New postseason odds (August 17, 2003)
Fanhome's dackle checks in with his sim...
--posted by TangoTiger at 05:38 PM EDT
Posted 12:21 a.m.,
August 19, 2003
(#1) -
Jeff
If the Royals have a 70% chance of winning their division, then the Braves must have, oh, a 140% chance of winning theirs.
Posted 12:35 a.m.,
August 19, 2003
(#2) -
Noffs
BP's seem more reasonable to me...
Posted 12:39 a.m.,
August 19, 2003
(#3) -
kamatoa
I still think Dackle's got the Yankees overranked - even at five games back, the Red Sox offense is better than the Yanks', and New York's pitching has started to lose some wheels. I doubt they'll put another two games on Boston, as Dackle predicts. Converted into odds, Dackle has 13-to-1 against Boston taking the division. Although I'm not sure if the Sox can squeeze out all five games, that would be a tempting bet to take.
Posted 3:40 a.m.,
August 19, 2003
(#4) -
Mark
The Red Sox (71-53) have a 58.6% chance of taking the wild card, and the A's (71-53) have a 28.6% chance? And the Phillies are at 62.5% of winning the wild card, and the Marlins 20.4%, when the Phillies are 0.5 games up? There must be some serious strength-of-schedule effects going on here.
Posted 8:37 a.m.,
August 19, 2003
(#5) -
Scoriano
Yanksare +5.5 and 6 in the loss column. They also have all their Tiger games left (the Red Sox were, IIRC, 9-1 vs. Detroit and are finished with them).
Posted 11:28 a.m.,
August 19, 2003
(#6) -
tangotiger
It's worth noting that BP and Dackle offer very close odds for everything except the 2 central divisions.
The differences between approaches are:
BP: head-to-head matchups do not have binary approaches
Dackle: does not do a good job at valuing a team's true talent level
What does this mean? As long as the number of games remaining is large enough, the BP estimates are more reliable. However, as soon as you've got say 20 games left, the BP estimate would have to be discarded, and Dackle's estimates would take precedence. You can't have the possibility as BP has it that the Yanks and Redsox can win the same game and hope it evens out over such a small number of remaining games.
Posted 12:58 p.m.,
August 19, 2003
(#7) -
Matthew(e-mail)
(homepage)
Mark,
I just did a side-by-side comparison of the Marlins and Phillies remaining schedule. Removing common matchups and head-to-head games, here are the differences in their schedules:
Florida Away: Colorado 3 games, San Francisco 3
Florida Home: Montreal 2, Pittsburgh 3
Philadelphia Away: Montreal 1, Milwaukee 3, St. Louis 3
Phila. Home: Boston 1, New York Mets 1, Cincinnati 3
It appears to me that Philadelphia's away schedule is easier than Florida's (Colorado is 41-19 at home this year), thus accounting for the strength of schedule effect you've observed.
Posted 2:37 p.m.,
August 19, 2003
(#8) -
Minnesota Twins
We'll stick with Baseball Prospectus, thank you very much.
Posted 3:20 p.m.,
August 19, 2003
(#9) -
Mark
Matthew: Interesting, it's still amazing that the SoS effect is so strong (more than tripling the chance, given virtually identical records). Also I think Colorado's home record is a bit of a mirage, and SF hasn't been playing all that well lately. I notice the Baseball Prospectus projections have slightly more favorable, but still similar, numbers: 60.4% for Philly and 29.4% for Florida.