Cheech Tremendous Posted October 5, 2007 Report Posted October 5, 2007 I was just thinking about the Diamondbacks and the potential for the next few years. I honestly wouldn't be surprised if they win multiple world series titles. Their lineup is young and scary good. A full year of development for some of these guys is going to mean a lot and Justin Upton isn't that much now but he's supposed to be the best of the group. And he's 20. It's kinda nuts that they are so successful now because they could be scary in 2 years. Arizona also outperformed their pythagorean record by about 11 wins this year, so they are extremely lucky to even be in this position. They are going to have to improve a lot in the offseason to even have a hope of getting back to the playoffs next year.
The Man in Blak Posted October 5, 2007 Report Posted October 5, 2007 The pythagorean record has got to be one of the most overrated analytical tools in the sabermetric toolbox. It doesn't account for extra innings, it isn't park adjusted, it holds an inherent bias against winning home teams (who don't have as many innings to score runs as the losing road team), it obscures changes in the roster over the sample period, and it doesn't really deal with "noise" in run distribution (blowouts vs. one-run games). Everybody's so quick to point out that pythagorean record has more predictive value than overall W-L record, but how much more predictive value does it really give you? Recent numbers cited at BTF indicate that first half pythag has had a 51% (moderate, but not really strong) correlation with second half wins in this decade, which is only 4% better than 1st half W-L record. And never mind that the run environment and managerial tactics in the regular season are quite different than what we see in the playoffs, which is a crucial point to consider when you're dealing with a team like the 2007 Diamondbacks that has been more prone to losing blowouts games by using absolutely terrible relievers in mop-up situations. These crappy relievers, who aren't even the same league as the "core" part of the Arizona bullpen, won't likely see the same percentage of innings in the playoffs, where long-term resource management (i.e. saving the arms of your best relievers) takes a backseat towards getting any chance for a win. Pythagorean record does have some utility as an estimator for future success, but it is far from law in any case, especially in the context of a team like the Diamondbacks, whose bullpen usage directly led to an abnormal run distribution throughout the regular season.
Ginger Snaps Posted October 5, 2007 Report Posted October 5, 2007 Okay non-Yanks/Sawx fans, time to jump on that Indians' bandwagon with me. They are our only hope. I'm VERY happy that the Indians won. And decisively. I want them to sweep so bad. I like you now although I'm kinda hoping that series goes 4 games with Cleveland winning. That way if the Sox sweep the pitching match-ups at first will slightly be in their favor. I'm from Ohio so they're one of the few teams I can really get passionate about. And I hate the Yankees, so it'd be really satisfying if Cleveland swept. I'd have loved for it to be Cleveland / Cubs in the Series (cause Chicago is the other team I can be passionate about) but that's not going to be happening. I just hope the Indians can pull it off, I don't want to see the Yankees go any further and I will be disgusted if they lose to the Saux too. Those two are probably the teams that I hate the most. I'm going to stop caring about this postseason altogether if it's Saux / NY again, unless somehow the Cubs pull off a miracle comeback.
Cheech Tremendous Posted October 5, 2007 Report Posted October 5, 2007 The pythagorean record has got to be one of the most overrated analytical tools in the sabermetric toolbox. It doesn't account for extra innings, it isn't park adjusted, it holds an inherent bias against winning home teams (who don't have as many innings to score runs as the losing road team), it obscures changes in the roster over the sample period, and it doesn't really deal with "noise" in run distribution (blowouts vs. one-run games). Everybody's so quick to point out that pythagorean record has more predictive value than overall W-L record, but how much more predictive value does it really give you? Recent numbers cited at BTF indicate that first half pythag has had a 51% (moderate, but not really strong) correlation with second half wins in this decade, which is only 4% better than 1st half W-L record. And never mind that the run environment and managerial tactics in the regular season are quite different than what we see in the playoffs, which is a crucial point to consider when you're dealing with a team like the 2007 Diamondbacks that has been more prone to losing blowouts games by using absolutely terrible relievers in mop-up situations. These crappy relievers, who aren't even the same league as the "core" part of the Arizona bullpen, won't likely see the same percentage of innings in the playoffs, where long-term resource management (i.e. saving the arms of your best relievers) takes a backseat towards getting any chance for a win. Pythagorean record does have some utility as an estimator for future success, but it is far from law in any case, especially in the context of a team like the Diamondbacks, whose bullpen usage directly led to an abnormal run distribution throughout the regular season. Excellent post, MIB. I still like to use them as a tool to measure underlying tendencies, but I've made the point here many times that a good team should overperform its record to some degree and vice versa. A good shut-down closer and a young team is a perfect storm for outperforming the PWL (e.g. Arizona). Next year will be a true test for the Diamondbacks. How will they perform with regards to their pythagorean record? Is their performance a function of the team's construction, or an exception that proves the rule? EDIT: By the way, run distribution and win expectancy makes my head hurt.
cabbageboy Posted October 5, 2007 Report Posted October 5, 2007 I agree that stuff like point differential and the Pythagorean record isn't that important in baseball (whereas I do think it is often useful to look at in football). Over 162 games there can be so much that happens, maybe a few pitchers get hurt and the replacement guys give up more runs, maybe like with the D'Backs they win close games but get squashed in several losses.
Dobbs 3K Posted October 5, 2007 Report Posted October 5, 2007 I agree. The way some statheads talk, the Pythagorean record for a team should almost be the one in the record books, rather than the actual, real life, win-loss record. "Oh, that team just over-performed, or were lucky." "Gee, are you really sure? Are you sure your precious stat metric wasn't just flawed to begin with, since you're trying to predict the future in a sport full of intangibles?"
EVIL~! alkeiper Posted October 5, 2007 Author Report Posted October 5, 2007 The pythagorean record is useful because since you can reasonably project win/loss records with it, you can evaluate individual players' direct contributions to winning games. As far as predictive value it's good from year-to-year, but not so good in-year.
Cheech Tremendous Posted October 5, 2007 Report Posted October 5, 2007 I hope that the September version of Kelvim Escobar shows up tonight.
CanadianChris Posted October 5, 2007 Report Posted October 5, 2007 This trivia question they just asked on TBS may be the stupidest one I've ever seen: How many players have gotten three hits with two outs and runners in scoring position in a postseason game?
strummer Posted October 5, 2007 Report Posted October 5, 2007 If they pull this out that is your standard Yankee 1996-2001 playoff victory. Oh man
HarleyQuinn Posted October 5, 2007 Report Posted October 5, 2007 Busting out the bug spray to fend off the things... awesome.
KingPK Posted October 5, 2007 Report Posted October 5, 2007 They're breaking out the Off at the Jake right now.
CanadianChris Posted October 5, 2007 Report Posted October 5, 2007 Blecch...didn't need the close-up of Joba's neck in HD there.
EVIL~! alkeiper Posted October 5, 2007 Author Report Posted October 5, 2007 This trivia question they just asked on TBS may be the stupidest one I've ever seen: How many players have gotten three hits with two outs and runners in scoring position in a postseason game? Sounds like someone at TBS subscribed to Baseball-Reference.com's Play Index. Despite the bugs, this is easily the best game of the postseason thus far.
strummer Posted October 5, 2007 Report Posted October 5, 2007 1996-2001 all over gain. Bunting rules! Edit: or not. Tie game!!!
HarleyQuinn Posted October 5, 2007 Report Posted October 5, 2007 Wow! That's one way of scoring the game tying run... go Indians go.
Hawk 34 Posted October 5, 2007 Report Posted October 5, 2007 I just tuned in, what's with all the bugs?
CanadianChris Posted October 6, 2007 Report Posted October 6, 2007 We've found Joba's weakness. The kid doesn't like bugs.
HarleyQuinn Posted October 6, 2007 Report Posted October 6, 2007 Seriously... that has gotta suck playing like that. Perspiring, getting bitten by a hundred swarming bugs... ugh.
CanadianChris Posted October 6, 2007 Report Posted October 6, 2007 I know I've seen a game called due to bugs, I just can't think of when.
Danville_Wrestling Posted October 6, 2007 Report Posted October 6, 2007 Yea, those bugs present an icky situation for any pitcher. I'd hate to be on the mound right now.
Slayer Posted October 6, 2007 Report Posted October 6, 2007 Cleveland's trained army of gnats will win the game!
EVIL~! alkeiper Posted October 6, 2007 Author Report Posted October 6, 2007 The big problem for the Yankees is that even if they get the win, Torre might be gun-shy on using Chamberlain again in a big spot.
Mr. S£im Citrus Posted October 6, 2007 Report Posted October 6, 2007 This is one of the seven signs, isn't it? THE END IS NEAR!
CanadianChris Posted October 6, 2007 Report Posted October 6, 2007 It's going to start raining frogs sometime this inning.
CanadianChris Posted October 6, 2007 Report Posted October 6, 2007 Who's it gonna be...Hurray-Rod or Go Away-Rod?
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now