PWR Math

[KaMiGo]

With the Big Ten adding Notre Dame and moving to a new playoff format next season, I was wondering what implications it might have on PWR math. I'm not an expert by any stretch of the imagination, so I'll put it out to the community to see how it would work.

The new format is a 3-weekend event, all games hosted at campus rinks (well maybe not all, with Wisconsin and Ohio State having multi-purpose facilities, but that isn't important for this discussion)
Weekend 1:
#1 gets a bye
3 highest seeds host 3 lowest seeds
This will be a best of 3 weekend series

Weekend 2:
2 highest seeds host 2 lowest seeds
This will be a single game

Weekend 3:
Highest seed hosts lowest seed
This will be a single game


So the question is this: what is the difference in going 0-1 in a single game format vs going 1-2 in a 3-game format? Do the 2 losses cancel out any gains from the win?

 

[FlagDUDE08]

Re: PWR Math

With the Big Ten adding Notre Dame and moving to a new playoff format next season, I was wondering what implications it might have on PWR math. I'm not an expert by any stretch of the imagination, so I'll put it out to the community to see how it would work.

The new format is a 3-weekend event, all games hosted at campus rinks (well maybe not all, with Wisconsin and Ohio State having multi-purpose facilities, but that isn't important for this discussion)
Weekend 1:
#1 gets a bye
3 highest seeds host 3 lowest seeds
This will be a best of 3 weekend series

Weekend 2:
2 highest seeds host 2 lowest seeds
This will be a single game

Weekend 3:
Highest seed hosts lowest seed
This will be a single game


So the question is this: what is the difference in going 0-1 in a single game format vs going 1-2 in a 3-game format? Do the 2 losses cancel out any gains from the win?

Let's extrapolate to a 34-game season, let's say they go 20-14 in the regular season (.5882). With the one game format, 20-15 (.5714). With the three game, 21-16 (.5676). Remember that, prior to weighting, this accounts for 25% of your RatingsPI. Not to mention the effect on other teams with Common Opponents record.

 

[Agganis]

Re: PWR Math

prior to weighting

Including weighting if you're the road team you'd rather lose a round 2-1 than 1-0, but if you're a home team you'd probably rather lose 1-0 than 2-1.

 

[Numbers]

Re: PWR Math

And, in general, that 0-1 versus 1-2 thing is going to affect the final PWR less than whatever Notre Dame's OOC record is next year. Since, as above, win% is only 25% of RPI, and oppwin% and oppoppwin% matter far more, then what your conference mates do in the OOC games is more valuable than adding or subtracting a 1-1 record from your own.

 

[Patman]

Re: PWR Math

once you get to tournament format what happens you cap the number of losses. Hockey gets a bit weirder with its best of 3 and is better visualized for basketball.

Generally, you can't move down much. Sure, a bad loss is a bad loss but you only lose once. Your security before and after can be changed but if you were well secure before you aren't going to become unsecure.

On the other hand, if you win a lot you can move up a fair bit. Now, of course other teams that tend to win will also be winning so this gets obscured slightly. A longer playoff means that generally you can get a bit more forward motion.

Now, that being said, with a best-of-3 you can lose twice. The principles still applies, nobody is losing four games in the post-season (though various formats open it up to losing 3).

----

All the above being said, any change in conference tournament format will not dramatically change the world. Best of 3 is the biggest impact because of the ability to take more losses.