Great Comments

There were some great comments the other day by a user with the name of bonifacemundane69 on Keith Laws blog. Here they are:

"Sean (Seattle): I keep hearing the M's weren't an 88 win team last year, but the only record I can find is 88-74. Isn't it possible that a bad stretch with a LOT of blowout wins (Weaver and HoRam starts) can mess with the run differential. They also had an outstanding bullpen, leading to many close wins.
Keith Law: People love to try to explain away divergences between Pyth records and actual W-L records. It doesn't work.
Jacob (Amherst): How come no one is talking about the Rockies in 2008? They lead the league in almost every batting category and were 1st in ERA in the 2nd half. I don't know if I've ever seen a defending champ get so little love.
Keith Law: They were 76-72 with two weeks to go in their season, and of course they played some (not all) of those remaining games against teams with nothing to play for. I think quality-wise they were closer to a .500 club than a true 90-win club, and people realize that."
So... on one hand, The M's were much closer to a .500 team due to the Pyth and any explanation by examination of incongruent data in the sub-set "doesn't work". Yet, on the other hand, the Rockies who fell exactly in line with their Pyth (90-73 v. 91-72) were also closer to .500 because of circumstances that caused incongruent data in both their W-L records AND their Pyth W-L? So, then the Pyth can be wrong as well, if there is a 3rd (undefined) parameter that "explains away" convergence in W-L and Pyth? Yikes!

Now 76-72 is an interesting place to pick-up the Rockies. That would be after Sept 15th, when they were coming off 3 losses dropping them from 76-69 to a much closer to .500 team. And it's very interesting that they played "some" teams that had "nothing to play for". Of course, those 15 games being against: (7)LAD, (4)SDP, (3)ARZ, (1)FLA. The Rockies were in 4th behind the Dodgers by 1 G ( (79-71 v. 78-72)when they played the 1st of those 7 games, LA had just as much to play for as COL. The Rockies won all 3 of a series against SDP. If the Padres had won 1, there never would've been a play-in game. The D'backs did win one, and it won them West by 1 game. Seems those were all pretty important games for all those teams. The Rockies were in no better or more advantageous position playing against teams "that had nothing to play for". If anything, they were playing 3 teams with everything to play for and 1 game against FLA. Nevermind the other NL WC contenders. The Mets played 13G against WAS & FLA in their last 15 and went 5-8. Maybe it's not so easy to just beat teams that have nothing to play for!
Nevermind the total subsequent dismissal of run differential in the sub-set. The Rockies in Sept went 20-8 and outscored opponents +63. Makes sense. However, in Aug, they outscored opponents +36 and only went 15-14. So, despite a +36 run differential, a .517 record was actually more representative of the Rockies "quality" than the Pyth? And that "works"?