Add new comment

A little reminder for those who may not recall this tool from several years back (or for new readers):
Pythagorean W% was simply RS^2 / (RS^2 + RA^2) - this is what Doc quickly refers to to make his basic point that the Mariners are better than their W/L record if you account for the luck of how run scoring happens to have been distributed.
PythagenPat W% is a metric that varies the exponent (from 2) based on the run scoring that is happening per game using the formula X = ((RS + RA) / G)^0.285 basic on an empirical curve-fitting relationship found by another well known saberdude who went by the web handle "Patriot"
PythagenMatt is PythagenPat applied to each game individually and summed over the course of a season (with a minor linear adjustment to account for the fact that doing it this way pulls everyone too far toward .500 due to the lack of decisive 0s and 1s such a method produces).
I'm afraid, Doc, in this case, PythagenMatt doesn't agree with the back of the envelope Pythags. Seattle has benefited from four blow-out wins, but has also been shut out twice already and has decisively lost a number of other low-scoring games. The net result is a club on an 7-9 pythagenMatt track (7.2-8.8 to be exact).
Caution bulbs though...it's a very short number of games, I am not sold that the correct thing to do when a team gets shut out is give it a 0 for its win probability for that day, and some of the final scores are misleading early on here because the Mariners have piled on late and have gotten their brains beat in late on different occasions.
I am inclined to attempt an even lower-level Pythagorean W% estimate based on inning by inning run differential or at the very least...using Bayesian logic for each game (rather than assuming that a team scoring zero runs would score zero runs if it played identically in 1000 rematches under the same conditions...it is probably better to a assume a team is average and let the runs scored and allowed deviate the projection from average).
I would, however, say that PythagenMatt and my visceral reactions to the team are lining up well right now. If we didn't anticipate any roster changes at all, we should expect this team to go about 73-89...the pitching is LOUSY (and was lucky early in the season) and the hitting is average and streaky. Luckily...we expect to get Iwakuma, Paxton, and Pryor back soon and Walker MIGHT also come back.