Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of letters

[TonyTheTiger20]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Thanks, Grant. I sort-of figured that, but thought asking was worth it, anyway.

Cheers!

PS -- I added a decimal place to the BCI RPI for fun, since the top two are so close.

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Sorry, that was probably nonsense to most of you!

EDIT: Forgot to take into account the fact that UM has the 1.5 fewer wins over UMD. So yes, USCHO's PWR is undercounting those UMD QWB points for Wisconsin and it could conceivably flip the two teams. Made some clarifying edits above.

The part that follows the disclaimer is the actual nonsense. Both UM and UW have two wins over UMD. You're counting wins for the Badgers that they have yet to earn. Or you are assuming both will win out, in which case you're then shorting UM by one win over UMD. You stats guys are terrible at logic.

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Really not that important for the purpose of the question, is it?

No, but neither is Bob Costas, and that doesn't stop you from bringing him up over and over again.

 

[TonyTheTiger20]

Or you are assuming both will win out, in which case you're then shorting UM by one win over UMD.

Lol yes I am -- wouldn't that make UM 2-1-1 and UW 4-0-0, a difference of +1.5 wins for Wisconsin?

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Lol yes I am -- wouldn't that make UM 2-1-1 and UW 4-0-0, a difference of +1.5 wins for Wisconsin?

No. If you're going to do the stats thing, you have to get your stats correct. The Gophers are 2-1-1 vs UMD with the teams' fifth meeting taking place today. They played a NC game in January at the Minnesota Cup.

 

[TonyTheTiger20]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

They played a NC game in January at the Minnesota Cup.

Ohhhhhh yes, that's right.

This was really just an example, it wasn't intended to be a rigorous proof lol. The point is, any difference in games won against any teams in the top 12 could conceivably make USCHO's and BCI's Pairwise differ if UM/UW's RPIs (or anyone else's) are close enough.

 

[robertearle]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

This was really just an example, it wasn't intended to be a rigorous proof lol.

Yes, while my question was triggered by the current UW-UM-UMD numbers and situation, it really was meant to be more generic and hypothetical than all this.

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

The point is, any difference in games won against any teams in the top 12 could conceivably make USCHO's and BCI's Pairwise differ if UM/UW's RPIs (or anyone else's) are close enough.

Got it.

Maybe the real reason that the NCAA uses RPI/PWR instead of a decent ranking system is that it gives us plenty to talk about. I imagine if they ever do replace it, then they'll go to something even worse, like what they started to use for basketball this year.

 

[D2D]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Pending the outcome of the Minnesota/Minnesota Duluth game today, I see that the Gophers and Badgers temporarily have exactly the same RPI, at .6432. Does anyone remember the two top teams in the PairWise having the same RPI, down to the 4th decimal?

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Even I'm not enough of a geek to remember something as random as that.

 

[robertearle]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Even I'm not enough of a geek to remember something as random as that.

I guess I'm close, but not quite there. I can remember discussions of 'beyond four decimal places' but not the specific teams or year, etc.

Maybe Grant does (!)

 

[D2D]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

With their win this afternoon the Gophers now vault ahead of the Badgers in the RPI and lead by an insurmountable 0.0043.

 

[TonyTheTiger20]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

I guess I'm close, but not quite there. I can remember discussions of 'beyond four decimal places' but not the specific teams or year, etc.

Maybe Grant does (!)

Oh I'm sure it's happened mid-season before... couldn't tell you a specific time, but I feel like I remember it happening! haha

 

[Still Eeyore]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

With their win this afternoon the Gophers now vault ahead of the Badgers in the RPI and lead by an insurmountable 0.0043.

KRACH gives Minnesota a 49.9% chance of beating Wisconsin in a single game.

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

KRACH gives Minnesota a 49.9% chance of beating Wisconsin in a single game.

So you're saying there's a chance?

 

[Still Eeyore]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

USCHO's PWR page is showing teams' records without the games played tonight. Have the ratings been updated? I had remembered BC, in particular, is being higher than the 10th (9th w/o St Anselm) that they are shown in now, but it seems strange that the ratings would be recalculated but still show the outdated W/L records.

Edit: Never mind. Looking at Grant's calculator, it looks like the USCHO page hasn't been updated at all.

 

[TonyTheTiger20]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

I don't know what's up with USCHO, but I updated BCI's.

BC is ... ok in the Pairwise, they're basically where they were before the loss in that they will be in 7th if they beat UNH and re-flip the BC/BU comparison back in their favor. The problem now is the gap up the Princeton in 6th is huge and the gap back to OSU is basically gone. BC really needs one of the 4-5-6 teams to lose twice and BC to win out to have a shot at staying east, unless the committee does what they did a few years ago and send BC to NU for no reason to avoid hotel rooms. And, of course, BC would need to hope there are no conference tournament surprises on top of that.

God, what a disaster this whole season has been. Already essentially 0-2 on trophies (Beanpot & WHEA regular season), looking like we're not even good enough to have a shot at the WHEA tournament trophy, and might not even get to play for the 4th one in the NCAA tournament unless we win the 3rd one too.

 

[TonyTheTiger20]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

BC is ... ok in the Pairwise, they're basically where they were before the loss in that they will be in 7th if they beat UNH and re-flip the BC/BU comparison back in their favor

Well, BC has a much bigger problem.

The bad news is if BU wins the Beanpot, that would lock in a 1.000 co-op record against Harvard for BU and a 0.000 record against Harvard for BC, pretty much locking in a BU comparison win over BC and knocking BC pretty much definitely out of the tournament even if BC wins out, barring something crazy with Princeton/Clarkson/Cornell.

The good news for BC though is that since BC and BU are the 2/3 seeds in Hockey East, they'll almost definitely meet in the tournament semifinals if they both get that far, in which case BC would be able to control their own destiny again by evening up Head to Head and taking the whole co-op point out of the equation entirely. And if they don't meet in the semifinals, it means either BU lost 2 of 3 against UNH, which would probably put BC back on top anyway, or BC lost two of three in their quarterfinal in which case they obviously don't belong in the tournament and probably would be out on RPI anyway.

In any case, needs to really hope Harvard beats BU next week, otherwise BC is going to be on the outside looking in at least until the Hockey East semifinals.

 

[ARM]

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

Re: Fun With Numbers 2019: PWR, RPI, KRACH, GRaNT, WCHODR, & other assortments of let

So BC is pretty much definitely out of the tournament, except that there are a ton of different scenarios where it will still be in. Thanks for that clarification.

 

[TonyTheTiger20]

So BC is pretty much definitely out of the tournament, except that there are a ton of different scenarios where it will still be in. Thanks for that clarification.

Sorry ARM, I'll be sure to keep all discussion of the nuances of the Pairwise out of the Pairwise thread from now on. That's my fault.