Power Ranking Question

[watchmaker49]

OK some of you really understand how the rankings work much better than I do. Could someone tell me if this is the right way to look at something or explain if I have it wrong. Sorry in advance if I make this confussing. Here goes nothing. The Gophers again have a weak non-conference schedule. They make one road trip to Vermont. Play a home series with Michigan State. A possible game with BC at Christmas, I am sure they will only play them the second night if they and BC win Friday night. A single game with Notre Dame will be played at home. Now my question is are they piggybacking on the win/loss statement of other teams? Example would be if they play the team that maybe upset BC and win does that give them a boost in the rankings? Would they get more credit for winning one game than they deserve? So if they beat MS at the begining of the season and MS goes on to having a great season will they get more credit for that? Thanks in advance for helping me with this.

[82SiouxGuy]

OK some of you really understand how the rankings work much better than I do. Could someone tell me if this is the right way to look at something or explain if I have it wrong. Sorry in advance if I make this confussing. Here goes nothing. The Gophers again have a weak non-conference schedule. They make one road trip to Vermont. Play a home series with Michigan State. A possible game with BC at Christmas, I am sure they will only play them the second night if they and BC win Friday night. A single game with Notre Dame will be played at home. Now my question is are they piggybacking on the win/loss statement of other teams? Example would be if they play the team that maybe upset BC and win does that give them a boost in the rankings? Would they get more credit for winning one game than they deserve? So if they beat MS at the begining of the season and MS goes on to having a great season will they get more credit for that? Thanks in advance for helping me with this.

It depends on which system you are talking about because they are all different. You are probably referring to Pairwise since that is the one used for the tournament. Pairwise does use the success of everyone's opponents as a part of their ranking. But I'm not an expert on the rankings so I won't try to explain it. I just know that losing to high ranked teams doesn't hurt as much as losing to low ranked teams, except in the head to head categories.

[sagard]

Strength of schedule was majorly devalued several years ago. The most important thing is now to have a great record. Strength of schedule still plays a role but much less.

Current RPI Format 25/21/54

1 Boston Coll (HE)

2 Michigan (CC)

3 Union (EC)

4 North Dakota (WC)

5 Miami (CC)

6 Ferris State (CC)

7 Minn-Duluth (WC)

8 Minnesota (WC)

9 Boston Univ (HE)

10 Maine (HE)

11 Denver U (WC)

12 Mass-Lowell (HE)

13 Cornell (EC)

14 Western Mich (CC)

15 Mich State (CC)

Old RPI format 25/50/25

1 Boston Coll (HE)

2 Michigan (CC)

3 North Dakota (WC)

4 Union (EC)

5 Boston Univ (HE)

6 Miami (CC)

7 Maine (HE)

8 Mass-Lowell (HE)

9 Western Mich (CC)

10 Denver U (WC)

11 Minn-Duluth (WC)

12 Ferris State (CC)

13 Mich State (CC)

14 Minnesota (WC)

15 Cornell (EC)

The teams would have stayed the same, but the brackets would have been MUCH different.

[watchmaker49]

Strength of schedule was majorly devalued several years ago. The most important thing is now to have a great record. Strength of schedule still plays a role but much less.

Current RPI Format 25/21/54

1 Boston Coll (HE)

2 Michigan (CC)

3 Union (EC)

4 North Dakota (WC)

5 Miami (CC)

6 Ferris State (CC)

7 Minn-Duluth (WC)

8 Minnesota (WC)

9 Boston Univ (HE)

10 Maine (HE)

11 Denver U (WC)

12 Mass-Lowell (HE)

13 Cornell (EC)

14 Western Mich (CC)

15 Mich State (CC)

Old RPI format 25/50/25

1 Boston Coll (HE)

2 Michigan (CC)

3 North Dakota (WC)

4 Union (EC)

5 Boston Univ (HE)

6 Miami (CC)

7 Maine (HE)

8 Mass-Lowell (HE)

9 Western Mich (CC)

10 Denver U (WC)

11 Minn-Duluth (WC)

12 Ferris State (CC)

13 Mich State (CC)

14 Minnesota (WC)

15 Cornell (EC)

The teams would have stayed the same, but the brackets would have been MUCH different.

Could you explain the 25/50/25 numbers please.

[The Sicatoka]

RPI is based on three factors:

• Your winning percentage
  
• Your opponent's winning percentage
  
• Your opponent's opponent's winning percentage
  

The 25/50/25 is the weighting given to each in the calculation of your RPI.

[watchmaker49]

RPI is based on three factors:

• Your winning percentage
  
• Your opponent's winning percentage
  
• Your opponent's opponent's winning percentage
  

The 25/50/25 is the weighting given to each in the calculation of your RPI.

Then am I in the right ballpark thinking that the Gophers are piggy backing on other teams success? Or at the same time the failure of the other teams?

[The Sicatoka]

The formula is the same for all teams.

[watchmaker49]

The formula is the same for all teams.

What I mean is since they play such a lame non-conference schedule does that help them?

[as15]

What I mean is since they play such a lame non-conference schedule does that help them?

I'm not an expert either, but to answer your question, if they win those non-conference games it will help them slightly, whereas if they lose, it could hurt them a lot.

[The Sicatoka]

What I mean is since they play such a lame non-conference schedule does that help them?

If the conference is bad in non-conference play that "opponent's opponent's" number takes a hit and hurts the conference overall. The WCHA had that problem somewhat this past season.

[siouxnews]

bc from hockey east and notre dame/michigan state from ccha are weak non conference opponents? bc and notre dame might enter the year as their conference preseason favorites

[watchmaker49]

bc from hockey east and notre dame/michigan state from ccha are weak non conference opponents? bc and notre dame might enter the year as their conference preseason favorites

They will not play BC for sure though and the games are at home.

[Snake]

I'm not an expert either, but to answer your question, if they win those non-conference games it will help them slightly, whereas if they lose, it could hurt them a lot.

Case-in-point is the Gophers' losses to Vermont and Northeastern this past season.

[Southernmn]

Case-in-point is the Gophers' losses to Vermont and Northeastern this past season.

Correct........if the Gophers had won JUST that one game against Vermont, they would have finished #3 in the Pairwise.

[GFG]

After seeing some of the non-conference games Minnesota is picking up when they join the B1G conference, it seems to me that they usually make a lighter non-conference schedule in the WCHA because it is usually the top conference in college hockey. If you're playing top teams almost night in and night out I assume their logic is "we don't need top teams to boost us up in the rankings for non-conference games."

If you look at their non-conference games once they go to the B1G it's almost ridiculous, which makes sense because it's a small conference and you're going to need to play some big teams in non-conference games to get a bump in the rankings. They have non-conference series against BC, ND, possibly DU and CC or UND (still hasn't been clarified for that), UMD just about every year and the other Minnesota schools (granted not all of those schools will help a lot). They're kicking it up a notch, which leads me to assume it's because they know their future conference won't be as strong as the current WCHA and they know they need the boost with non-conference games now.

Just my 2 cents, chime in if you see another reason.

[yzerman19]

The RPI described is right on. The TUC and the COP components have not been addressed. If you beat say New Hampshire in a one game out of conference situation, and say BC goes 3-1 vs them in 4 games. You win the COP category vs BC ( all else being equal). Now it is likely that you face more than just the one COP (common opponent), so it isn't as simple as I am describing.

In theory, you are best playing your out of conference games against the middle tier teams- you are likely to win one game, and they are likely to pull off one win vs the top teams in a 4 game series.

[watchmaker49]

RPI is based on three factors:

• Your winning percentage
  
• Your opponent's winning percentage
  
• Your opponent's opponent's winning percentage
  

The 25/50/25 is the weighting given to each in the calculation of your RPI.

If a middle of the road team goes on a little rampage and wins alot of big games say over their last 6 games. Then depending on who they had played that year and how they had done, against certain teams, can really throw a wrench and change the rankings heavily? Do I have this right? Is this what happened with UND this year?

[yzerman19]

If a middle of the road team goes on a little rampage and wins alot of big games say over their last 6 games. Then depending on who they had played that year and how they had done, against certain teams, can really throw a wrench and change the rankings heavily? Do I have this right? Is this what happened with UND this year?

definitely true.

The pairwise is determined by RPI (which was well dscribed earlier as your record, your opponents' records, and your opponents' opponents' record), your comparison of record head to head, your comparison of record against "teams under consideration" for the NCAA tournament, and your comparison of record against "common opponents". RPI is the tie breaker. RPI, COP, and TUC are all worth "one goal" each win head to head is wort "one goal". Example: If UND has a better RPI, better record against COPs, and better record against TUCs, but gets swept in a 4 game series by the Gophers, the Gophers would win the pairwise comparison vs UND 4-3. They do not double count TUC and head to head when calculating TUC, so in this example, say both UND and UMN are tournament teams, the Gophers four wins would not hurt us twice vs the Gophers- those 4 losses would be removed from the TUC record for purposes of that pairwise calculation i.e. UND vs UMN.

It is important to note that the pairwise is a comparison of one team vs each of the other teams under consideration for the tournament.

Wins against cupcakes do not help. The RPI of those games is often lower than the total RPI, in which case they are removed from the RPI calculation. Losing against cupcakes kills you.

Again, the best bet for sandbagging the Pairwise is to play non-conference games vs the middle tier opponents. If you're BC, it is ideal to play last years Wisconsin or a St. Cloud- they are TUCs, but they weren't tournament teams (falling outside the top 16). BC should win these games- in a one or two game series- boosting their record against TUCs and COPs for WCHA teams. Even in a great year, it is tough for a UND to sweep all four games vs a Wisconsin or St. Cloud (case in point- last years .500 record against them).

If BC plays Wisconsin once and St Cloud twice and goes 2-1 vs them, and UND plays those teams 8 times and goes 5-3, BC has the PWR advantage on both TUC and COP: 0.666 vs 0.625.

So, you want to play mid-tier teams that your non-conference powerhouses play most often.

That is sandbagging, however, I still believe to be the best, you need to play the best. In an ideal world UND would play BC/BU/Maine/NH/Cornell/Michigan type teams every year.

[watchmaker49]

definitely true.

The pairwise is determined by RPI (which was well dscribed earlier as your record, your opponents' records, and your opponents' opponents' record), your comparison of record head to head, your comparison of record against "teams under consideration" for the NCAA tournament, and your comparison of record against "common opponents". RPI is the tie breaker. RPI, COP, and TUC are all worth "one goal" each win head to head is wort "one goal". Example: If UND has a better RPI, better record against COPs, and better record against TUCs, but gets swept in a 4 game series by the Gophers, the Gophers would win the pairwise comparison vs UND 4-3. They do not double count TUC and head to head when calculating TUC, so in this example, say both UND and UMN are tournament teams, the Gophers four wins would not hurt us twice vs the Gophers- those 4 losses would be removed from the TUC record for purposes of that pairwise calculation i.e. UND vs UMN.

It is important to note that the pairwise is a comparison of one team vs each of the other teams under consideration for the tournament.

Wins against cupcakes do not help. The RPI of those games is often lower than the total RPI, in which case they are removed from the RPI calculation. Losing against cupcakes kills you.

Again, the best bet for sandbagging the Pairwise is to play non-conference games vs the middle tier opponents. If you're BC, it is ideal to play last years Wisconsin or a St. Cloud- they are TUCs, but they weren't tournament teams (falling outside the top 16). BC should win these games- in a one or two game series- boosting their record against TUCs and COPs for WCHA teams. Even in a great year, it is tough for a UND to sweep all four games vs a Wisconsin or St. Cloud (case in point- last years .500 record against them).

If BC plays Wisconsin once and St Cloud twice and goes 2-1 vs them, and UND plays those teams 8 times and goes 5-3, BC has the PWR advantage on both TUC and COP: 0.666 vs 0.625.

So, you want to play mid-tier teams that your non-conference powerhouses play most often.

That is sandbagging, however, I still believe to be the best, you need to play the best. In an ideal world UND would play BC/BU/Maine/NH/Cornell/Michigan type teams every year.

Ok how the heck can you keep this all straight?

[82SiouxGuy]

Ok how the heck can you keep this all straight?

We let Jim Dahl do it. He does a great job and we don't have to worry about it.

[watchmaker49]

We let Jim Dahl do it. He does a great job and we don't have to worry about it.

I sure can understand that. If you knew what you were doing would it take that long to enter all the info?

[82SiouxGuy]

I sure can understand that. If you knew what you were doing would it take that long to enter all the info?

Actually, he has it built into this site. I assume that he has the site gather the information from other places, I know that he doesn't sit at home entering games as they happen. But if you were going to do it by hand it would take quite a while.

[watchmaker49]

Actually, he has it built into this site. I assume that he has the site gather the information from other places, I know that he doesn't sit at home entering games as they happen. But if you were going to do it by hand it would take quite a while.

Then overall this system really is the fairest (is that a word?) to decide and seed the teams it seems. It keeps the politics out of the equation. Do I have that right?

[82SiouxGuy]

Then overall this system really is the fairest (is that a word?) to decide and seed the teams it seems. It keeps the politics out of the equation. Do I have that right?

It is much more fair than systems used in other sports. Some people believe that the KRACH rankings would be more accurate. It is another system that ranks teams buy results. I am not as familiar with it so I couldn't tell you the differences. But many fans think it would be more representative.

[watchmaker49]

It is much more fair than systems used in other sports. Some people believe that the KRACH rankings would be more accurate. It is another system that ranks teams buy results. I am not as familiar with it so I couldn't tell you the differences. But many fans think it would be more representative.

I don't know if by results would be better or not. Scores are not as important as if you won or not. Way too many independant variables using scores. This has been a good lesson for me and I thank all of you who explained it to me in such detail.