April 25, 2005
From The Mailbag
The Flat Earth Society, Win Expectancies, Running Wild, and Baseball Rhymin'Prospectus Triple Play: Los Angeles Dodgers, Minnesota Twins, San Francisco Giants
Could this be the year? In the AL East, the consensus among the masses was that the Yanks and Red Sox would again duel it out, many picking the Yanks to claim another AL East title. In the AL West, most picked the A's to finish a distant 3rd or 4th (primarily based on their offseason trades). And in the NL West, skeptics continued to question DePo and picked the Giants to claim that division. Sample size rebuttles aside, the Red Sox and Dodgers are leading their divisions and the A's are hanging close proving many pre-season (mainstream media) picks incorrect. If things (standings) continue to hold, do you anticipate many realizing the error of their ways and finally acknowledging the benefits of performance analysis?
Given how many naysayers have come forward join the anti-stathead brigade even in the face of a stat-oriented team (the Red Sox) doing the equivalent of turning water into wine while moonwalking across the Charles River, I'm going to guess that the answer is a big NO, no matter what the outcome on the field.
Never underestimate the power of the Flat Earth Society, nor its mainstream media guardians to take potshots at ideas they don't understand while protecting the old guard that gave them access back when ballplayers were lucky to make $30,000 and took offseason jobs plowing snow while walking to work uphill both ways, and you could spot the hitters because, dang it, they knew how to hit instead of talking walks like a bunch of sissies. It will take the passage of a whole generation of those crusty old media types before those complaints become less prevalent.
But as an analyst and a fan, I do look forward to seeing whether these teams (particularly the Dodgers) can continue to give their strategies a good name.
Nice job finally revealing the secret history of baseball. But I can't believe you left out the Soviet/Eastern Bloc penetration of the game as juiced moles like Steve Sax, Steve Swisher, Scott Sanderson and Steve Stone came on the scene (Stone given the broadcasting job to shut him up). What do you think was behind all of Sax's throwing errors? I mean, hell's bells, at one point there were TWO guys named Scott Servais/Scott Service playing at the same time...why do people need these things spelled out?
Glad you twigged to the rest of the conspiracy, even if I couldn't list every member. I didn't want to report on anyone living for fear they might use the libel laws as an excuse to silence we truth-tellers. At least you'll be safe when the revolution comes. Until then, beware the Juiced Moles.
I just read the BP 2005 section on Win Expectancies. Quite a job. Didn't understand everything, but think I got the gist. Three questions:
1 - There are a couple of subtle points here. The probability of winning the game, assessed at the top of the 2nd with the home team ahead 2-0, is the same whether the pitcher is a starter or reliever (and ignoring the quality of the pitcher himself).
What I think you're referring to is the part of the article where I contrast a visiting starting pitcher with a reliever entering the game in bottom of the first inning (none out, no score, bases empty). The win expectations are not the same for the pitchers in those two situations.
The difference is that a starting pitcher is committed to the game when the lineup cards are submitted. He is "in" the game as of that point. That decision is made before any of the game is played, and thus it's not the same as a reliever entering after the game starts. The win expectation for the reliever must take into account the then-certain knowledge of what occured in the top half of the inning.
2 - You're right -- Win Expectancy is inherently a context-sensitive measurement (indeed, that's what separates it from context-independent measures like VORP or EqA). It would be possible to measure something like "percentage of possible increase in win probability attained", and in some circumstances that might be an interesting measure to look at.
Another way to look at it, though, is to realize that teams who have significantly fewer high-leverage situations means that they benefit proportionally less from having an excellent bullpen. Tactical use of a team's best relievers in close games isn't terribly valuable if there aren't many close games. That insight might be useful to a GM deciding what he might be able to part with in a trade.
3 - We didn't have space or time to exhaust all of the possibilities of Win Expectancy in BP2005, so I chose to focus more on the pitching side, because of the close relationship with BP stats like Support Neutral Win/Loss records (SNWL). But check back -- we're not done with Win Expectancy, and applying it batting is on the list.
Thanks for writing, and for reading Baseball Prospectus.
How is it that a website so enamored of statistics and with all that brainpower could possibly publish this line:
Let me first admit that I only play Scrabble recreationally, but I'm amazed how many BP readers have pointed out the scoring impossibilities presented in the article. Here's the list of all players tied at 50 PSS, many of whom you'll find are also technically not worth as many points because of double z's and other single letters:
Kazuhiro Shibazki Kazuharu Yamazaki Kazuaki Miyazaki Jeffrey Randazzo Jeffrey Marquez Juan Velazquez Jose Velazquez Jack Krawczyk Jorge Vazquez
Thanks for the very interesting article on the effect of baserunners on batting success. A couple of thoughts, if I may:
If we add sacrifice flies to the denominator in AVG and SLG (and still remove IBB), we get the following numbers:
YEAR RUN AVG OBP SLG ISO MLVr 2004 3rd .267 .361 .418 .151 .039 2004 1st .280 .337 .440 .160 .037 2004 2nd & 3rd .253 .347 .415 .162 .000 2004 1st & 3rd .268 .337 .418 .150 -.003 2004 Empty .262 .325 .426 .164 -.018 2004 1st & 2nd .258 .327 .415 .157 -.032 2004 2nd .247 .340 .394 .147 -.045 2004 Loaded .254 .314 .408 .154 -.068
As you can see, while situations in which there's a runner on third fall back to the pack a little, they're still three of the four best situations for batters. The numbers above are very consistent with previous seasons as well, so while sacrifices and sac flies have some bearing on the results, there does still appear to be distinct advantages to batting with certain runner combinations on base.
Your second point is a good one that I had not considered. It certainly agrees with conventionally espoused tactics in those situations. Thanks for pointing it out.
Judging from your chart that breaks down 2004 plate appearances by baserunning situation, you've got about 40,000 PA's to work with where a runner is on first with second base open. It would be interesting to see how the trend pans out with, say, four or five baserunner classifications rather than three.
Breaking things up into five groups yields the following results:
AVG OBP SLG .295 .349 .458 .281 .335 .447 .277 .337 .441 .277 .336 .435 .274 .332 .428
While there's still some consideration that the faster runners bat in front of more impressive hitters, the trends are impressively steady and significant given the sample size.
Unless I misunderstood your article, it appears that your Phillies bet to win the world series is a big mistake. If you got 15-1 to win the pennant, that's good. But if they win the pennant, they are probably still only even money to win the next short series against an equally good team. I would expect to get closer to 30-1 on my money for the same team to go another round. Maybe 25-1 if I thought the pennant bet was really soft, but 20-1 just makes me want to put a larger bet on the pennant, and cheer for the AL team in the WS (so I don't feel bad about lost opportunities).
I'd say you're 100% right. And in fact the Phils were listed at 35-1 on the sheet I'd picked up to win the WS too, attractive odds. The teller told me the odds had fallen all the way to 20-1, I started to walk away...and circled back, because it was $20, and it was Vegas. As I said in the article, as a first-time baseball bettor, it was easy to fall into traps, and I imagine that was one of them.
If nothing else I'll have the cool element of knowing if they go all the way, I win an even $1,000. And I definitely won't have any regrets if they win the pennant and I "only" win $580, net.
I can't hear that Sublime song without thinking:I don't Edgar Renteria I don't Rafael Furcal I had Luis Gonzalez, but I Mike Lieberthal