2004-05 KRACH

[Trotsky]

In a move with Earth-shattering significance, I am incorporating KRACH into the TBRW Predicitions algorithm.

I'm searching for the final KRACH from 2004-05.  Assistance will be rewarded.  Yeah, even with corn nuggets (shudder).

[jtwcornell91]

[Q]Trotsky Wrote:I'm searching for the final KRACH from 2004-05.  Assistance will be rewarded.  Yeah, even with corn nuggets (shudder).[/q]

http://slack.net/~whelan/tbrw/tbrw.cgi?2005/krach

Or did you mean including the NCAA tournament as well?

Note that the ECAC-only KRACH http://slack.net/~whelan/tbrw/tbrw.cgi?2005/ecac.rrwp gives the same ranking as the final standings because the ECAC schedule is balanced.

[Trotsky]

Well, here's what I am aiming for.

I am thinking about a substitute for RS.  My first thought was KRACH, since it covers all teams unlike PWR (and also since PWR's mystery epicycles are dumb).  I prefer KRACH to RPICH or other rating systems because some physicist guy whose name escapes me has brainwashed me with that over and over again...

KRACH itself is difficult to explain, which is a drawback for the intentionally obtuse, "simple enough to put on a postage stamp," Pred.  But the middle of the RS gets very bunched up, and one stray win or loss can push you 3-5 places.  The crappiness of RS as a metric dramatically increases in the 4-8 spots.  I want a metric that is homoskedactically crappy.