Changes to Pairwise

[LynahFan]

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.

Doesn't seem like there are really quite enough details released for anyone to do the number crunching yet - certainly not in the CHN article. I'm a little surprised that they seem to be applying this to conference as well as non-conference games. Seems like an unfair advantage for the mid-to-bottom feeders of a top conference (who will have multiple opportunities to win bonus points by playing top teams) over the best teams from weaker conference who will have fewer such opportunities.

 

[CLS]

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.

I doubt there's much they would do anyway. There's too much money involved for, say, Minnesota to start scheduling games in Huntsville because because it might affect their chances to make the tournament or their seeding in the tournament. What we'd be more likely to see is micro-adjustments to the algorithm (e.g. 1.1/0.9 instead of 1.2/0.8) if there are some anomalous results, and the micro-adjustments will change until the "traveling" teams are satisfied that they're being recognized for the scheduling disparity and the home-heavy teams are satisfied that the adjustments don't go too far.

I'm a little surprised that anyone thinks that this will result in any significant scheduling change at all. It think it's really clear than the home/away scheduling disparity is here to stay because it benefits both parties. Adjusting schedules is expensive for both parties; adjusting the algorithm is a political process that can be resolved with phone calls, emails, and text messages.

 

[Patman]

Doesn't seem like there are really quite enough details released for anyone to do the number crunching yet - certainly not in the CHN article. I'm a little surprised that they seem to be applying this to conference as well as non-conference games. Seems like an unfair advantage for the mid-to-bottom feeders of a top conference (who will have multiple opportunities to win bonus points by playing top teams) over the best teams from weaker conference who will have fewer such opportunities.

It really seems to be if they mean percentage points or not. I mean, playing six games against the top 10 and splitting would presumably mean a bonus of 7.5%

Now, I take that to mean an RPI of .5000 becomes .5750. Of course if I apply that in a lower decimal place it becomes different .5075 or .5008

Considering how much of a schedule can be against the top 20 in HEA and B10 those are going to add up quickly.

I haven't attacked this because I don't have games identified by home/road/neutral. Even then it'd require some annoying work but it wouldn't be prohibitive. However, further, is SOS measured off of the weighted win %? Does the game for SOS and SOSOS factors get weighted with the home/road weight or do they only count once?

Once you get past those gymnastics then you have to code it up. This can be a tricky math problem in the end.

 

[vizoroo]

Re: Changes to Pairwise

Why not just use Krach?

 

[gfmorris]

Re: Changes to Pairwise

I'd like to hear a more quantitative statement from Patman on that; what is the home ice advantage?. If, say, the most home-heavy schedule is 16 home and 14 on the road and the most road-heavy schedule us 14 home and 16 on the road, then I'd agree it doesn't make much difference. But it appears to me that the home/away breaks are more uneven than that. And if that's the case, then even a small advantage in each home game can have a large cumulative effect. If a coin comes up heads 50.001% of the time, you can go broke (or make a fortune) if you flip it often enough.

UAH is at 16 home, 22 road, and that's using the Alaska exemptions for both of our trips. You can't compare us to last year (20-of-22 road) because of how our schedule was hosed. In our CHA days, we generally had 14 home games (10 league, two NC series) and 18-20 road games depending on how many NC games we could chock into our schedule. Clearly being in a 28-league-game schedule helps us, but we'll always have four NC weekends available, and we'll have five in some years. If this shift means that we'll see more non-conference games at home, I'm all for it. The years when we played just twelve games at home were hard.

[The SCSU series for this year was in place by Motzko wanting to get Nic Dowd time in front of friends and family.]

GFM <— doesn't know what to do with all the games.

 

[FlagDUDE08]

Re: Changes to Pairwise

Why not just use Krach?

I was wondering how long it'd take to see a Western Biased post like this.

 

[mnstate0fhockey]

Re: Changes to Pairwise

I've heard it on good authority that quite a few coaches weren't aware these changes were going to take place (which I think adds to Slap Shot's sentiment that this seems awfully knee jerk). I would expect to see quite a bit of push back on this and wouldn't be surprised in the least bit if there are significant changes to this after this season. Pretty bad when something gets pushed thru last minute and coaches aren't even made aware of the changes until it goes public.

 

[Patman]

Re: Changes to Pairwise

I've heard it on good authority that quite a few coaches weren't aware these changes were going to take place (which I think adds to Slap Shot's sentiment that this seems awfully knee jerk). I would expect to see quite a bit of push back on this and wouldn't be surprised in the least bit if there are significant changes to this after this season. Pretty bad when something gets pushed thru last minute and coaches aren't even made aware of the changes until it goes public.

I mean, its getting to the point where I am almost ready to ask my university if they want me to consult them on this stuff (probably pro-bono, or pro-swag, not offering yet).

To be honest, if things are looking the way they are, I'd advise them that this is good for them right now with the way the program is in that I think we can schedule better OOC than before (due to winning). For example, CC, DU, Mich, MSU, Penn State, I believe will all owe us trips in the next couple of years. As a fan, I'm against. The home/away thing is a sideshow... the main event is the bonus depending upon its actual degree.

I will admit that I haven't gotten into the nitty gritty of any of these articles.

 

[Priceless]

Re: Changes to Pairwise

I've heard it on good authority that quite a few coaches weren't aware these changes were going to take place (which I think adds to Slap Shot's sentiment that this seems awfully knee jerk). I would expect to see quite a bit of push back on this and wouldn't be surprised in the least bit if there are significant changes to this after this season. Pretty bad when something gets pushed thru last minute and coaches aren't even made aware of the changes until it goes public.

That runs counter to what the coaches asked for in Naples per the USCHO story. Maybe they weren't expecting it to get fast-tracked, but it would seem the vast majority of coaches asked for this.

 

[lugnut92]

Re: Changes to Pairwise

It really seems to be if they mean percentage points or not. I mean, playing six games against the top 10 and splitting would presumably mean a bonus of 7.5%

Now, I take that to mean an RPI of .5000 becomes .5750. Of course if I apply that in a lower decimal place it becomes different .5075 or .5008

Considering how much of a schedule can be against the top 20 in HEA and B10 those are going to add up quickly.

I haven't attacked this because I don't have games identified by home/road/neutral. Even then it'd require some annoying work but it wouldn't be prohibitive. However, further, is SOS measured off of the weighted win %? Does the game for SOS and SOSOS factors get weighted with the home/road weight or do they only count once?

Once you get past those gymnastics then you have to code it up. This can be a tricky math problem in the end.

I took it to mean that teams would get a few extra wins averaged in for their quality wins. If a team got 17 bonus points from QWs over their 34 games, they'd get an extra 0.5 win when their win percentage was figured at the end of the season.

 

[Patman]

Re: Changes to Pairwise

I took it to mean that teams would get a few extra wins averaged in for their quality wins. If a team got 17 bonus points from QWs over their 34 games, they'd get an extra 0.5 win when their win percentage was figured at the end of the season.

Yeah, that'd change things... god this stuff is silly/weird.

 

[mnstate0fhockey]

That runs counter to what the coaches asked for in Naples per the USCHO story. Maybe they weren't expecting it to get fast-tracked, but it would seem the vast majority of coaches asked for this.

I was told by a person who covers a program in the West (not a B1G program) that they were told by numerous coaches that they were unaware that this was going to happen. And this wasn't the only place I heard that. Don Lucia himself said at the B1G media day he was unaware. Sounds like this was far less of a consensus as some think.

 

[Freddie]

Re: Changes to Pairwise

I'm OK with adding weight to away wins, not sure I understand why they are dropping the TUC comparison.

 

[lugnut92]

Re: Changes to Pairwise

I'm OK with adding weight to away wins, not sure I understand why they are dropping the TUC comparison.

Because TUCs don't exist any more. Every D-I team is considered in the pairwise now. The QWB is somewhat akin to TUCs, but on a slope as opposed to the cliff.

 

[FlagDUDE08]

Re: Changes to Pairwise

Because TUCs don't exist any more. Every D-I team is considered in the pairwise now. The QWB is somewhat akin to TUCs, but on a slope as opposed to the cliff.

This does not seem to be true. The only thing that is going away is the Record vs. TUC comparison, and is being replaced by the top 20 bonus.

 

[lugnut92]

Re: Changes to Pairwise

This does not seem to be true. The only thing that is going away is the Record vs. TUC comparison, and is being replaced by the top 20 bonus.

From the NCAA memo:
"The selection criterion of Record Against Teams with an RPI of 50.00 or greater has been eliminated. Correspondingly, all Division I teams are compared against one another using the criteria of RPI (including QWB), head-to-head record, and record against common opponents."

TUCs no longer exist.

 

[goblue78]

Re: Changes to Pairwise

What you're missing, lugnut92, is that the new Quality Win adder is just TUC on a sliding scale, with a top 20 cutoff. The old TUC cliff won't matter much because being just in or just out of the top 20 won't matter much.

 

[FlagDUDE08]

Re: Changes to Pairwise

What you're missing, lugnut92, is that the new Quality Win adder is just TUC on a sliding scale, with a top 20 cutoff. The old TUC cliff won't matter much because being just in or just out of the top 20 won't matter much.

I can understand why they're doing it, because the TUC cliff drastically changes things. Are they still doing PWR points, but now for all of the teams?

 

[goblue78]

Re: Changes to Pairwise

Sounds like it: http://www.uscho.com/2013/09/20/sel...orn-from-coaches-wishes-committee-chair-says/
"Now, all 59 Division I schools will be compared against every other team using three criteria: RPI (including the quality wins bonus), head-to-head record and record against common opponents."

 

[lugnut92]

Re: Changes to Pairwise

Because TUCs don't exist any more. Every D-I team is considered in the pairwise now. The QWB is somewhat akin to TUCs, but on a slope as opposed to the cliff.

What you're missing, lugnut92, is that the new Quality Win adder is just TUC on a sliding scale, with a top 20 cutoff. The old TUC cliff won't matter much because being just in or just out of the top 20 won't matter much.

I'm not missing anything.