In the middle video with the first base side view he's almost doing a no stride, but I also notice that this is an older Ted Williams than in the other videos. Even though the stride is short he does seem to get all of his weight into the front leg by kind of rocking down on it. Another thing I notice, in the last video where he's hitting the low pitch, is that his swing is actually a perfect golf swing in that instance.

I'm going to violate the laws of the universe and suggest that pitching is harder than hitting. Mentally you have to stay locked in for a longer period of time. You're the first mover so you have to contend with "yips." You're performing your job under conditions of accumulating fatigue. Or to be fair to hitting, hitting is more physically precarious, and probably leaves less margin for error in execution, but hitting is less mentally and emotionally difficult
.

If the Bosox ever let Ortiz retire, they can carry redundancy in the infield, using DH for half rest days like a normal team. Then you could have Betts, Pedroia, Middlebrooks, and Bogaerts sharing DH and the 4,5, and 6 positions. Of course, you're still in a position where there's no place for Cecchini, Coyle, or Marrero and it seems likely one of those guys will be good enough to play. Maybe you bundle some of these guys together into a prospect backed security and flip it for pitching.

WRT your answer to the question of what teams you would choose to be in your division:
If you were a small market team--say, on the east coast--why not choose BOS, NYA, NYN, and PHI to be in your division? This guarantees you're going to have big market gate receipts to something like 36 of your 81 home games. Then if you get half the attendance at the rest of your home games, your gate receipts will be the equivalent of 36+54/2=63 big market home games, which is 7/9 of what the other four teams in your division earn. So you've minimized your revenue disadvantage. I'm not sure to what extent you could make a similar argument for television revenues, etc. Furthermore, as your own fans don't expect you to win, you have the strategic advantage of being able to maximize the long run value of your homegrown talent, rather than having to sell pieces to compete every year. After an initial period of losing and building up your farm system you would may find yourself in a Rays-like position of being on an even competitive footing with big market teams in most seasons. Maybe?

Just checked your standings... The Yankees are almost +5 relative to their 2nd and 3rd order winning pythagorean %'s. That's the real story so far this year, not the individual performances but the chasm between the record and the lack of performances justifying it.

Only two players in the Live Ball Era have managed to pick up a golden sombrero and smack two home runs in the same game.
I can't parse this sentence. It sounds like it means only two players have ever had a two homer game, but obviously that's not what it means. Please re-write this article.

Well done. This was a great article. And I agree worst system would be a great comparison piece. You're always getting surplus value from your farm system, the question is how much. So if a great farm system nets you 180MM or whatever, it would be good to know what a bad farm system nets you too to give a sense of the incremental value of improving a system.

The team side comps of Lannan, Bailey etc. were ridiculous. As soon as I saw those names, I knew there was no case to be made for the team. A welcome relief to have a hearing that I could decide easily to make me feel smart. Most of the time I'm scratching my head and feeling stupid.

The panel cites Choo's status as a "national hero" in Korea as part of the reason for their award. But the player's case never makes this argument. Is there any rule in arbitration hearings that states you have to use the evidence that was presented--kind of like the instructions that juries received? Or is it acceptable for the arbitrators to go out and use any information they want?

+1 to the idea of removing the pitcher's PA's from his team's BABIP (I wouldn't be surprised if this is what JC did). Within the pitcher's PA's, the relationship between the two statistics is tautological, as we've established, so including them is going to bias the observed relationship toward that y=1-x line. Removing them gets us closer to the relationship we're really trying to study.
Anyway, thanks all. I understand the article's argument much better now as being somewhat about the lack of steepness in the line's slope and somewhat about the high variance of the team-conditional distribution.

Thanks for pointing me to your glossary. Here are the definitions I found:
BABIP = (H - HR) / (AB - SO - HR + SF)
Defensive Efficiency = 1- (H-HR)/(AB-SO-HR+SH+SF)
So, for all intents and purposes, these two numbers are mathematical inverses (the strange omission of SH from the BABIP formula notwithstanding). Within the same set of plate, appearances, if you compute the two numbers, BABIP = 1- DE.
Each data point of your analysis includes a number for BABIP and a number for DE. But DE does not appear to be 1-BABIP (that is, the points don't all lie along y=1-x. But I don't understand why. Are you computing DE over a larger sample of PA's than the sample used to compute BABIP? Maybe one is computer for just the PA's against the pitcher and the other is computed for all of the defense's PA's? That would make sense. You can ignore this if you think I'm getting too technical but it seems like an ambiguity that obscures the whole focus of your analysis--to me. So maybe I'm not the only one wondering this. No worries either way.

Aside from the fact that Defensive Efficiency lumps errors and hits together into "outs not made," BABIP is just the inverse of Defensive Efficiency. Isn't it? As such, the correlation of the two measures within a sample is kind of meaningless, isn't it? A high DE and a low babip go together almost by definition so that finding doesn't really say anything about how a defense impacts a pitchers BABIP. It just says that if a pitcher has a high babip he probably also has a low DE. But who cares? What am I missing?

I like the Stephen Drew signing. It completes the pattern of filling every Red Sox position with major league average level talent. Hopefully this experiment will result in a "workingman's juggernaut" of higher-than-median expected performance across the board, fueled by high morale borne of satisfaction with communistic ideals. Three cheers for the reddest Red Sox team in many years!

An interesting hypothesis. Of course one year of data really only provides anecdotal support. Makes me really curious how a more detailed an analysis would look though. Perhaps you could do a similar analysis to the one BP did with high Guillen ratio teams. I.e. ask the question, "does the post-season reduce the winning percentage of high payroll teams less than it reduces the winning percentage of low payroll teams?" Sample size would be a bit of a problem because there are so few playoff games, even if you go back as far as the modern era of free agency, say post-collusion in the early 90's. You get even less data if you try to capture the period of time that Selig's talking about, which would not include the late nineties and early 2000's when the Yankees had their most dominant modern run. I wouldn't be surprised to find that our impression that big market teams do better can't be supported at a high level of statistical significance. Thanks for the food for thought.

From the New York Times, "While most players dump their bats in cylindrical canvas bags when they are not using them, Suzuki neatly stacks his best eight bats inside a shockproof, moisture-free black case"
That's awesome and ninja-like, and is consistent with many other bad-ss behaviors practiced by Ichiro that belie his cultural link to the mythic Bushido mercernaries of feudal Japan. I suggest we add 287 hits to his career total to account for the fact that he is this awesome. Don't even wait until he retires. Hall. Of. Fame. Now.

It looks pretty good. And it's a great story anyway so somebody would have to really screw up to make it bad. Questionable choice on the soundtrack though. Maybe there's a better world where you can make this movie and Jay-Z doesn't compose an obligatory song but I fear that only my children's children will to see that world. I like Jay-Z as much as the next guy but enough already.

Great article. Your explanation of the situational use of shifts against a left handed hitter makes sense of something I never really understood before.
Obviously I don't have the sample sizes for the batting average splits you cite but Ryan Howard only has about 4666 career plate appearances. Once you break that out into a 2 or 3 way split the standard error for a difference between the split batting averages might be 20 points or more. Howard is one player out of thousands and his "skill" probably isn't even two standard errors from no skill at all. Statistically, this isn't even enough of an outlier to bother taking notice. Still, a great article well worth the read.

So we're looking at cross sectional variance here... What about within-player longitudinal variance? It could be that the users are hitting more home runs and the nonusers are retiring and getting demoted to the minors. In other words, selection into the major leagues is highly endogenous to steroid use. In that scenario, maybe everybody in the majors has more home runs... And the non-steroid users lose roster spots and don't really impact your cross sectional variance.
Trying to think of the best approach to showing this... I guess one way to show evidence for this story would be that player performances vary more relative to their own ex ante expected performances. So based on history up to 1995, you expect so many home runs from a player in 1996, and the variance of his actual home runs in 1996 relative to what you expected is larger than that variance would've been using a similar forecasting method for 1983 using historical data through 1982.
The sophisticated version might be a panel data model of player level home run totals with player and year fixed effects. Use, say, 1970-2012. Then you test the null hypothesis that the variance of the white noise errors is larger for the 1993-2006 period than for the preceding and following years. If you reject the null that the variance increases for the steroid years, that supports the "steroids mattered" theory.

+1. The Orioles are playing with house money right now. Who knows what you're going to get from Machado at this stage in his development? Maybe nothing. Take a risk and try to get lucky. Sabr-metrically, I don't hate it. As a fan, I love it.

Coming at this from another angle... If you assume every plate appearance is an independent random trial, you can actually compute confidence intervals around (e.g.) a batting average. Roughly, that line of reasoning leads you to the conclusion that 100 at bats gives you a confidence interval of about +-.100. 400 at bats gives you a confidence interval of +-.50, 1600 AB's: +-.25. For Wade Boggs' entire career of about 9000ish at bats, the confidence interval is still +-.10 around his .328 average. And I expect that these confidence intervals are, if anything, too narrow. A plate appearance is not an independent random trial because players run hot-and-cold, and their abilities improve and decline over time, so there is also unaccounted for time series variation. Very similar logic would apply to something like a walk-rate or a homerun rate, which we normally think of as more reliable (and which stabilizes faster according to your methodology). The first conclusion I'd draw is that we are kidding ourselves when we report the 3rd digit of detail on anyone's batting average. We should just say Robinson Cano is hitting 32% (plus or minus 7%). It'd be nice to see somebody tackle the issue of how useful statistical analysis can be in baseball given that we're typically drawing inferences from such imprecise statistics.

How about Adam Dunn? For this season, his true outcomes as a share of plate appearances must be close to the all-time record. That's a match up that will probably happen shortly.

The really smart question here would be whether home runs are serially random, or are they bunchier (positive serial correlation) or less bunchy (negative serial correlation) than they would be if they were serially random. Whip out your ARIMA modeling tool kit. Nonetheless I think this article does a good job showing a more intuitive analysis that is related to the "smart" analysis. Given the findings of this article, I would be surprised if it is possible to reject a null hypothesis of no serial correlation out to a lag of, say, 50 plate appearances. Surprised, but not shocked. It's still worth doing.

"The game is dragging, and the talent on the field isnâ€™t exactly a plate of chicken fried steak with delicious cream gravy held close enough to tempt but far enough away to limit my accessibility." Huh?

Am I missing something? Why doesn't Josh Reddick qualify for mention in this discussion. He has a 1+ WARP and basically has JD Drew's job now. Maybe he doesn't have a minor league contract? He started the season in AAA...

Comment rating:

0