Anyone going to attempt to publish a PairWise ranking for D III?

[spwood]

I asked this question on "the other forum" and never received a response. Is anyone going to put together a PairWise ranking for DIII since the NCAA will be using it this year to select at-large teams?

[NUProf]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

I asked this question on "the other forum" and never received a response. Is anyone going to put together a PairWise ranking for DIII since the NCAA will be using it this year to select at-large teams?

The big question about a PWR is what are the components to be used? For example, how is SOS to be measured? I doubt that the same factors will be used as in the DI PWR. If the factors are known, it is pretty easy for someone to do.

[Fishman'81]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

The big question about a PWR is what are the components to be used? For example, how is SOS to be measured? I doubt that the same factors will be used as in the DI PWR. If the factors are known, it is pretty easy for someone to do.

I hear you.

But, if the D-1 PWR is employed, it'll shake-out almost identically to the Regional RPI. (And, nearly to the KRACH too, usually, which of course is a better metric than either of the former. )

[spwood]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

The big question about a PWR is what are the components to be used? For example, how is SOS to be measured? I doubt that the same factors will be used as in the DI PWR. If the factors are known, it is pretty easy for someone to do.

And if the NCAA doesn't disclose this, then there is no transparency, which is their stated purpose for changing to a Pairwise, right? I guess this was kind of my point in asking the question. I knew the components and weighting haven't been publicized yet.

[joecct]

And if the NCAA doesn't disclose this, then there is no transparency, which is their stated purpose for changing to a Pairwise, right? I guess this was kind of my point in asking the question. I knew the components and weighting haven't been publicized yet.

The NCAA D1 side never publishes a PWR. The various web sites do. It so happens that the USCHO/CHN pwrs mirror the committee's charter.

On the NC women's side, Grant (TonytheTiger) has the most accurate PWR.

[Russell Jaslow]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

USCHO will after an "x" number of games are played. Keep your eye out for it.

[NUProf]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

The NCAA D1 side never publishes a PWR. The various web sites do. It so happens that the USCHO/CHN pwrs mirror the committee's charter.

On the NC women's side, Grant (TonytheTiger) has the most accurate PWR.

The DI factors are known and disclosed for both genders. If we know the factors, anybody with a computer can figure it out. I would assume that the factors would be similar to what is published for the other DIII sports, but applied as a PWR instead of having the committee use the smoke filled room in its process. If that is the case, then the question becomes how will RNK work, since that is part of the DIII criteria.

[Fishman'81]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

The DI factors are known and disclosed for both genders. If we know the factors, anybody with a computer can figure it out. I would assume that the factors would be similar to what is published for the other DIII sports, but applied as a PWR instead of having the committee use the smoke filled room in its process. If that is the case, then the question becomes how will RNK work, since that is part of the DIII criteria.

The "RNK" farce should not in any way factor-in to a true PWR. Otherwise, why attempt making the process appear objective via yet another artificial process in the first place..? Do the powers that be consider the fans so clueless as to think that s*** will flush?

My opinion -and I don't know what D-3 has planned- is that we'll see a transparent and easily verifiable PW this season, a la D-1. There would be no reason in the world to trot-out another BS smokescreen, when they already have a perfectly cynical and fungible one in place.

[Russell Jaslow]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

The "RNK" farce should not in any way factor-in to a true PWR. Otherwise, why attempt making the process appear objective via yet another artificial process in the first place..? Do the powers that be consider the fans so clueless as to think that s*** will flush?

My opinion -and I don't know what D-3 has planned- is that we'll see a transparent and easily verifiable PW this season, a la D-1. There would be no reason in the world to trot-out another BS smokescreen, when they already have a perfectly cynical and fungible one in place.

I don't know why people are throwing in RNK. I always thought it was going to be a straight up PWR just like D1. I believe your last paragraph is correct and the way it's going to be done.

[Ed Trefzger]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

I don't know why people are throwing in RNK. I always thought it was going to be a straight up PWR just like D1. I believe your last paragraph is correct and the way it's going to be done.

It's all explained in this NCAA memo which is found linked on the D-III men's hockey championships page at the NCAA: https://ncaaorg.s3.amazonaws.com/cha...ot_Program.pdf

As Russell noted, USCHO will have D-III men's and women's PWR based on the D-I algorithm very early in 2019.

[NUProf]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

It's all explained in this NCAA memo which is found linked on the D-III men's hockey championships page at the NCAA: https://ncaaorg.s3.amazonaws.com/cha...ot_Program.pdf

As Russell noted, USCHO will have D-III men's and women's PWR based on the D-I algorithm very early in 2019.

Here is the relevant quote from the memo which should give pause

As you know, the PairWise system calculates rankings based on a comparison matrix that applies an equal weight to the selection criteria to compare each team to all other teams.

What has not been made clear is what the selection criteria are. I support the change. My point is that the official selection criteria include RNK and there is not indication that this is not still the case. That is my only concern

[joecct]

Here is the relevant quote from the memo which should give pause



What has not been made clear is what the selection criteria are. I support the change. My point is that the official selection criteria include RNK and there is not indication that this is not still the case. That is my only concern

If D1 is any guide
RPI
H2H
COP

RPI is the tiebreaker.

[NUProf]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

If D1 is any guide
RPI
H2H
COP

RPI is the tiebreaker.

My point is that nobody has said that.

[Russell Jaslow]

Here is the relevant quote from the memo which should give pause



What has not been made clear is what the selection criteria are. I support the change. My point is that the official selection criteria include RNK and there is not indication that this is not still the case. That is my only concern

You need to see appendix A. That’s where they say the details are.

[Ed Trefzger]

Re: Anyone going to attempt to publish a PairWise ranking for D III?

You need to see appendix A. That’s where they say the details are.

In addition to appendix A, USCHO has confirmed that the formula will be identical to D-I.

It's about time that the NCAA tournament field and pairings was not manipulated. Maybe the retirement of a prominent coach is part of that.