Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Tuesday, August 29, 2017

Why Are Some New Statistics Embraced and Not Others? - The New York Times

Tango and company get some kudos for the stuff they are doing at MLB.com.

‘‘Bill James had a great line,’’ Tom Tango, the senior database architect of stats for MLB Advanced Media, told me. (If James is largely considered the godfather of sabermetrics, Tango could be its Michael Corleone.) ‘‘If you have a metric that never matches up with the eye test, it’s probably wrong. And if it never surprises you, it’s probably useless. But if four out of five times it tells you what you know, and one of out five it surprises you, you might have something.’’

Jim Furtado Posted: August 29, 2017 at 12:41 PM | 39 comment(s) Login to Bookmark
  Tags: sabermetrics, statistics

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Rally Posted: August 29, 2017 at 01:01 PM (#5522853)
(If James is largely considered the godfather of sabermetrics, Tango could be its Michael Corleone.)


Interesting idea. Who's the Fredo?
   2. Renegade (((JE))) Posted: August 29, 2017 at 01:08 PM (#5522862)
And who's Enzo the Baker?
   3. Pasta-diving Jeter (jmac66) Posted: August 29, 2017 at 01:08 PM (#5522864)
(If James is largely considered the godfather of sabermetrics, Tango could be its Michael Corleone.)


Interesting idea. Who's the Fredo?

MGL
   4. villageidiom Posted: August 29, 2017 at 01:51 PM (#5522893)
And unlike catchall measures of a player’s value, whether Wins Above Replacement (WAR) or Value Over Replacement Player (VORP), exit velocity is free of the stink of the actuarial tables deployed by team executives, or worse, fantasy-baseball players. It simply confirms, on a repeating, easily digestible basis, what you see on the field. And unlike on-base percentage (OBP), which rewards the boring walk, exit velocity coincides with something most fans come to the ballpark to see.

It occurs to me that these things are catching on for that very reason. It puts an actual number to what we already observe.

Value is subjective, which is why we have countless debates on what the MVP award is supposed to recognize. (Is there value in being good on a last place team? In being a good hitter in a great offense? In pitching well? Should leverage matter?) Any metric that purports to quantify value is doing so within a defined context where some activities have value and others do not, which in turn is either for subjective reasons or practicality. But we know objectively that hitting the ball hard will increase the likelihood of a hit and hitting it in the air at the proper angle will increase the likelihood of a hard-hit ball going for extra bases. The new Statcast stats just put a number on it, and those numbers generally make sense.

Something like barrels, crafted from these intuitive numbers, is in a sense the first sabermetric stat that has been constructed in front of the general baseball audience. You could argue stuff like UZR or DIPS was crafted mostly in the public domain, with somewhat instant peer review. But the audience then was the people on this site or in usenet a long time ago. It wasn't the general baseball audience. It had the air of an exclusive club if you were there for the start. Barrels, route efficiency... the entire MLB audience is learning them at the same time. There's no condescension, no differing camps.
   5. Tom Nawrocki Posted: August 29, 2017 at 02:02 PM (#5522898)
Interesting idea. Who's the Fredo?


Mike Gimbel. MGL is Sonny.
   6. TDF, FCL Posted: August 29, 2017 at 03:35 PM (#5522982)
You could argue stuff like UZR or DIPS was crafted mostly in the public domain, with somewhat instant peer review. But the audience then was the people on this site or in usenet a long time ago. It wasn't the general baseball audience. It had the air of an exclusive club if you were there for the start. Barrels, route efficiency... the entire MLB audience is learning them at the same time. There's no condescension, no differing camps.
And to get back to what you said earlier - "barrels" and "route efficiency" are things we see and things we think are good and things we teach little leaguers are good.

You can teach "throw the ball over the plate"; you can teach deception with pitches. But you can't teach "limit HRs" and no one actually sees any of those things (and MWE has argued DiPS only works with MLB players; that lower-level players do show different ability to limit BABIP). You can kind of see better fielding, but it's often been said that everyone makes great looking plays at the limits of their range; mediocre defense can often look "great".

So not only are we measuring things we observe, we're measuring things we can teach; it's putting a number on things the Marty Brennemans of the world spout off about during every game.
   7. PreservedFish Posted: August 29, 2017 at 03:41 PM (#5522988)
Isn't Billy Beane the Michael Corleone?
   8. Pops Freshenmeyer Posted: August 29, 2017 at 05:14 PM (#5523070)
Isn't Billy Beane the Michael Corleone?

Paul DePodesta is Moe Greene.
   9. Zach Posted: August 29, 2017 at 05:22 PM (#5523081)
(If James is largely considered the godfather of sabermetrics, Tango could be its Michael Corleone.)

Hey everyone! Michael is a terrible godfather, just like his brothers.

Each of the brothers lacks something that the father had. Sonny is too passionate. Fredo is too weak. Michael is too cold.

   10. WKRP in Cincinnatus Posted: August 29, 2017 at 05:39 PM (#5523096)
Isn't Billy Beane the Michael Corleone?

Paul DePodesta is Moe Greene.


So Brad Penny is a "Moe Greene Special?"
   11. Walt Davis Posted: August 29, 2017 at 06:32 PM (#5523133)
Yeah, Tango is more Tom Hagen. He never gets riled up about nothing.

I've got to believe that "limiting HRs" (i.e. keep the ball down in the zone and up out of the zone) is a lot easier to teach than "optimize the swing to maximize the chance of a barrel."

I don't think there's anything particularly magical about exit velocity and the fascination with it will likely fade in a few years. It will always be there in the same way that pitch velocity has been there for a long time.

Folks might want to keep that Bill James comment in mind. What do people really think this is measuring and why is it important? Joey Gallo is #2 in average exit velocity. That makes sense as he crushes the ball when he hits it ... but he's not one of MLB's best hitters. A kid named Yandy Diaz sits 6th ... and he's got an OPS+ of 68 and an ISO of 75. Alex Avila is 10th, tied with Matt Olson (?) and Yosmany Tomas. Kendry Morales has a higher avverage EV than Ryan Braun. Mike Trout doesn't appear to be in the top 50 (they don't number, I'm not counting) and has an average EV equal to VMart and Matt Kemp.

So the super-high K-brigade have high average exit velocities while actual good hitters are all over the place. Average EV seems to fail on both of James's criteria -- it tells us what we already know yet it doesn't conform to the "eye test" of being able to identify the best hitters.

Barrels/PA isn't doing much better. The top 8 or 9 look a little better (but still no Trout, no Harper, no Goldschmidt) but falls apart quicky after that -- Grichuk, Rasmus and Napoli are just outside the top 10. And Trout? Tied with Trey Mancini, Ryah Schimpf, Logan Morrison. Across the board, Salvador Perez's numbers are a close match to Trout's -- sometimes a bit better, sometimes a bit worse.

It looks a bit like clutch hitting. Clutch hits exist, it's not clear that clutch hitters do. Balls with high exit velocity, hit at certain launch angles (but not to certain parts of the park) have very good outcomes -- just like clutch hits -- but that doesn't guarantee anybody can do this on a particularly regular basis. Clearly the high-K guys are all about generating high exit velocity but, except for those guys, it doesn't appear to be a particularly important component of successful hitting. And if we want to capture those guys who totally wail on it when they do hit it, we have the hR list, the HR/AB list, the HR/FB list, could have the HR/contact list,

I can see it may have some projective value -- I haven't looked but I assume Yandy Diaz pounds the ball into the ground and maybe if he can be taught to elevate the ball a bit that EV might translate into better production. This seems to have been the case for Ryan Zimmerman who has been near the top in average EV for the last three years but maybe not as I see hie B/F ratio has returned towards the norms of the last few years and his 2nd half OPS is almost 200 points below the first half (but still way better than last year).

Like not every guy who can throw 97 can achieve the right mix of control and movement, doesn't necessarily have anything but velocity, may in the end stink as a pitcher .... maybe batter EV is closer to that. A high EV guy has the potential to be a Stanton, maybe even a Thome ... and is almost certainly a better hitter than a guy who generates a much lower EV while still hitting the ball in the air a lot (Billy Hamilton say). I'm far from convinced it tells us much that's useful other than quantifying "boy, he hit that a ton!"

Here's a good one -- Starlin Castro, Joe Mauer, Daniel Murphy and Jeff Samardzija have the same barrels/PA rate this year (3.4%).

DM 496 PA, 20 HR, 133 OPS+
SC 354 PA, 12 HR, 112 OPS+
JM 469 PA, 6 HR, 110 OPS+
JS 59 PA, 1 HR, -32 OPS+

Mauer leads that little group in avg EV, avg EV_FB/LD and % 95+; Castro leads in max EV and max distance; Murphy leads in avg FB distance.
   12. Rennie's Tenet Posted: August 29, 2017 at 06:44 PM (#5523140)
I think that WAR has a couple of time bombs in it for general consumption. First, offensive WAR plus defensive WAR has to equal WAR. Second is bWAR and fWAR differing quite a bit in some instances.
   13. Mike Emeigh Posted: August 29, 2017 at 07:04 PM (#5523147)
MWE has argued DiPS only works with MLB players; that lower-level players do show different ability to limit BABIP


Well...

What I have argued (based on data from Clay Davenport and others) is that minor-league BABIP tends to be a predictor of who moves up and who doesn't, so that over time the differences between pitchers becomes small enough so that DIPS works. It's not that pitchers don't control H/BIP; it's that those who can't do it as well don't make it to the majors.

Bill James used weight for offensive linemen in football as a point of comparison. The differences in weight between the lightest and heaviest offensive linemen are fairly small, so that if you correlated weight to performance you'd conclude that weight doesn't have any measurable effect on performance. Well, it's not that weight doesn't matter (as you will find if you try to put a 220-pounder on an NFL O-line); it's that there's a defined minimum standard that restricts your choice.

It's not quite the same for strikeouts, or walk prevention, or home run prevention; there is much more variance for those stats among major league pitchers than there is for H/BIP, so while there is a minimum threshold there as well, the tolerance for differing levels of performance is much larger so you can actually identify measurable differences in performance.

-- MWE
   14. cardsfanboy Posted: August 29, 2017 at 07:11 PM (#5523152)
I think that WAR has a couple of time bombs in it for general consumption. First, offensive WAR plus defensive WAR has to equal WAR. Second is bWAR and fWAR differing quite a bit in some instances.


Why? I think that the way oWar and dWar is correct. It's annoying at times, but good. If someone wants to argue which player provided the best defensive value, say Keith Hernandez vs Mike Piazza, it's a good tool to use, and illustrates how important the concept of positional differences is.
   15. PreservedFish Posted: August 29, 2017 at 08:57 PM (#5523205)
Mike Emeigh, with respect I don't think I buy that argument. Baseball players vary wildly in every trackable human ability, from running speed to strength to body fat. There are obese players, and players that are dumb as rocks, and we've even seen an outfielder that throws like a girl. If BABIP is a real ability then there must be pitchers out there that have immense ability to limit BBs, or to rack up Ks, to the point where it would allow them to surmount their subpar BABIP ability. I don't understand how BABIP would be the one thing that gets weeded out at the minor league level to such an extent that it becomes the one ability that all players share in common.

Personally, I think BABIP is a red herring. Removing HRs from a pitcher's "balls in play" line makes sense from one very narrow perspective (trying to measure the effect of defense) and is hilariously stupid from the larger, more vital perspective of "can this guy limit hard contact?" BABIP is a fundamentally weird and unnatural statistic and we should expect it to have weird results.
   16. Everybody Loves Tyrus Raymond Posted: August 30, 2017 at 02:34 AM (#5523296)
This is not a criticism, but I find it oddly fascinating this guy continues to go by "Tom Tango."
   17. Rally Posted: August 30, 2017 at 09:17 AM (#5523364)
Personally, I think BABIP is a red herring. Removing HRs from a pitcher's "balls in play" line makes sense from one very narrow perspective (trying to measure the effect of defense) and is hilariously stupid from the larger, more vital perspective of "can this guy limit hard contact?" BABIP is a fundamentally weird and unnatural statistic and we should expect it to have weird results.


Good point. If you want to evaluate defense, then use BABIP. If you want to evaluate the pitcher's ability to prevent hard contact, keep the homers in. Batting Average on CONtact - BACON! I wish we'd see it more often. Everybody loves bacon!
   18. Rally Posted: August 30, 2017 at 09:29 AM (#5523370)
I really doubt this weeding out process happens. There is a group of pitchers for whom this weeding out process is never applied. I'm talking about position players who take the mound in blowouts. BBref has a list.

Here's what I did: Take all players on that list whose careers started after 1993 (BABIP has been roughly .300 since then, it was lower before that year). Remove Rick Ankiel - he was a real pitcher before he was an outfielder. That leaves me with:

180 innings
233 hits
43 homers
105 walks
66 strikeouts
18 HBP
920 BFP

I get 688 at bats here - that is probably a bit high since the table does not have sacrifices. There should be a few sac flies here but I doubt there is a single sac bunt in these situations.

190 non homer hits. That's a BABIP of .276. Maybe it's .285 if I had sac flies, maybe .290. But it sure is not .350 or anything like that.

These "pitchers" are horrible at preventing homers and walks. Horrible at striking people out. But downright decent at preventing hits in play.
   19. PreservedFish Posted: August 30, 2017 at 09:41 AM (#5523379)
Thank you Rally. I was aware of that fact but I couldn't find a reference on the internet so I didn't cite it. The fact that they allow an egregious amount of homeruns also disproves any claim that the hitters are trying less hard or altering their behavior because of the unpredictable pitching or blowout situation.
   20. GuyM Posted: August 30, 2017 at 02:46 PM (#5523568)
These "pitchers" are horrible at preventing homers and walks. Horrible at striking people out. But downright decent at preventing hits in play.

Well, maybe. But it certainly seems possible that these hitters, on the winning side of blowouts, are swinging at a lot of pitches they might take under normal conditions. The BB% is 11%, not that much higher than against real pitchers (around 8% in these years), while the K% is far lower (7%, vs. about 18% with real pitchers). I would guess that if these hitters were really trying, they could be much more selective, resulting in both more walks and much better contact on balls put in play. And contra #19, a high home rate would not be inconsistent with the idea that these hitters are being very aggressive at the plate.
   21. Sunday silence Posted: August 30, 2017 at 03:02 PM (#5523579)
what exactly are you proving in no 18? If you turn flyballs into HRs, I suppose BABIP will decline. I think the normal HR rate is about 2% (maybe up recently) and these pitchers are throwing them at a rate of about 7%.

If instead we take 5% of those HRs and turn them into screaming line drives, maybe what 70% of them drop in? So add 3.5% to your .276 babip and its now at .310.

The point is, I just dont know what you prove by having poor pitchers out there throwing meatballs. This probably changes BABIP and not in a way that would have survived the natural selection of the minor leagues. So your point about natural selection is probably not being made either.
   22. Ron J Posted: August 30, 2017 at 03:04 PM (#5523580)
#12 Finally got around to doing a basic sanity check of WAR. It pretty much confirms what we've been saying all along. The standard error of team WAR to team wins is roughly 4.65. Meaning that the standard error of WAR for a full-time player has to be in the range of 1. "Only" explains ~81% of the variation in team wins. (To be clear that's pretty good)

Not that this should be a surprise. The standard error of pythags is around 3.5 wins per 162 games and WAR is working with counter stats ( and the conversion of counter stats to batting runs has a standard error in the range of 14 runs. Should be a similar range of standard error on the defensive side)

And all I'm arguing against is the presumption of second decimal place accuracy. Simply confirms what (among others -- but very relevant names in any WAR discussion) Sean Smith or Sean Forman have been saying from the get go. It's a good overall measurement of value. Best place to start a discussion. It can't be the last word.
   23. PreservedFish Posted: August 30, 2017 at 03:04 PM (#5523581)
I dunno Guy. I'm not convinced.

I find #21 even less convincing.
   24. Rally Posted: August 30, 2017 at 03:50 PM (#5523602)
If instead we take 5% of those HRs and turn them into screaming line drives, maybe what 70% of them drop in? So add 3.5% to your .276 babip and its now at .310.


43 home runs. 5% of that is 2.15 homers. So 690.15 BIP. 70% of those drop in - 1.5 more hits. That changes BABIP to .2775, not .310.

From 1993-2017 overall BABIP is .301

BACON is .328. Add those 43 homers back in and BACON for the fake pitchers is .319

Probably what we should be looking at is wOBA on contact - WOBACON? Since homers are more valuable than other hits that might show that non-pitchers are giving up worse results on contact than real pitchers.

All this demonstrates to me why DIPS looks reasonable. The pitcher tries to throw a ball you can't hit. You try to square it up. The results fall into this spectrum:

1. No contact
2. weak contact
3. medium contact
4. hard contact
5. Extreme hard contact - over the wall.

Good pitchers have a lot more 1 and less 5, bad pitchers the opposite. When you take both 1 and 5 out of the equation, no wonder all pitchers look a lot more similar to each other.
   25. Rally Posted: August 30, 2017 at 03:58 PM (#5523606)
If instead we take 5% of those HRs and turn them into screaming line drives, maybe what 70% of them drop in? So add 3.5% to your .276 babip and its now at .310.


Perhaps I interpreted this wrong. Normal pitchers in the same number of AB would give up about 24 homers. So these guys are hit for an additional 19. Turn them into line drives and that's 19 more AB, 13.3 more hits, a .287 BABIP.

If I turn all 43 homers into line drives in play, that would be 30 more hits, and a BABIP of .301. That's unrealistic because normal pitchers do indeed give up some homeruns. I see no way to get to .310.
   26. PreservedFish Posted: August 30, 2017 at 04:10 PM (#5523616)
It takes a lot of creative thinking to turn HRs into line drives. HRs that miss are usually fly balls, which usually get caught. If we're saying that hitters in blowout situations have a very different approach that leads them to totally eschew a line drive swing, well, we're getting pretty speculative now.
   27. GuyM Posted: August 30, 2017 at 04:27 PM (#5523631)
I dunno Guy. I'm not convinced.

If someone really wanted to, they could figure this out. Taking the data from the past decade, you could see if hitters do swing at more out-of-zone pitches in these situations. I'd be surprised if they don't, but admittedly I'm speculating. But I don't think the burden of proof here falls on those who accept Clay Davenport's study. Rally presents some interesting data in #18, but the sample is small and the potential for bias is big (because these are all late-inning blowout situations). It's more than an anecdote, but far short of a study. Maybe Davenport missed something -- it's been years since I read his article -- but someone needs to show that if you want to reject the idea that MLB selects for pitchers with an ability to induce weak contact.
   28. PreservedFish Posted: August 30, 2017 at 04:50 PM (#5523646)
Taking the data from the past decade, you could see if hitters do swing at more out-of-zone pitches in these situations. I'd be surprised if they don't, but admittedly I'm speculating.


Me too, but we'd also be speculating on the effect of that, which is a double speculation situation.

Maybe Davenport missed something -- it's been years since I read his article -- but someone needs to show that if you want to reject the idea that MLB selects for pitchers with an ability to induce weak contact.


I need a link. Can't debate something I'm not familiar with.

However, I'd be curious to see if you can mount an argument against my reasoning in #15. Why on earth would BABIP be the one single skill that all major league pitchers share equally? (I don't find the offensive lineman thing remotely compelling)
   29. cardsfanboy Posted: August 30, 2017 at 05:04 PM (#5523652)
Why on earth would BABIP be the one single skill that all major league pitchers share equally?


They don't share it equally, they share it within a tight grouping, but there is evidence that veteran pitchers have a slightly better babip, evidence that knuckleball pitchers have a slightly better babip, and evidence that some pitchers are better able to get infield popups, which also affects babip, and there is evidence that there is a difference in babip for flyball pitchers vs groundball pitchers

Ultimately though there is only so many square feet available in play that can't be covered by defenders, and the pitcher doesn't really control that much the location where a hitter hits the ball, other than a general concept.
   30. Infinite Yost (Voxter) Posted: August 30, 2017 at 05:16 PM (#5523658)
Why on earth would BABIP be the one single skill that all major league pitchers share equally?


I'm not sure that's what's going on. I think the truth is that BABIP is extremely hard to control, which is why so few pitchers have a trackable skill therein. Now, it may be that there's a "control BABIP" floor above which you don't see the bigs, or it might be that controlling BABIP isn't worth it for most pitchers, or it might be that major league hitters, even the bad ones, are so good that you can't control their BABIP beyond a very narrow band. But I gravely doubt it's a skill that all people share equally.
   31. GuyM Posted: August 30, 2017 at 05:27 PM (#5523664)
However, I'd be curious to see if you can mount an argument against my reasoning in #15.

I don't disagree with it. Hit prevention ability clearly does vary among MLB pitchers. But the amount of variance is presumably reduced by the screening process Davenport described. (Of course, that should also be true for other pitching skills, at least somewhat.)

   32. Ziggy: The Platonic Form of Russell Branyan Posted: August 30, 2017 at 06:00 PM (#5523680)
More speculation time:

There's (effectively) a ceiling on the ability for a human to limit major league BABIP. Now, AT THE MAJOR LEAGUE LEVEL, the ability to limit BABIP may not have much to do with the ability to limit HR, BB, or increase K's. But that's because this is an extremely non-random group, all of the members of which are pretty much hard up against the limit on limiting BABIP. (Which, if I understood correctly, is what Davenport is saying.)
It wouldn't be any surprise (someone must have the data) if, especially at lower minor league levels, there is some correlation between BABIP and HR or BB rates.* If there is, then the generally bad pitchers get weeded out, everyone that we're left with is maximally good (with knuckleball exceptions & etc.) at limiting BABIP, and so the link to other aspects of pitching well disappears when we're talking about major league players.

* The position players pitching seems to tell against this, but I really don't know how much to read into those data. Those are junk-time at bats against a team that has all but surrendered. What difference that might make, I don't know, but it might be different than minor league players who are trying to stay in the league, with pitchers who are pitching in front of minor league defenses.
   33. You're a clown, RMc! I'm tired of it! Posted: August 30, 2017 at 06:49 PM (#5523706)
BILL JAMES: I never wanted this for you. I work my whole life - I don't apologize - to take care of my family, and I refused to be a fool, dancing on the string held by all those bigshots. I don't apologize - that's my life - but I thought that, that when it was your time, that you would be the one to hold the string. Senator SABR; Governor SABR. Well, it wasn't enough time, Tango. It wasn't enough time.

TANGO: We'll get there, Bill. We'll get there.
   34. What did Billy Ripken have against ElRoy Face? Posted: August 30, 2017 at 06:52 PM (#5523707)
Can't debate something I'm not familiar with.

You're new to this country, aren't you?
   35. PreservedFish Posted: August 30, 2017 at 09:38 PM (#5523781)
There's (effectively) a ceiling on the ability for a human to limit major league BABIP. Now, AT THE MAJOR LEAGUE LEVEL, the ability to limit BABIP may not have much to do with the ability to limit HR, BB, or increase K's. But that's because this is an extremely non-random group, all of the members of which are pretty much hard up against the limit on limiting BABIP. (Which, if I understood correctly, is what Davenport is saying.)


I just read the Davenport article. That's not what he says. Davenport simply says that BABIP is a regular skill just like K, BB, HR.

That doesn't clear up the mystery. There's no earthly reason why there can't be bad BABIPers in the majors. There are slow runners and weak hitters and flaws of every sort present. There must be bad BABIPers in the majors! It makes no sense otherwise. And so we have to assume that there are. It's just that there is a shockingly thin spread in BABIP results from the best to the worst - best explained by Rally's #24.
   36. Rally Posted: August 31, 2017 at 08:36 AM (#5523886)
but someone needs to show that if you want to reject the idea that MLB selects for pitchers with an ability to induce weak contact.


MLB selects for pitcher who can prevent runs. That is the extent of the selection process. Whether they do so by missing bats, throwing strikes, or preventing hard contact. That is the extent of the selection process.

I just don't see the idea that all MLB pitchers are in a tight band of BABIP prevention but minor league pitchers are above that as even remotely possible. The reason is you just don't have the time to make such a selection. The better pitching prospects just don't spend very much time at each minor league stop before they are promoted.

Here's an article showing the stabilization point for various stats, the amount of playing time where you can regress 50% to the mean. For pitcher strikeout rate, you just need 70 batters faced. That's pretty easy to meet and have an idea of how good a guy is at missing bats. For pitcher BABIP, you need 2000 balls in play to make a judgment with the same degree of confidence. That's probably 4 years of full season minor league starting pitchers. They don't have that much time. If they are rejecting pitchers who are good at other things but have a high BABIP in the minors, then they are making bad decisions - they are almost certainly weeding out more decent pitchers with bad luck or bad defensive support than weeding out guys who truly can't avoid getting hit.
   37. Ron J Posted: August 31, 2017 at 09:42 AM (#5523929)
#36 Goes farther than what you're saying in that even if it's true that there's a base minimum for a BABIP skill (stipulating for the moment that such a skill exists) it's certain that some pitchers who do in fact possess the skill will wash out due to bad luck. For the precise reasons you lay out.

To my knowledge nobody's worked out what the normal variation is over the course of a career. A quick look at Greg Maddux gives a standard deviation of .019.

IOW ~ +/- 16 hits a year most of the time (for a healthy starter who puts a lot of balls in play to be sure -- but I picked him for a reason) with a non-trivial chance of a bigger variation. Which is in essence why the article you cite suggests it takes so long for BABIP to stabilize.

EDIT: article author mistake
   38. GuyM Posted: August 31, 2017 at 01:12 PM (#5524127)
For those interested, here is the original Clay Davenport article. (ungated)
   39. Rally Posted: August 31, 2017 at 01:26 PM (#5524138)
Here are the 2017 BABIP by ball in play type (by bbref):

GB .246
LD .614
FB .089

That varies by year but probably pretty similar for past years to the extent we have consistency of data collection and classification. All flyballs have a BA of .217, .089 is for the ones that stay in the ballpark.

What's probably happening is that the non-pitchers are really weak at keeping the ball down. So they get hammered with a homerun rate that is nearly double what true pitchers give up. But BABIP calculation takes those events out of the formula no matter how many are given up. What we're left with is probably more flyballs, and fewer groundballs, than normal pitchers. Since flyballs in the park turn into outs more often, we are left with the counter-intuitive result of bad pitchers having more success on balls in play.

Imagine we take a similar approach to tee shots golf. Remove the 300 yard drives that land in the middle of the fairway. Rory Mcilroy might hit these most of the time. I won't hit one without a lot of luck and wind. Remove the misses, popups, balls that find water, go into the adjacent fairway or woods, etc. He might make a mistake now and then, but those are probably 50-75% of my tee shots. Now look at what's left in the middle. Now my average drive isn't that much worse than his, or at least closer before we remove the extremes.

I think Mcilroy still would be noticeably better, the "what's left" for him is probably a bunch of 275-299 yarders, while my "what's left" will be more 225-250 yarders. Baseball goes one level deeper in that the "near success" for a hitter, the flyout that doesn't make it out, is less likely to be a hit than the "near failure", the weak grounder that might sneak through or something a runner can beat out before the infielder gets to it.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Support BBTF

donate

Thanks to
Harveys Wallbangers
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOTP 18 September 2017: Ex-Baseball Star Darryl Strawberry Criticizes Jemele Hill, Praises POTUS: Trump is ‘A Great Man’
(675 - 4:41am, Sep 21)
Last: David Nieporent (now, with children)

Newsblog‘Friends,’ the Sitcom That’s Still a Hit in Major League Baseball
(326 - 3:24am, Sep 21)
Last: Ben Broussard Ramjet

NewsblogPat Neshek rips Zack Greinke for not signing autograph | SI.com
(17 - 2:15am, Sep 21)
Last: Baldrick

NewsblogThis is baseball’s home run breaking point | New York Post
(25 - 2:12am, Sep 21)
Last: Baldrick

NewsblogJosh Donaldson Wants You To Know: This Wasn’t A Catch
(48 - 1:49am, Sep 21)
Last: villageidiom

NewsblogOMNICHATTER'S not the man they think he is at home, oh, no, no, no, for September 20, 2017
(88 - 1:35am, Sep 21)
Last: LA Podcasting Hombre of Anaheim

NewsblogDayton Moore on why Royals did not sell | MLB.com
(13 - 12:59am, Sep 21)
Last: QLE

NewsblogYoenis Cespedes reveals his new offseason training plan | New York Post
(6 - 11:03pm, Sep 20)
Last: What did Billy Ripken have against ElRoy Face?

Gonfalon Cubs15 To Go
(67 - 11:01pm, Sep 20)
Last: Moses Taylor, aka Hambone Fakenameington

NewsblogMets' Jose Reyes: Continues late-season offensive bender
(17 - 10:38pm, Sep 20)
Last: Russlan thinks deGrom is da bomb

NewsblogA’s preferred stadium site criticized for causing gentrification, killing waterfowl | Field of Schemes
(59 - 9:39pm, Sep 20)
Last: Mellow Mouse, Benevolent Space Tyrant

NewsblogHow Tracking Technology Helped Baseball’s Best Fielding Outfielder - WSJ
(3 - 7:24pm, Sep 20)
Last: Sunday silence

NewsblogIf you’re the Red Sox, do you want to play the Indians or Astros in the divisional round of the playoffs?
(63 - 6:54pm, Sep 20)
Last: jacksone (AKA It's OK...)

NewsblogMust C: Five straight for Olson | MLB.com (video)
(12 - 5:08pm, Sep 20)
Last: Walt Davis

NewsblogOT-NBA off season thread
(2408 - 5:07pm, Sep 20)
Last: PASTE, Now with Extra Pitch and Extra Stamina

Page rendered in 0.4256 seconds
47 querie(s) executed