PDA

View Full Version : Changes to Pairwise



Pages : [1] 2 3 4 5 6

mar7967
09-18-2013, 04:39 PM
CHN is reporting that the NCAA Cabinet approved a change to the selection process for the tournament. They will be considereding a "quality wins bonus" and "weighting for home and away games."

http://www.collegehockeynews.com/news/2013/09/18_home-road_balance_will_be.php

Thoughts?

Patman
09-18-2013, 04:43 PM
Yeah, incredibly ill advised without knowing the impacts of such things.

Do they even know the impact of their fudge factors. Home ice advantage does exist but it's not extreme. If you want to penalize teams for too many home games do it via a cap, not via pairwise.

Edit: we knew this was coming, the real question was how soon. I wanted to evaluate possibilities but I couldn't come up with a notion of how to compare to simulated lists of results... Could I have done something, likely, but I wanted the correct something.

ExileOnDaytonStreet
09-18-2013, 05:45 PM
The more factors you introduce, the more subjective the process becomes.

I like the idea of forcing the major programs to go on the road now and again, but the "cold hard math" side of the PWR is one of the reasons I like it so much as a selection criterion.

The Exiled One
09-18-2013, 05:53 PM
The more factors you introduce, the more subjective the process becomes.

I like the idea of forcing the major programs to go on the road now and again, but the "cold hard math" side of the PWR is one of the reasons I like it so much as a selection criterion.
I think they can add as many factors as they like. As long as they are objective, it's still "cold hard math". Google's algorithm is just math, but it's incredible complexity is why it works so well.

I say, go ahead and add complexity, just make sure it's transparent. Selection Sunday should be just a formality.

Patman
09-18-2013, 06:01 PM
maybe they are trying to nudge programs like Michigan into playing some away games and penalizing them for their refusing to play away games that they absolutely do not need to play?

and home ice CAN matter. not so much at Lowell, but at Wisconsin or the Alfond back in the 90s

Rare circumstances not deserving of a one size fits all method

Patman
09-18-2013, 06:14 PM
I think they can add as many factors as they like. As long as they are objective, it's still "cold hard math". Google's algorithm is just math, but it's incredible complexity is why it works so well.

I say, go ahead and add complexity, just make sure it's transparent. Selection Sunday should be just a formality.

It's not objective. It's just formulaic...

I am starting to wonder if this was a compromise of the form "we will give you this if you give us that". The schools that may advantage from a home favoring system would be disadvantaged with a quality wins construction and vice versa. Quality wins would incentivize power conferences. Quality non-conf would only incentivize the top teams to play each other.

In any system, if the "model" is not true then it can be gamed. Current pairwise is gameable but its not so certain how, especially with so many conference tilts kind of averaging things out for a league. If you believe krach then its more important for PWR to pick up wins against lesser competition as krach favors team w strong SOS and PWR doesn't as much.

Granted, the strength of said gaming was never clear. Adding factors usually creates more undesirable incentives than it solves due to the additional complexity and de-weighting factors that were more useful to capture things.

The key question is whether or not the new system captures team ability more than the old knowing people will take advantage of the system.

I am pessimistic here. I see gameable features here being crafted by people who aren't sure what they are doing.

The fact that i would have some difficulty with the idea of calibrating these factors appropriately should give pause.

CLS
09-18-2013, 08:14 PM
I think this is more of a compromise in the form of (a) home/away disparities are unfair from a competitive standpoint but that (b) home/away disparities are desirable financially, so they're not going away. The NCAA realizes all too well that both teams benefit financially from a game being at, for example, Yost rather than at a 1200 seat stadium in North Bumfugg.

FiveHole12
09-18-2013, 08:21 PM
I have always known there is a East Bumfugg, but had no idea there was a North Bumfugg. Learn something new every day. :D

Biddco
09-18-2013, 08:49 PM
Less 2 for 1 agreements that schools do the better it is for college hockey. 2 for 1's suck.


Don't know how I feel about no longer having the TUC line.

purpleinnebraska
09-18-2013, 09:52 PM
As long as they are objective, it's still "cold hard math".

But how do you make it objective? Is a win at Yost worth more or less than a win at Mariucci? If more, how much more? If less, how much less? If it's worth more than a win at Sacred Heart (which I think we would all agree that it is), how much more (which is where we'd all disagree)? Wins and SOS were already figured in. Now we're adding some kind of a "degree of difficulty" that doesn't make the Pairwise a whole lot different than the basketball "smoke-filled room," or the gymnastics/figure skating "Hey, give her a good score. She's the favorite" model.

mnpokecheck
09-18-2013, 09:56 PM
So with a frozen four last year of Quinnipiac, St. Cloud State, Mass.-Lowell and Yale.....what are they trying to fix? Wait for the complaining to go back to the old formula if four past champs are in the frozen four this season.

CLS
09-18-2013, 11:20 PM
But how do you make it objective? Is a win at Yost worth more or less than a win at Mariucci? ....I don't see anything that proposes that. Just that a win on the road is worth more than a win at home.





Don't know how I feel about no longer having the TUC line.I never understood why it needs to be a cliff. Why not a slope?

Priceless
09-18-2013, 11:31 PM
Let's just put the four most lucrative programs in the Frozen Four every year. That will make the sponsors happy, ESPN happy, the arena happy and the NCAA happy.

The Rube
09-19-2013, 12:05 AM
Let's just put the four most lucrative programs in the Frozen Four every year. That will make the sponsors happy, ESPN happy, the arena happy and the NCAA happy.


I'll take all the titles for my team, thankyouverymuch. No argument from me.

Priceless
09-19-2013, 12:27 AM
I'll take all the titles for my team, thankyouverymuch. No argument from me.

It'll be Minnesota, Wisconsin and North Dakota every year. The 4th team will either be Michigan, New Hampshire, Maine or a private school that can show revenue above $2.75M/year

4four4
09-19-2013, 12:36 AM
It'll be Minnesota, Wisconsin and North Dakota every year. The 4th team will either be Michigan, New Hampshire, Maine or a private school that can show revenue above $2.75M/year

I don't see the problem. Lol

The Rube
09-19-2013, 12:37 AM
It'll be Minnesota, Wisconsin and North Dakota every year. The 4th team will either be Michigan, New Hampshire, Maine or a private school that can show revenue above $2.75M/year

Like I said....;)

Agganis
09-19-2013, 08:38 AM
I never understood why it needs to be a cliff. Why not a slope?

RPI (and KRACH) are effectively a sloped TUC cliff where each team is valued slightly less than the team above them. Having a second one of these metrics would be redundant.

jcarter7669
09-19-2013, 09:40 AM
Western Bias

ExileOnDaytonStreet
09-19-2013, 09:51 AM
I think they can add as many factors as they like. As long as they are objective, it's still "cold hard math". Google's algorithm is just math, but it's incredible complexity is why it works so well.

I say, go ahead and add complexity, just make sure it's transparent. Selection Sunday should be just a formality.

Complex formulae require more subjective decisions (do we add these extra variables? Do we average them? What multiplying coefficient do we use?), so it's impossible to add extra variables or criteria without increasing the risk of being more subjective.

That said, I agree with your statement as far as transparency is concerned. At large selection should remain a formality.