over LSSU has. The fact that my previous understanding and SiouxSports details doesn't allow me to calculate it myself is frustrating. I can't even get close to their numbers for Opp W% or Opp Opp W%.

My initial thought was that they start at the beginning and add each game and see if it positively or negatively impacts the RPI and exclude the negatives, after further review its obvious they go through and remove each game and see what effect it has.
Click to expand...

Flagdude used to have a simulator for the PWR at wa2aea.com. It looks like he has issues with his website right now. I have it saved to a jump drive somewhere. I'll look for it and try to re-post it this weekend. You *should* be able to tinker with it, eliminate the LSSU games, and run the RPI simulation.
 
Re: WCHA 2014-15: Semi-colons make things seem important

Shirtless Guy said:
I never said that, ...
Click to expand...

Shirtless Guy said:
Hey guess what guys, I think MTU is getting screwed in the PWR ranking simply because they played LSSU first.
Depending on rounding errors from the info provided by http://siouxsports.com/hockey/rankings/rpidetails.php?teamid=4
MTU's RPI should be somewhere around 0.0179 points higher when listing an RPI as 0.####, not ##.## as SiouxSports.com does on their PWR page.
Michigan Tech would be likely around 4th in the PWR instead of 6th. Not sure how those changes adjust Quality Win Bonuses or anything else...
Click to expand...

;)

Anyway, here's the link to Flagdude's RPI simulator:

http://www.nmubaseball.net/uscho/NCAARPIPWR.zip
Here's a sample Excel file that shows the format you need for it to build upon:
http://www.nmubaseball.net/uscho/NCAARPIPWRHockey20131208.csv

I've tinkered with it for NCAA, and used it to see what it would do for the AHL.


*edit* I haven't gotten down to the nuts and bolts of the simulation program, so the weighted bonuses may not be calculated right.
 
Last edited:
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
Flagdude used to have a simulator for the PWR at wa2aea.com. It looks like he has issues with his website right now. I have it saved to a jump drive somewhere. I'll look for it and try to re-post it this weekend. You *should* be able to tinker with it, eliminate the LSSU games, and run the RPI simulation.
Click to expand...

Yeah, I found the old slack.net one but it hasn't been updated since 2013 I think?
 
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
;)

Anyway, here's the link to Flagdude's RPI simulator:

http://www.nmubaseball.net/uscho/NCAARPIPWR.zip
Here's a sample Excel file that shows the format you need for it to build upon:
http://www.nmubaseball.net/uscho/NCAARPIPWRHockey20131208.csv

I've tinkered with it for NCAA, and used it to see what it would do for the AHL.


*edit* I haven't gotten down to the nuts and bolts of the simulation program, so the weighted bonuses may not be calculated right.
Click to expand...

ok, let me rephrased, upon further inspection I have backed off that statement ;) when I posted that, I had not checked other teams and noticed the BU exclusion, and still can't get my numbers to match close enough to SIouxSports to verify anything. I'm just really confused why the average of the "Win% vs others" column doesn't equal the unadjusted "Opp Win%" on this page. Even when I verify winning percentages of the teams MTU has played, the rounding errors of using their 4 decimal numbers doesn't fix that.
 
Re: WCHA 2014-15: Semi-colons make things seem important

Ok, I figured out how SiouxSports calculates a teams Opp Win%.

you manipulate each opponent's Win% vs others by multiplying by the Weighted GP value (1.2 for road, 1 for neutral, 0.8 for home).
Sum each of those numbers and divide by the sum of the Weighted GP.

I get the same value as listed on the Sioux Sports website.

The same calc does not work for Opp Opp Win%
 
Re: WCHA 2014-15: Semi-colons make things seem important

aparch said:
;)

Anyway, here's the link to Flagdude's RPI simulator:

http://www.nmubaseball.net/uscho/NCAARPIPWR.zip
Here's a sample Excel file that shows the format you need for it to build upon:
http://www.nmubaseball.net/uscho/NCAARPIPWRHockey20131208.csv

I've tinkered with it for NCAA, and used it to see what it would do for the AHL.


*edit* I haven't gotten down to the nuts and bolts of the simulation program, so the weighted bonuses may not be calculated right.
Click to expand...

Thanks for the links!

GFM
 
Re: WCHA 2014-15: Semi-colons make things seem important

I know that we have a WCHA TV thread, but I wanted to let everyone know that there are half-season packages for $70 available immediately. Matt Hodson and I have had a good dialogue about the service's problems, and things have gotten a lot better.

The biggest thing from our email exchange today is that there will be HD-quality feeds from eight of the ten league facilities, and the league is working to get the other two up ASAP. Akamai has done a great job in helping the league schools to get up to snuff. Also, they're doing their best to gate the game-switcher up Very Soon Now.

I think that 2015 is going to bring better service from WCHA TV.

GFM
 
Re: WCHA 2014-15: Semi-colons make things seem important

gfmorris said:
I know that we have a WCHA TV thread, but I wanted to let everyone know that there are half-season packages for $70 available immediately. Matt Hodson and I have had a good dialogue about the service's problems, and things have gotten a lot better.

The biggest thing from our email exchange today is that there will be HD-quality feeds from eight of the ten league facilities, and the league is working to get the other two up ASAP. Akamai has done a great job in helping the league schools to get up to snuff. Also, they're doing their best to gate the game-switcher up Very Soon Now.

I think that 2015 is going to bring better service from WCHA TV.

GFM
Click to expand...
which 8?
 
Re: WCHA 2014-15: Semi-colons make things seem important

The bottom line with the RPI & PWR is that it means absolutely nothing until the last game of the season is finished.
 
Re: WCHA 2014-15: Semi-colons make things seem important

gfmorris said:
Matt didn't say in his email, and I didn't ask. I'm betting that Fairbanks will be one. I'd expect Lake next and maybe UAH. I know that UAH was getting more bandwidth into the rink.

GFM
Click to expand...

Fairbanks HD should be up and running....GF was the UAH-AK series not???
 
Status
Not open for further replies.
Back
Top