Changes to Pairwise

[Patman]

According to CHN
Weight a win @ home or a loss on the road by 0.8
Weight a win on the road or a home loss by 1.2

Also a sliding bonus for beating a top 20 (on Selection Sundae??) RPI team from 5.0 to 0.25.

CHN crunching numbers to how it pans out. Somewhere, wherever Jon Whelan is, the CPU usage just tripled.

Actually the hamsters aren't a real issue unless its Monte Carlo. Read in and out is a fraction of a second for a competent coder.

Did they give any definition for neutral site games?

Edit: as in when is a neutral site game really neutral? Minny at xcel? Maine in Portland? UAF playing a lower 48 in anchorage? Etc?

 

[Biddco]

Re: Changes to Pairwise

I'm calling these changes the "Michigan Modification". It's all about getting teams to travel OOC.

 

[5mn_Major]

Re: Changes to Pairwise

ESPN likes a Cinderella story every once in a while.

Now if it only they liked college hockey.

 

[The Exiled One]

Re: Changes to Pairwise

Actually the hamsters aren't a real issue unless its Monte Carlo. Read in and out is a fraction of a second for a competent coder.

Did they give any definition for neutral site games?

Edit: as in when is a neutral site game really neutral? Minny at xcel? Maine in Portland? UAF playing a lower 48 in anchorage? Etc?

I'd love if they made the North Star College Cup a home series for the Gophers! They could justify the decision because the Gophers don't rotate out of the tournament every fifth year.

*prepares self for Gopher comment onslaught*

 

[ExileOnDaytonStreet]

Re: Changes to Pairwise

Also a sliding bonus for beating a top 20 (on Selection Sundae??) RPI team from 5.0 to 0.25.

Well that just seems stupid. You're going to apply RPI rankings to games played early in the season (when RPI is still fluctuating a lot)? Or are you going to apply season ending RPI, which doesn't necessarily describe how good the team was at the time that you played them?

 

[5mn_Major]

Re: Changes to Pairwise

According to CHN
Weight a win @ home or a loss on the road by 0.8
Weight a win on the road or a home loss by 1.2

Was it the same at home and on the road before? This gives a flat 0.2 points to an away team...assuming there is no home rink advantage. If you factor that it...its perhaps 0.1 pts. If you want to play at home, you just better be better on your home rink or youre giving away points.

 

[Priceless]

Re: Changes to Pairwise

According to CHN
Weight a win @ home or a loss on the road by 0.8
Weight a win on the road or a home loss by 1.2

Also a sliding bonus for beating a top 20 (on Selection Sundae??) RPI team from 5.0 to 0.25.

CHN crunching numbers to how it pans out. Somewhere, wherever Jon Whelan is, the CPU usage just tripled.

Can't access his site If the site comes back, we should be able to apply these changes to previous seasons back to 2003.

 

[FlagDUDE08]

Re: Changes to Pairwise

According to CHN
Weight a win @ home or a loss on the road by 0.8
Weight a win on the road or a home loss by 1.2

Also a sliding bonus for beating a top 20 (on Selection Sundae??) RPI team from 5.0 to 0.25.

CHN crunching numbers to how it pans out. Somewhere, wherever Jon Whelan is, the CPU usage just tripled.

Would the weighting be 1.0 for neutral site?

 

[Ralph Baer]

Re: Changes to Pairwise

Well that just seems stupid. You're going to apply RPI rankings to games played early in the season (when RPI is still fluctuating a lot)? Or are you going to apply season ending RPI, which doesn't necessarily describe how good the team was at the time that you played them?

I would be very surprised if it isn't the end-of-season RPI.

 

[Priceless]

Re: Changes to Pairwise

CHN update:

For example, among teams in the new Big Ten, Wisconsin has 14 non-league games, 10 home and four away; Ohio State has 14 non-league games, 11 home and three away; and Minnesota has 16 non-league games, 12 at home (including a tournament at the Xcel Center) and four away.

In the new plan, the RPI (Ratings Percentage Index) will be weighted so that road wins and home losses count by a factor of 1.2. Also, home wins and road losses will count by a factor of 0.8. In other words, road wins and home losses will be a significantly bigger factor in affecting your RPI, which is, generally speaking, the ranking system used by the NCAA to adjust your winning percentage by a strength of schedule.

In addition, your RPI will get a "bonus" for wins against teams ranked in the Top 20 of RPI, based on a sliding scale. If you defeat the No. 1 team, you get an RPI bonus of 5.0 percentage points, sliding down to a 0.25 bonus for defeating the 20th team.

The Quality Win Bonus (QWB) will be multiplied by a factor consistent with the home-road weighting for that game. So, a win on the road against the No. 1 team, for example, would give you a bonus of 6.0 (5.0 x 1.2), whereas a win against the No. 1 team at home, would give you a bonus of 4.0 (5.0 x 0.8).

 

[Patman]

Re: Changes to Pairwise

CHN update:

Yeah, the quality win aspect seems the most bogus to me. Aside from the fact that the #1 team can't beat itself, it seems to me the power conferences will rake in on that factor alone. I assume its going to be cumulative over all of your opponents and not just a singular opponent. Maybe this is making up for taking out TUC but I don't see this as a good replacement.

 

[joecct]

Re: Changes to Pairwise

I would be very surprised if it isn't the end-of-season RPI.

According to CHN, it is.

Would the weighting be 1.0 for neutral site?

According to CHN, it is.


A few years ago, D-III tried a concept of a 15 point SoS metric based on whether you beat a team with a great WP on the road (15), home (14), down to losing to a poor team on the road (2) or at home (1). I think it lasted all of 2 seasons before they came up with the current strategy (which nobody knows how it is applied).

 

[ExileOnDaytonStreet]

Re: Changes to Pairwise

Yeah, the quality win aspect seems the most bogus to me.

No need to elaborate. It is bogus. RPI already takes the quality of opponent into account, and the PWR uses record against TUC to give a boost for quality wins. What are we gaining with this, I'll say it, awfully BCS-ish kind of weighting factor?

 

[4four4]

Re: Changes to Pairwise

No need to elaborate. It is bogus. RPI already takes the quality of opponent into account, and the PWR uses record against TUC to give a boost for quality wins. What are we gaining with this, I'll say it, awfully BCS-ish kind of weighting factor?

We all knew if the BiG decided to start their own conference they were going to change the rules.

 

[4four4]

Re: Changes to Pairwise

I'd love if they made the North Star College Cup a home series for the Gophers! They could justify the decision because the Gophers don't rotate out of the tournament every fifth year.

*prepares self for Gopher comment onslaught*

Meh. Life isn't fair.

 

[4four4]

Re: Changes to Pairwise

When was the objective to be like College football and basketball?

Two years ago when the Big Ten decided to form their own conference.

 

[lugnut92]

Re: Changes to Pairwise

So it really looks like these weights are going to exist throughout the season, which means teams also get punished for being at home in the early rounds of their conference playoffs. A team that wins in three games at home ends up with an effective record of 1.6-1.2 instead of 2-1 because they were the better team in the regular season. I'm all in favor of getting the giants out west to travel more than twice a season, but I think it should really only count for OOC play.

 

[Slap Shot]

Re: Changes to Pairwise

If this was really added to impact scheduling, why not wait 1-2 seasons to allow to adjust future scheduling? Not much teams can do about it this season.

 

[Priceless]

Re: Changes to Pairwise

The USCHO piece

Those actions reflect what the committee heard from coaches at their annual meeting in Florida in April, said incoming chair Jim Knowlton, the athletic director at Rensselaer.

“Their voice was pretty loud on the fact that they would like to see the committee look at both the home-and-away piece and then also the teams under consideration piece,” said Knowlton, who took over as committee chair at the start of September. “Both of those had the potential to have an impact on the selection criteria, and so that’s what the committee since Naples in April has looked at. What you saw come out of the NCAA was what the committee thought made the most sense for college hockey.”

 

[Ralph Baer]

Re: Changes to Pairwise

So it really looks like these weights are going to exist throughout the season, which means teams also get punished for being at home in the early rounds of their conference playoffs. A team that wins in three games at home ends up with an effective record of 1.6-1.2 instead of 2-1 because they were the better team in the regular season. I'm all in favor of getting the giants out west to travel more than twice a season, but I think it should really only count for OOC play.

An excellent point.