Master_Of_Districts
Registered User
As I understand it, however, SRS does not consider the scoring environment. Playing in a lower-scoring era, goal differentials will be more tightly bunched, drawing the best and worst teams toward zero relative to higher-scoring times.
A great team playing in 1986 can be +110 while in 1999 they might be +80 (a difference of about 0.37 in SRS as I understand it), with the same point total, because there were simply fewer goals scored in 1999. Their SRS, assuming equal SOS, would be lower in 1999 than in 1986, simply because of the lower scoring environment.
It's no coincidence that the absolute values of SRS in the late 1990s/early 2000s are significantly lower than in the 1980s, which makes the statistic of dubious value for comparing teams from different times.
Which in turn suggests that a scoring level-adjusted SRS would be desirable. Anyone done any work on that? If not I might have to.
I have.
1998-99 DAL
GF - 236
GA - 168
Schedule Difficulty (1=hardest) - 25th
ADJ GF - 234.0
ADJ GA - 170.2
ADJ Pythagorean Expectation - 0.634
ADJ Pythagorean Expectation Rank - 1st
1985-86 MTL
GF - 330
GA - 280
Schedule Difficulty (1=hardest) - 5th
ADJ GF - 329.6
ADJ GA - 272.7
ADJ Pythagorean Expectation - 0.582
ADJ Pythagorean Expectation Rank - 3rd
Schedule difficulty correction was based on 4 iterations.