Playoff possibilities

[ugarte]

I don't mind the sharp cutoff as much as you, though I acknowledge that it is a problem given how closely clustered the RPIs will be between, say, 22 and 30.  A weighting system would also be a reasonable way of capturing this criteria.

[ugarte]

The tournament is at the end of the year.  A selection criteria that captures, and gives "extra-credit" to, recent performance is a way of getting the teams playing the best hockey come tournament time into the tournament.  I think it is a valid criteria, as long as it is properly treated alongside other criteria that analyze the whole season.

There is a good reason that everyone living outside of Lincoln and Miami were furious when Nebraska got to play for the BCS championship. Nebraska was busy playing shitty football while Oregon and Colorado were gaining steam.  Something that penalized late-season failure in that case would have been nice.

[jeh25]

From a post I made on USCHO:

"I absolutely agree that L16 needs a SOS adjustment. However, I absolutely disagree that L16 should be removed as a component of the PWR.

Humor me with a little thought experiment. For this example, assume all 3 teams have the same SoS. Team A returns all its starters and has all its lines worked out and starts winning a decent number of games from week 1. Team B lost a lot of senior starters but has a few incoming blue chip recruits including a highly ranked goalie. Team C has a few amazing star players but lacks depth in it's 3rd and 4th lines.

Before the Christmas break, Team A wins 75% of their games while Team C wins 85% with their high powered offense. Meanwhile Team B struggles to figure out who should play on what line and only wins 40% of the time.

But after Xmas, we see a different picture emerge. Team A continues to win 75% of its games. But by this time, Team B has their lines figured out and has worked their new better goalie into the rotation and is tearing up the league, winning 90% of its games. Meanwhile Team C's star players are dinged up and the inability of the other lines to score causes their winning percentage to drop to 50%.

Thus, come tourney selection time, we have the following records.

Team A: 27-9-0
Team B: 24-12-0
Team C: 24-12-0

Now imagine we have 2 slots available. Who deserves to go? Obviously Team A deserves to go as consistancy deserves to be rewarded. But how do we decide between Team B and Team C? Personally, I'd rather see Team B go on to the postseason given that they have gone 18-2 since christmas as compared with Team C who has gone 10-10.

In effect, L16 rewards 3 things:
a) depth
b) coaching
c) player development"

[Greg Berge]

Or it may reward:

a) an easy conference (no NC games in the second half)
b) a fortuitous home/away breakdown
c) a well-placed hot streak being over-represented

This is a philosophical difference, but I still haven't heard a good argument for why a February game is intrinsically more important than a November game.

Having the concept of auto-bid in the tourny is the biggest boon of all to a team on a late hot streak.  That's enough for me.