Didn't BC split with BU whonlost 2-0 this week to NU? Doesn't the computer rankings take strength of schedule and opponents record and strength of schedule into consideration???
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.
2011-12 DI Rutter Computer Rankings
- Thread starter LakersFan
- Start date
Didn't BC split with BU whonlost 2-0 this week to NU? Doesn't the computer rankings take strength of schedule and opponents record and strength of schedule into consideration???
granddaddyscout
Registered User
Re: 2011-12 DI Rutter Computer Rankings
Not sure what your point is. My point is that the strength on schedule for the teams in the WCHA is beyond what the computers factor in; also with the limited amount of crossover play it is very hard to accurately reflect the difference. Look at MSU 1-9 in the WCHA with a win over Mercyhurst who just beat Cornell. I am not saying that MSU is better than Cornell or MC, but they clearly are pretty capable of playing a good game and UW, UM, UND, etc... play them 4 times every year. They are the 2nd weakest team in the WCHA.
Not sure what your point is. My point is that the strength on schedule for the teams in the WCHA is beyond what the computers factor in; also with the limited amount of crossover play it is very hard to accurately reflect the difference. Look at MSU 1-9 in the WCHA with a win over Mercyhurst who just beat Cornell. I am not saying that MSU is better than Cornell or MC, but they clearly are pretty capable of playing a good game and UW, UM, UND, etc... play them 4 times every year. They are the 2nd weakest team in the WCHA.
Re: 2011-12 DI Rutter Computer Rankings
The NCAA selection criteria consider strength-of-schedule, but not in a way that is accurate. The way the NCAA does it they just take a weighted average of your record, your opponents' records, and your opponents' opponents' record (known as the ratings percentage index, the RPI, not to be confused with the ECAC school)
The problem with the RPI is that in theory, losing to a perfect team or beating a perfectly awful team should provide no new information about the quality of your school, but both results can affect your RPI (though the NCAA does have an "adjustment" that drops wins that lower your RPI from consideration). There are actually no perfect teams and no perfectly awful teams, but this problem still exists to some extent. Another way of looking at this:if the number of conference games gets large compared to the number of nonconference games, then the RPI is largely determined by your conference schedule, and a team that's 5th in a strong conference has no realistic chance of achieving a better RPI than the best team in the weaker conference.The KRACH and Rutter rankings are actually based on estimating statistical models, and they do not suffer from this fallacy.
Why does the NCAA not change this? The main reason I'd say is they have a strong institutional history with the RPI. Another reason is there's pressure from weaker conferences to maintain the status quo, since they are the ones hurt by moving to a more accurate system. (Of course, the RPI is still a better system than basing 1/3 of the selection of your national championship participants based on 5 computer rankings that are not made public, but that's off topic)
I had a conversation with a coach last year from Eastern school. I was interested that the coach offered the perception that a ranking like KRACH is fundamentally unfair because there are Eastern schools that simply wouldn't have the funds available to travel to top Western schools. This is interesting because it is a common misperception -- that second-tier teams in strong conference get a higher ranking simply because they get to play more games against top teams. But that's not the case. These schools need some actual results to earn the higher ranking. It's not like St. Cloud was getting ranked ahead of Cornell last year.
When you move to KRACH or Rutter though, the games against your relatively scarce quality of opponent do become relatively more important -- e.g. if you're Cornell and play only two games against a top 5 team, those games matter a lot for your ranking, whereas if you're Minnesota and Wisconsin each of your games against each other matters relatively less since you play each other four teams. Similarly, if you're Wisconsin and only play two games against teams in the bottom half of the ECAC, you'd better not lose them, whereas a team like Cornell that plays 12 such games has a bit more margin of error. So there are pros and cons to the system for a team like Cornell -- you stand to benefit a lot from going 2-0 in those games against top teams, while you stand to lose a lot if you go 0-2 in those games.
The NCAA selection criteria consider strength-of-schedule, but not in a way that is accurate. The way the NCAA does it they just take a weighted average of your record, your opponents' records, and your opponents' opponents' record (known as the ratings percentage index, the RPI, not to be confused with the ECAC school)
The problem with the RPI is that in theory, losing to a perfect team or beating a perfectly awful team should provide no new information about the quality of your school, but both results can affect your RPI (though the NCAA does have an "adjustment" that drops wins that lower your RPI from consideration). There are actually no perfect teams and no perfectly awful teams, but this problem still exists to some extent. Another way of looking at this:if the number of conference games gets large compared to the number of nonconference games, then the RPI is largely determined by your conference schedule, and a team that's 5th in a strong conference has no realistic chance of achieving a better RPI than the best team in the weaker conference.The KRACH and Rutter rankings are actually based on estimating statistical models, and they do not suffer from this fallacy.
Why does the NCAA not change this? The main reason I'd say is they have a strong institutional history with the RPI. Another reason is there's pressure from weaker conferences to maintain the status quo, since they are the ones hurt by moving to a more accurate system. (Of course, the RPI is still a better system than basing 1/3 of the selection of your national championship participants based on 5 computer rankings that are not made public, but that's off topic)
I had a conversation with a coach last year from Eastern school. I was interested that the coach offered the perception that a ranking like KRACH is fundamentally unfair because there are Eastern schools that simply wouldn't have the funds available to travel to top Western schools. This is interesting because it is a common misperception -- that second-tier teams in strong conference get a higher ranking simply because they get to play more games against top teams. But that's not the case. These schools need some actual results to earn the higher ranking. It's not like St. Cloud was getting ranked ahead of Cornell last year.
When you move to KRACH or Rutter though, the games against your relatively scarce quality of opponent do become relatively more important -- e.g. if you're Cornell and play only two games against a top 5 team, those games matter a lot for your ranking, whereas if you're Minnesota and Wisconsin each of your games against each other matters relatively less since you play each other four teams. Similarly, if you're Wisconsin and only play two games against teams in the bottom half of the ECAC, you'd better not lose them, whereas a team like Cornell that plays 12 such games has a bit more margin of error. So there are pros and cons to the system for a team like Cornell -- you stand to benefit a lot from going 2-0 in those games against top teams, while you stand to lose a lot if you go 0-2 in those games.
Last edited:
OnMAA
Vort Doen.....Heija Heija Heija
Re: 2011-12 DI Rutter Computer Rankings
Meanwhile, with enough games in the books to do the stats, the WCHODR stuff from SLU has started up for this season:
http://it.stlawu.edu/~chodr/wchodr/current.html
Minny actually ranked ahead of Wisco in this one.
Meanwhile, with enough games in the books to do the stats, the WCHODR stuff from SLU has started up for this season:
http://it.stlawu.edu/~chodr/wchodr/current.html
Minny actually ranked ahead of Wisco in this one.
Re: 2011-12 DI Rutter Computer Rankings
Only through 11/27. With the Gophers splitting their series against UND this past weekend, I suspect the Badgers will be ranked No. 1 when the poll is updated to reflect the more recent games.
Only through 11/27. With the Gophers splitting their series against UND this past weekend, I suspect the Badgers will be ranked No. 1 when the poll is updated to reflect the more recent games.
OnMAA
Vort Doen.....Heija Heija Heija
Re: 2011-12 DI Rutter Computer Rankings
ARM might be right. With Minny having 5 goals against in 2 games, it might push their total GAA on D up enough for Wisco to over come the small delta, as the mean reason Minny are on top is their 0.94 GAA defensive rating. Wisco outdid Minny by 6 in the combined GF-GA stat this weekend.
ARM might be right. With Minny having 5 goals against in 2 games, it might push their total GAA on D up enough for Wisco to over come the small delta, as the mean reason Minny are on top is their 0.94 GAA defensive rating. Wisco outdid Minny by 6 in the combined GF-GA stat this weekend.
Re: 2011-12 DI Rutter Computer Rankings
For games played through December 4th, 2011
For games played through December 4th, 2011
Code:
Team Rating
1 Wisconsin 2.6024
2 Minnesota 1.9134
3 Cornell 1.4382
4 North Dakota 1.1789
5 Mercyhurst 1.0072
6 UMD 0.9870
7 Boston College 0.9055
8 Bemidji State 0.8565
9 Harvard 0.7852
10 Ohio State 0.7784Re: 2011-12 DI Rutter Computer Rankings
For games played through December 11th, 2012
The teams ranked five-seven are really close at this point. BC has a little separation from Bemidji. Northeastern is ranked 9 in PWR, but around 12th in Rutter and KRACH. Assuming no tournament upsets, is this the NCAA eight?
For games played through December 11th, 2012
Code:
Team Rating
1 Wisconsin 2.6510
2 Minnesota 1.9811
3 Cornell 1.4510
4 North Dakota 1.1228
5 UMD 0.9804
6 Mercyhurst 0.9702
7 Harvard 0.9642
8 Boston College 0.8991
9 Bemidji State 0.7776
10 Ohio State 0.6671The teams ranked five-seven are really close at this point. BC has a little separation from Bemidji. Northeastern is ranked 9 in PWR, but around 12th in Rutter and KRACH. Assuming no tournament upsets, is this the NCAA eight?
OnMAA
Vort Doen.....Heija Heija Heija
Re: 2011-12 DI Rutter Computer Rankings
While you could argue ad nauseum about the order of 4-8, me thinks, that this is the right set of teams at this moment.
Having said that, a few teams on this list may at some point come knocking at the door. Those would be BU if they can get it back together, Quinnipiac if they continue their recent run, and North Eastern for the same reason.
In essence, the way it looks now you'll have four WCHA teams, and then Mercyhurst and the Eastern teams vying for the other four spots. Two of those eastern spots will be Autobids, so if Cornell does not win ECAC for example, it could knock a team like Harvard out of the equation. Same could happen to BC in HE.
While you could argue ad nauseum about the order of 4-8, me thinks, that this is the right set of teams at this moment.
Having said that, a few teams on this list may at some point come knocking at the door. Those would be BU if they can get it back together, Quinnipiac if they continue their recent run, and North Eastern for the same reason.
In essence, the way it looks now you'll have four WCHA teams, and then Mercyhurst and the Eastern teams vying for the other four spots. Two of those eastern spots will be Autobids, so if Cornell does not win ECAC for example, it could knock a team like Harvard out of the equation. Same could happen to BC in HE.
Last edited:
Re: 2011-12 DI Rutter Computer Rankings
Re: OnMAA, I don't think BC's in any danger of ending up on the tournament bubble this year. I expect Minnesota, Wisconsin, and then the leaders of the other three conferences (BC, Mercyhurst, Cornell) are pretty much in unless there's a significant drop off.
Re: Lakersfan, In terms of whether this is the 8, it's hard to say. The teams for the last three at-large bids are so close in the NCAA criteria, there's going to be a lot of randomness involved.
For instance, look at how awful the record vs. top 12 criteria is. Harvard's sitting pretty now because Clarkson and Dartmouth are in the top 12 and Harvard is 3-1 against them. If both of them slips out of the top 12, then Harvard's suddenly 0-3 against teams under consideration. It ends up making a huge difference who ends up in those 11-12 slots in determining whether a team finished 6-8th or 9-10. If say Ohio State makes it, then that's a huge boon to UND, bad for Bemidji. It's highly unlikely the WCHA gets 6 teams in the RPI 12, and most likely I'm thinking they'll end up with 4, though there is some chance they'll still have 5.
For example, if you look at last season, one reason why Dartmouth gets in over N. Dakota is that N. Dakota doesn't get enough credit for going 2-1-1 vs. Bemidji, who should've been a top 12 team.
Re: OnMAA, I don't think BC's in any danger of ending up on the tournament bubble this year. I expect Minnesota, Wisconsin, and then the leaders of the other three conferences (BC, Mercyhurst, Cornell) are pretty much in unless there's a significant drop off.
Re: Lakersfan, In terms of whether this is the 8, it's hard to say. The teams for the last three at-large bids are so close in the NCAA criteria, there's going to be a lot of randomness involved.
For instance, look at how awful the record vs. top 12 criteria is. Harvard's sitting pretty now because Clarkson and Dartmouth are in the top 12 and Harvard is 3-1 against them. If both of them slips out of the top 12, then Harvard's suddenly 0-3 against teams under consideration. It ends up making a huge difference who ends up in those 11-12 slots in determining whether a team finished 6-8th or 9-10. If say Ohio State makes it, then that's a huge boon to UND, bad for Bemidji. It's highly unlikely the WCHA gets 6 teams in the RPI 12, and most likely I'm thinking they'll end up with 4, though there is some chance they'll still have 5.
For example, if you look at last season, one reason why Dartmouth gets in over N. Dakota is that N. Dakota doesn't get enough credit for going 2-1-1 vs. Bemidji, who should've been a top 12 team.
Re: 2011-12 DI Rutter Computer Rankings
For games played through 1/8/12
For games played through 1/8/12
Code:
Team Rating
1 Wisconsin 2.3368
2 Minnesota 2.0555
3 Cornell 1.4392
4 North Dakota 1.0904
5 UMD 0.9812
6 Mercyhurst 0.8371
7 Boston College 0.8293
8 Harvard 0.7285
9 Bemidji State 0.7169
10 Boston Univ. 0.5937Watson Rink
Registered User
Naive question about "bunched" schedules
Naive question about "bunched" schedules
Perhaps Dave can shed some light on this somewhat random question.
It seems that the ECAC schedule has become more bunched this year in the sense that instead of playing one game in the fall against Travel Pairs #1 through #5 and your own travel buddy and then repeating the same process in more or less the same order later in the winter, you're playing Travel Pair #1 at home and then away a couple of weeks later, and then you're done with Travel Pair #1 for the season. This can be less fun if you have an intense rivalry with your travel buddy, as Harvard does with Dartmouth (both games were played months ago), and it seems to involve a greater amount of luck, as when you play both games with Travel Pair #1 two weeks apart just when your best player (or their best player) is injured. It also makes it a little harder to interpret the ECAC standings at midseason when some teams have played harder or easier schedules rather than everybody having played the same full round-robin once through, although you can mentally adjust the standings for strength/weakness of each team's remaining opponents.
My question is whether this bunching of results has any significant effect on PWR, RPI, Rutter or other statistical indices of performance? Or does everything even out? It seems intuitively that playing the same TUC at different times of the season, rather than a quick home-and-away, would involve less volatility in the week-by-week statistical indices, same as in conference results, but is that a false analogy?
Naive question about "bunched" schedules
Perhaps Dave can shed some light on this somewhat random question.
It seems that the ECAC schedule has become more bunched this year in the sense that instead of playing one game in the fall against Travel Pairs #1 through #5 and your own travel buddy and then repeating the same process in more or less the same order later in the winter, you're playing Travel Pair #1 at home and then away a couple of weeks later, and then you're done with Travel Pair #1 for the season. This can be less fun if you have an intense rivalry with your travel buddy, as Harvard does with Dartmouth (both games were played months ago), and it seems to involve a greater amount of luck, as when you play both games with Travel Pair #1 two weeks apart just when your best player (or their best player) is injured. It also makes it a little harder to interpret the ECAC standings at midseason when some teams have played harder or easier schedules rather than everybody having played the same full round-robin once through, although you can mentally adjust the standings for strength/weakness of each team's remaining opponents.
My question is whether this bunching of results has any significant effect on PWR, RPI, Rutter or other statistical indices of performance? Or does everything even out? It seems intuitively that playing the same TUC at different times of the season, rather than a quick home-and-away, would involve less volatility in the week-by-week statistical indices, same as in conference results, but is that a false analogy?
Re: Naive question about "bunched" schedules
Re: Naive question about "bunched" schedules
Well, it goes further than that because in the ECAC, for example, Princeton/Quinnipiac have already played Colgate/Cornell twice each (and Colgate/Cornell have played each other twice as well), but Colgate/Cornell have yet to play St. L/Clarky at all. So, in theory one could argue that they have had 4 easier or harder games (depending on who's perspective) already and have 4 easier or harder to go.
At the end of the day, the Rutter Ranking will (should) take into account the whole season so that's what really matters come mid-Feb., right?
Re: Naive question about "bunched" schedules
Well, it goes further than that because in the ECAC, for example, Princeton/Quinnipiac have already played Colgate/Cornell twice each (and Colgate/Cornell have played each other twice as well), but Colgate/Cornell have yet to play St. L/Clarky at all. So, in theory one could argue that they have had 4 easier or harder games (depending on who's perspective) already and have 4 easier or harder to go.
At the end of the day, the Rutter Ranking will (should) take into account the whole season so that's what really matters come mid-Feb., right?
brookyone
Deport Right Wing Nuts
ARM
Fan of chipmunk-like mascots.
Re: 2011-12 DI Rutter Computer Rankings
If they keep winning, it doesn't matter if they do so in regulation or OT. For the PWR, a win is a win, as long as it isn't a Shootout win, which is just a tie.
Re: 2011-12 DI Rutter Computer Rankings
PWR, RPI, Rutter and KRACH do not factor in when games are played. So at the end of the season, order of games are played is not an issue. PWR used to have a "last 16 games" component, but that was eliminated three or four years ago.
As for Mercyhurst, I think they have a little wiggle room. If they win out, they are in. Every tie/loss is going to hurt their RPI, although the teams behind them in the PWR have lower RPIs, so they have some room for error (as long as a run by another team doesn't increase their RPI). Now, if losing 1 out of 4 to a Robert Morris team that wins all their other games might actually help Merychurst. If RMU can become a team under consideration by winning a bunch of games, then Hurst going 3-1 against them will help them in the PWR. I don't think RMU making the top 12 is possible, so I think Mercyhurst could absorb one loss and still make it.
PWR, RPI, Rutter and KRACH do not factor in when games are played. So at the end of the season, order of games are played is not an issue. PWR used to have a "last 16 games" component, but that was eliminated three or four years ago.
As for Mercyhurst, I think they have a little wiggle room. If they win out, they are in. Every tie/loss is going to hurt their RPI, although the teams behind them in the PWR have lower RPIs, so they have some room for error (as long as a run by another team doesn't increase their RPI). Now, if losing 1 out of 4 to a Robert Morris team that wins all their other games might actually help Merychurst. If RMU can become a team under consideration by winning a bunch of games, then Hurst going 3-1 against them will help them in the PWR. I don't think RMU making the top 12 is possible, so I think Mercyhurst could absorb one loss and still make it.
Watson Rink
Registered User
Re: 2011-12 DI Rutter Computer Rankings
Your point about the final seasonresults appears valid, but part of my question was whether bunching of games against the same opponents makes weeek-by-week computerized results more volatile or whether the computer algorithms iron out any such volatility
Your point about the final seasonresults appears valid, but part of my question was whether bunching of games against the same opponents makes weeek-by-week computerized results more volatile or whether the computer algorithms iron out any such volatility