2009-2010 DIII Rutter Computer Rankings Thread

[nbound]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

you're still evading my main question lol who replaces midd that doesn't have just as many losses and to lesser opponents? seems to me...no one...

No evading your question, just not sure if you are worthy of a response Kidding aside
In the fan pole I had them replaced by St Kate or Stevens Point. take your pick. This weeks results will determine if Midd is back in my fan pole. By Eastern reasoning Point should move up in the standings. They lost to GAC . But they will drop out this week in my pole. In my Berry Patch Pole, based on combined average percentages(the average percentage of wins against teams at .500 or better and seasonal winning percentage), Midd was #9 and RIT was 12. Normally Midd would have been in my poll at # 9 but but they lost to Hamilton who they should have beat. Same reason eastern pollsters dropped Lake Forest after their Jan 2 loss to UMass. We'll see what this weeks results are, and how they affect the polls. "Keep smiling until then"(old line from Roy Rogers .)

 

[nyi19]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

No evading your question, just not sure if you are worthy of a response Kidding aside
In the fan pole I had them replaced by St Kate or Stevens Point. take your pick. This weeks results will determine if Midd is back in my fan pole. By Eastern reasoning Point should move up in the standings. They lost to GAC . But they will drop out this week in my pole. In my Berry Patch Pole, based on combined average percentages(the average percentage of wins against teams at .500 or better and seasonal winning percentage), Midd was #9 and RIT was 12. Normally Midd would have been in my poll at # 9 but but they lost to Hamilton who they should have beat. Same reason eastern pollsters dropped Lake Forest after their Jan 2 loss to UMass. We'll see what this weeks results are, and how they affect the polls. "Keep smiling until then"(old line from Roy Rogers .)

still not entirely sure how i feel about it......but......alas! an answer. and an arguably good one

 

[LakersFan]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

For games played through Feb 20, 2010

	Team  		Rating  RPI.Rank  	RPI
1	Plattsburgh	2.3712	1	0.6567
2	Elmira		2.0449	3	0.6282
3	Amherst		1.8826	2	0.6349
4	RIT		1.6932	5	0.6052
5	Trinity		1.4521	4	0.6065
6	Middlebury	1.3847	9	0.5849
7	Manhattanville	1.2287	10	0.5800
8	Gustavus Ad.	1.1808	6	0.5971
9	Wis-River Falls	1.0674	8	0.5885
10	Lake Forest	1.0580	7	0.5904

 

[123kidd]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

opps! Link appears to be broken.

-Google

 

[Plugger]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

For games played through Feb 20, 2010

	Team  		Rating  RPI.Rank  	RPI
1	Plattsburgh	2.3712	1	0.6567
2	Elmira		2.0449	3	0.6282
3	Amherst		1.8826	2	0.6349
4	RIT		1.6932	5	0.6052
5	Trinity		1.4521	4	0.6065
6	Middlebury	1.3847	9	0.5849
7	Manhattanville	1.2287	10	0.5800
8	Gustavus Ad.	1.1808	6	0.5971
9	Wis-River Falls	1.0674	8	0.5885
10	Lake Forest	1.0580	7	0.5904

This is very interesting, however, I am sure the westerners will not like it.

 

[gojackets]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

This is very interesting, however, I am sure the westerners will not like it.

Yes there clearly is a flaw in the ratings..not sure what it is...but there is something wrong.

Teams like Middlebury/RIT who have not beaten anyone ranked as high as they are presents a flaw. Again...they are given a lot of credit for playing the good teams even though they lose to them.

 

[Mania]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

Yes there clearly is a flaw in the ratings..not sure what it is...but there is something wrong.

Teams like Middlebury/RIT who have not beaten anyone ranked as high as they are presents a flaw. Again...they are given a lot of credit for playing the good teams even though they lose to them.

If I recall he says his ratings are based on win percentage, opponents win percentage, and opponents opponents win percentage. Since the western teams that came out east didn't have a particularly positive showing it drags all of the western teams down.

You can probably blame it mostly on Adrian.
Record vs. East:
2-6-1
Record vs. West
11-4-1

 

[Plugger]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

Yes there clearly is a flaw in the ratings..not sure what it is...but there is something wrong.

Teams like Middlebury/RIT who have not beaten anyone ranked as high as they are presents a flaw. Again...they are given a lot of credit for playing the good teams even though they lose to them.

There are many more strong teams out east which help would impact the ratings. Unfortunately for the better western teams there is a pretty good drop off out west after the top 4 or 5 teams thus making these rutter rankings more favorable to the eastern teams. There someone had to say it!!

 

[gojackets]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

There are many more strong teams out east which help would impact the ratings. Unfortunately for the better western teams there is a pretty good drop off out west after the top 4 or 5 teams thus making these rutter rankings more favorable to the eastern teams. There someone had to say it!!

So there is more of a drop off between Plattsburgh, Elmira, Amherst and the rest than there is between Gac, River Falls, Lake Forest in the west?

The next best east team (RIT) is 0-4-2 against those teams. Trinity is 0-1-1, Middlebury 0-3-2, Mville 1-4 Since the link is not working I will assume the next best west team is St. Kate They are 1-2 against the top 3 in the west, if it is Superior they are 2-3-1, St Thomas 2-3-3, Adrian 2-2 Who has the bigger drop off?

 

[LakersFan]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

I have let the systems administrators at Penn State Behrend know the website is down. It should be back on line soon.

These D-III rankings are based on the same methodology as my D-I rankings. There are a couple of factors that make ranking D-III teams a little more difficult.

1. They play fewer games than their D-I counterparts.
2. There are very few crossover games between western teams and eastern teams. The lack of information makes it difficult to rank the two sets relative to each other.

I don't know enough about the D-III schedule to know what teams have the most impact on the west is rated compared to the east.

The only way to compare east to west is to compile a list of all the games in which an east team played a west team. Look at the record and the strength of schedule, that will determine the relative strength. As stated below, Adrian hurts the entire NHCA by losing six to the ECAC west. My model keeps track of all of these games and has a better handle on strength of schedule than 99.9% of human beings, as the model is able to keep track of the "web" of games that make up the D-III schedule.

The one thing that could be wrong in my system is the data, as I scrape results from collegehockeystats.net. Sometimes, a team name changes and that screws up the entire system. To the best of my knowledge, it is currently working correctly, but if you notice any missing games, please let me know.

My rankings are very similar to USCHO.com's D-III KRACH rankings (http://www.uscho.com/rankings/?data=krach3w) , with east teams making up the top 7. Masey (http://www.masseyratings.com/rate.php?lg=chw) and
cMax (http://www.cmaxxsports.com/icehockey.php) have that same characteristic. Perhaps the east has better D-III hockey teams.

 

[gojackets]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

I have let the systems administrators at Penn State Behrend know the website is down. It should be back on line soon.

These D-III rankings are based on the same methodology as my D-I rankings. There are a couple of factors that make ranking D-III teams a little more difficult.

1. They play fewer games than their D-I counterparts.
2. There are very few crossover games between western teams and eastern teams. The lack of information makes it difficult to rank the two sets relative to each other.

I don't know enough about the D-III schedule to know what teams have the most impact on the west is rated compared to the east.

The only way to compare east to west is to compile a list of all the games in which an east team played a west team. Look at the record and the strength of schedule, that will determine the relative strength. As stated below, Adrian hurts the entire NHCA by losing six to the ECAC west. My model keeps track of all of these games and has a better handle on strength of schedule than 99.9% of human beings, as the model is able to keep track of the "web" of games that make up the D-III schedule.

The one thing that could be wrong in my system is the data, as I scrape results from collegehockeystats.net. Sometimes, a team name changes and that screws up the entire system. To the best of my knowledge, it is currently working correctly, but if you notice any missing games, please let me know.

My rankings are very similar to USCHO.com's D-III KRACH rankings (http://www.uscho.com/rankings/?data=krach3w) , with east teams making up the top 7. Masey (http://www.masseyratings.com/rate.php?lg=chw) and
cMax (http://www.cmaxxsports.com/icehockey.php) have that same characteristic. Perhaps the east has better D-III hockey teams.

Like you said...the problem is the lack of crossover. I think 9 of the 15 interregion games were played by Adrian. So one team brings the whole region down...which skews the results IMO.

 

[nbound]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

I have let the systems administrators at Penn State Behrend know the website is down. It should be back on line soon.

These D-III rankings are based on the same methodology as my D-I rankings. There are a couple of factors that make ranking D-III teams a little more difficult.

1. They play fewer games than their D-I counterparts.
2. There are very few crossover games between western teams and eastern teams. The lack of information makes it difficult to rank the two sets relative to each other.

I don't know enough about the D-III schedule to know what teams have the most impact on the west is rated compared to the east.

The only way to compare east to west is to compile a list of all the games in which an east team played a west team. Look at the record and the strength of schedule, that will determine the relative strength. As stated below, Adrian hurts the entire NHCA by losing six to the ECAC west. My model keeps track of all of these games and has a better handle on strength of schedule than 99.9% of human beings, as the model is able to keep track of the "web" of games that make up the D-III schedule.

The one thing that could be wrong in my system is the data, as I scrape results from collegehockeystats.net. Sometimes, a team name changes and that screws up the entire system. To the best of my knowledge, it is currently working correctly, but if you notice any missing games, please let me know.

My rankings are very similar to USCHO.com's D-III KRACH rankings (http://www.uscho.com/rankings/?data=krach3w) , with east teams making up the top 7. Masey (http://www.masseyratings.com/rate.php?lg=chw) and
cMax (http://www.cmaxxsports.com/icehockey.php) have that same characteristic. Perhaps the east has better D-III hockey teams.

I guess my question or statement would be if you and your computer don't know that much about the west, Are they getting a fair shake on their SOS?
Not a gripe just a question.

 

[LakersFan]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

I guess my question or statement would be if you and your computer don't know that much about the west, Are they getting a fair shake on their SOS?
Not a gripe just a question.

Ah, there is the difference. My model makes no distinction about location, it just "sees" two groups of teams that have very little crossover. The fact that it is east vs. west is a geographical distinction that we make. The same problem exists in D-I, where the WCHA plays very few non-conference games, therefore it makes it more difficult to rank them against the non-WCHA teams. What my computer knows about the west is based on the games each team has played. I don't go over the schedule game by game like the computer does, therefore the computer has a much better understanding which conferences play non-conference games with each other than I do.

I am trying to provide a ranking that compares ALL teams. The NCAA does PWR for teams in the west and east separately, recognizing that few crossover games occur. Given the information in the few "crossover" games that have been played this year, the computer models all show the east is stronger than the west. SOS is what it is based on games played, and the only way to change that is to have the top teams in the west play the top teams in east more often.

 

[LakersFan]

Re: 2009-2010 DIII Rutter Computer Rankings Thread

For games played through Feb. 28th.

Rank	Team		Rating	RPI Rk.	RPI
1	Plattsburgh	2.3831	1	0.6559
2	Elmira		2.0493	3	0.6271
3	Amherst		1.8946	2	0.6335
4	RIT		1.7034	5	0.6067
5	Trinity		1.4684	4	0.6069
6	Middlebury	1.4268	9	0.5880
7	Manhattanville	1.2370	10	0.5795
8	Gustavus Adol.	1.1989	6	0.5949
9	Wis-River Falls	1.1256	7	0.5933
10	Lake Forest	1.0877	8	0.5922