PDA

View Full Version : Who controls how hard ball is hit, and how does quality of contact relate to BABIP



savafan
11-22-2011, 06:22 PM
Interesting studies. I've been reading through them. Not sure I fully understand everything put forth here, but I do find it very interesting.

Part I:

http://www.baseballprospectus.com/article.php?articleid=15532

Part II:

http://www.baseballprospectus.com/article.php?articleid=15562

757690
11-22-2011, 06:41 PM
Fascinating stuff. Looks like there will be a new breakthrough in understanding the game based on this new technology.

For those not up to date on this discussion, the first part of the first article does an outstanding job of summing up the history of the discussion up to now.

Thanks Savafan!

757690
11-23-2011, 02:21 AM
I've moved this discussion to this thread, since it makes more sense here. Btw, if you read these articles, you might change your mind.


That's not a logical fallacy. Given enough time, a pitcher's babip tends to regress to .300. While it may ultimately be shown to not be completely accurate, it's the most logical conclusion to hold that pitchers exert little control over whether a batted ball is a hit or not absent evidence to the contrary.
I'm not following this argument. If a pitcher could significantly effect his babip, then why wouldn't this effect become more apparent as his seasons piled up?
I'd love to trade often with the GM who holds the latter philosophy.

Just repeating myself here, but one argument is that it would become more apparent if more pitchers who are bad at it continued to pitch. If that happened, the range would be much bigger. But those who are bad at it, don't last long enough for their data to be included.

As for the "given enough time" argument, given enough time, pitcher's HR/9 will be around 1. Does that mean that pitchers can't control how many HR's they give up? No, it just means that in order to stay in the majors, you have to be good at controlling it.

jojo
11-23-2011, 07:24 AM
I've moved this discussion to this thread, since it makes more sense here. Btw, if you read these articles, you might change your mind.

I have read he articles as I've been pouring over these type of studies before saber metrics was cool.


Just repeating myself here, but one argument is that it would become more apparent if more pitchers who are bad at it continued to pitch. If that happened, the range would be much bigger. But those who are bad at it, don't last long enough for their data to be included.

But we know that there are plenty of bad pitchers in the majors at any given time. The Reds organisation has been a feeder system for just that during much of the lost decade. Clearly pitchers who aren't major league quality or good pitchers regressing to he point of being done will get their packing papers. That still does not explain why a pitcher's babip regresses back to .300 pretty rapidly after a significant deviation. There would seem to be two logical conclusions. First pitchers don't possess a significant enough ability to control babip for it to be reliably detectable statistically without huge sample sizes. Second, all major league quality pitchers must have a roughly similar ability to influence babip.

In either case (and I suppose they aren't mutually exclusive), babip wouldn't be a very useful tool to evaluate pitchers outright because it simply doesn't have alot of power (doesnt allow for selection).

Where babip could be useful in informing decisions would be in a situation like Harang. He had good peripherals over his last few seasons with the Reds but these seemed to outperform his actual performance (i.e you'd likely draw different conclusions about his future depending upon whether one looked at his FIP or is ERA). His peripherals also seemed to outperform the eye when watching him. Was his outcome poor luck or was it something legitimate that effected his effectiveness in a meaningful way that would suggest his true talent level had changed? To me, that is the lump that the Fast articles are poking at and where some real progress can and needs to be made.


As for the "given enough time" argument, given enough time, pitcher's HR/9 will be around 1. Does that mean that pitchers can't control how many HR's they give up? No, it just means that in order to stay in the majors, you have to be good at controlling it.

I don't believe this to be true at all. A pitcher's hr/9 does not necessarily regress to 1.

757690
11-23-2011, 07:38 AM
I have read he articles as I've been pouring over these type of studies before saber metrics was cool.



But we know that there are plenty of bad pitchers in the majors at any given time. The Reds organisation has been a feeder system for just that during much of the lost decade. Clearly pitchers who aren't major league quality or good pitchers regressing to he point of being done will get their packing papers. That still does not explain why a pitcher's babip regresses back to .300 pretty rapidly after a significant deviation. There would seem to be two logical conclusions. First pitchers don't possess a significant enough ability to control babip for it to be reliably detectable statistically without huge sample sizes. Second, all major league quality pitchers must have a roughly similar ability to influence babip.

In either case (and I suppose they aren't mutually exclusive), babip wouldn't be a very useful tool to evaluate pitchers outright because it simply doesn't have alot of power (doesnt allow for selection).

Where babip could be useful in informing decisions would be in a situation like Harang. He had good peripherals over his last few seasons with the Reds but these seemed to outperform his actual performance (i.e you'd likely draw different conclusions about his future depending upon whether one looked at his FIP or is ERA). His peripherals also seemed to outperform the eye when watching him. Was is outcome poor luck or was it something legitimate that effected his effectiveness in a meaningful wy tht would suggest his true talent level had changed? To me, that is the lump that the Fast articles are poking at and where some real progress can and needs to be made.



I don't believe this to be true at all. A pitcher's hr/9 does not necessarily regress to 1.

Last part first, the range of HR/9 for starting pitchers in the league for 10 or more years is the same as the range of BABIP for the same pitchers. I'm off to work so I don't have time to list them, but it doesn't take long, there aren't that many.

And I am not arguing about the value of using BABIP or DIPS or whatever in evaluating players. I use them all the time.

I was making more of a linguistic argument, which you summarized quote well in you explanations of the two logical conclusions. I just believe in the second option.

I think it's clear that pitchers can control their BABIP, and that only the ones that can do it well, last long enough. The best proof is this.

If Jason Vargas threw nothing but fastballs down the middle of the plate, his BABIP and his LD rate would go through the roof. He lowers them, that is controls them, but throwing other pitches, in other locations, at different speeds. And who are the pitchers who succeed? Not the one that only throw hard, but the ones that pitch well.

jojo
11-23-2011, 07:47 AM
If pitchers can control their babip, why do they allow it to settle at a level where a batter would be considered an all-star based upon batting average?

jojo
11-23-2011, 08:46 AM
Last part first, the range of HR/9 for starting pitchers in the league for 10 or more years is the same as the range of BABIP for the same pitchers. I'm off to work so I don't have time to list them, but it doesn't take long, there aren't that many.

Did you really mean to reference HR/FB%?

A pitcher's hr/9 is a function of his batted ball tendencies (how many fly balls he yields) and his HR/FB%.

A pitcher's batted ball tendencies is a repeatable skill that is largely a function of his repertoire and his stuff. Meanwhile is hr/fb% will tend to regress to a league average of 10% (i.e. he oesnt really exert much control over this).

Thus a pitcher's hr/9 wouldn't be expected to regress to 1 without a mitigating factor. For instance when looking at starting pitchers with the highest ground ball percentages between 2007 and 2011, precious few in the top 50 have a hr/9 of 1 or greater.....

757690
11-23-2011, 10:54 AM
If pitchers can control their babip, why do they allow it to settle at a level where a batter would be considered an all-star based upon batting average?

If pitchers can control HR's, why do they allow any?

Pitchers clearly control BABIP, the issue, as with all stats, is how much they control it.

PuffyPig
11-23-2011, 12:26 PM
Pitchers clearly control BABIP, the issue, as with all stats, is how much they control it.


It's not so clear to most of us.

I think the best you can say is that there may be a few pitchers who have been able to control BABIP to a marginal degree.

But when I see Pedro Martinez in 1999 going 23-4 with a 2.07 ERA with a 13.20 K/9, a 1.56 W/9 and a 0.38 HR/9, with undoubtedly the majors best stuff, yielding a BABIP that year of .323 (the highest of his career), it tells be that even the pitcher with the majors best stuff, enjoying his best year in the majors, had zero control over BABIP.

IslandRed
11-23-2011, 12:55 PM
The main takeaway I got from the linked articles is that pitchers do have measurable control over how hard the ball is hit when one of their pitches is hit. That won't always reflect in BABIP, but the more likely effect is in what type of hits are allowed. I wonder if anyone's broken down SLGBIP as thoroughly as BABIP?

defender
11-23-2011, 02:21 PM
Baseball is a game of skill. I think we have to start with the assumption that pitchers have control over the ball in play, and statistical analysis has failed to measure it. I donít think BAPiP works, because it does not consider what the pitcher is trying to do in each AB. Baseball is made up of many unique events, statistical analysis that treats every event (AB) as the same can result in incorrect conclusions.

dougdirt
11-23-2011, 02:22 PM
If pitchers can control HR's, why do they allow any?


Pitchers "control" the amount of HR's allowed based on the idea that pitchers, in a nuetral HR park, tend to give up a HR once every 10 fly balls and that pitchers can typically control the amount of fly balls they give up. That is where they "control" it.

757690
11-23-2011, 03:29 PM
It's not so clear to most of us.

I think the best you can say is that there may be a few pitchers who have been able to control BABIP to a marginal degree.

But when I see Pedro Martinez in 1999 going 23-4 with a 2.07 ERA with a 13.20 K/9, a 1.56 W/9 and a 0.38 HR/9, with undoubtedly the majors best stuff, yielding a BABIP that year of .323 (the highest of his career), it tells be that even the pitcher with the majors best stuff, enjoying his best year in the majors, had zero control over BABIP.

Nope, doesn't say that at all. It just says that he was very good at other things, and not very good at this one thing for that year.

In 1965, Bob Gibson lead the league in homers, and had almost double his career rate of HR/9, and yet he had a great year other wise, with one of his highest K totals ever. Does that mean that he can't control his HR's?

PuffyPig
11-23-2011, 05:30 PM
Nope, doesn't say that at all. It just says that he was very good at other things, and not very good at this one thing for that year.



So he was very good at every thing other than allowing hits on balls hit into play. Why would a pitcher with trhe best stuff in the league at the time, who had all world swing and miss abilty and control, be so bad at such a thing?

Randomness is the only logical answer.

In every other year, he had less swing and miss abilty, less abilty to allow HR's and less control, but was better at allowing less hits on balls hit into play.

I'm not buying it.

757690
11-23-2011, 06:03 PM
So he was very good at every thing other than allowing hits on balls hit into play. Why would a pitcher with trhe best stuff in the league at the time, who had all world swing and miss abilty and control, be so bad at such a thing?

Randomness is the only logical answer.

In every other year, he had less swing and miss abilty, less abilty to allow HR's and less control, but was better at allowing less hits on balls hit into play.

I'm not buying it.

Why in the world would the best pitcher in baseball in 1965 all of the sudden become the worst in the league at allowing HR's?

I'm not saying randomness wasn't involved with Pedro, I'm must saying that we can't conclude that he had no control over who got hits over the balls in play that year, just like we can't conclude that Gibson had no control over how many HR's he gave up.

You are completely missing my point.

My only point is that we can not conclude that pitchers have no control over their BABIP. In fact, all the evidence and logic shows that they do have control over it. The issue is how much control do they have over it.

Again, assume that Mike Leake throws nothing but fastballs over the heart of the plate. His BABIP and his LD rate would go through the roof, it would be the highest ever for a MLB pitcher.

However, he can lower that BABIP and LD rate, meaning he can control it, by throwing other pitches, in other locations, at various speeds, which is what he does, as do all other pitchers. They control their LD rate, and their BABIP by using different pitches, at different times, in different locations, at different speeds.

And the pitchers who are good at it, stick around, and their data gets analyzed. Those that aren't good at it, don't stick around, and we never use their data.

jojo
11-23-2011, 06:34 PM
If pitchers can control HR's, why do they allow

Pitchers can't control HRs.... That's why HR/FB% regresses to around 10%..

757690
11-23-2011, 07:19 PM
Pitchers can't control HRs.... That's why HR/FB% regresses to around 10%..

I have to get ready for the Holiday, so I'll end it here for now.

But you guys keep missing my point.

All I am asking is for people to stop saying that a pitcher can't control how many hits he gives up, or how many homers he gives up.

I understand all the stats and how they work, but they don't prove that pitchers can't control these things. All they show is that most MLB pitchers that last long enough, control them at a similar rate.

Airport waits, so I'll leave it at that.

Have a great Thanksgiving :)

jojo
11-23-2011, 07:26 PM
I have to get ready for the Holiday, so I'll end it here for now.

But you guys keep missing my point.

All I am asking is for people to stop saying that a pitcher can't control how many hits he gives up, or how many homers he gives up.

I understand all the stats and how they work, but they don't prove that pitchers can't control these things. All they show is that most MLB pitchers that last long enough, control them at a similar rate.

Airport waits, so I'll leave it at that.

Have a great Thanksgiving :)

You kep ignoring a central question that pokes at the heart of your position.... If pitchers can control their HR/FB, then why does it always regress back to 10%?

IslandRed
11-23-2011, 07:40 PM
It's not so clear to most of us.

I think the best you can say is that there may be a few pitchers who have been able to control BABIP to a marginal degree.

But when I see Pedro Martinez in 1999 going 23-4 with a 2.07 ERA with a 13.20 K/9, a 1.56 W/9 and a 0.38 HR/9, with undoubtedly the majors best stuff, yielding a BABIP that year of .323 (the highest of his career), it tells be that even the pitcher with the majors best stuff, enjoying his best year in the majors, had zero control over BABIP.

What it says, specifically, is that even if a pitcher has a better-than-typical ability to control BABIP, it can be masked by randomness when looking at a single season. For a typical starting pitcher, you can go 30-40 points either side of true ability for a season and have it explained totally by luck. That's just the standard deviation of the sample size.

But over larger time slices, the randomness influence goes down and the true ability starts to show through. According to baseball-reference.com, Martinez' career BABIP was .282 versus the MLB average over that span of .298.

PuffyPig
11-23-2011, 08:46 PM
But over larger time slices, the randomness influence goes down and the true ability starts to show through. According to baseball-reference.com, Martinez' career BABIP was .282 versus the MLB average over that span of .298.

So Martinez posted a 5% better BABIP than normal over his career.

Which shows that perhaps some pitchers control BABIP to a very minor degree, or it's simply random variance within an exceptable range.

I'm not sure either explantion really matters.

Pitchers have little or no control over BABIP.

wlf WV
11-23-2011, 09:29 PM
It's not so clear to most of us.

I think the best you can say is that there may be a few pitchers who have been able to control BABIP to a marginal degree.

But when I see Pedro Martinez in 1999 going 23-4 with a 2.07 ERA with a 13.20 K/9, a 1.56 W/9 and a 0.38 HR/9, with undoubtedly the majors best stuff, yielding a BABIP that year of .323 (the highest of his career), it tells be that even the pitcher with the majors best stuff, enjoying his best year in the majors, had zero control over BABIP.
It seems that a BIP against great stuff has greater hSOB,just not as many BIP?

RedsManRick
11-23-2011, 09:36 PM
The thing with BABIP is simply that there is a lower bound of what is possible to sustain due to the nature of the game. There are two primary factors at play.

1.) Number and positioning of the players. So long as batters are attempting to hit the ball hard/far, thet will continue to have a general spray chart which includes large sections of the field that are always/never covered by a fielder. That is to say, most balls hit in to play are almost always going to be a hit (e.g. A line drive in the gap) or always an out (pop-up to 1B). So when you consider the realm of possible influence, it's on the margins by necessity.

2) The nature of inducing weak contact. At its core, what causes weak contact is essentially what causes a swing and miss -- unexpected timing and/or location. So if you reach a certain point of skill at introducing weak contact, it will also manifest as strikeout ability. There's only so far you can go at being able to introduce weak contact before it stops being contact altogether.

When you combine these dynamics with the issue of selection at the bottom end removig laggards from the pool, you naturally wind up with a pretty tight distribution. Sure, pitchers have influence over the quality of balls hit against them. But the spread of that ability is very tight compared what we see in other skills (precisely because it is so crucial and thus a guy can't induce weak contact won't make it) and thet variance in the sustained level of ability is dwarfed by in-season variance.

This is actually a good example of a big point of misunderstanding with interpretation of stats in general - confusing an inability to prove the existence of a given skill level via a given sample with a denial that the performance may have been a result (at least partially) of having great skill. It's merely that a single season of, say .240, BABIP is much much more likely to be a result of luck than skill, so absent significant additional evidence, the safe bet is luck.

PuffyPig
11-23-2011, 11:16 PM
That is to say, most balls hit in to play are almost always going to be a hit (e.g. A line drive in the gap) or always an out (pop-up to 1B).





I'm glad we got that cleared up.

Balls hit into play are hits or outs. And here I thought they might turn into a choclate sundae....

Superdude
11-24-2011, 04:01 AM
The thing with BABIP is simply that there is a lower bound of what is possible to sustain due to the nature of the game. There are two primary factors at play.

1.) Number and positioning of the players. So long as batters are attempting to hit the ball hard/far, thet will continue to have a general spray chart which includes large sections of the field that are always/never covered by a fielder. That is to say, most balls hit in to play are almost always going to be a hit (e.g. A line drive in the gap) or always an out (pop-up to 1B). So when you consider the realm of possible influence, it's on the margins by necessity.

2) The nature of inducing weak contact. At its core, what causes weak contact is essentially what causes a swing and miss -- unexpected timing and/or location. So if you reach a certain point of skill at introducing weak contact, it will also manifest as strikeout ability. There's only so far you can go at being able to introduce weak contact before it stops being contact altogether.

When you combine these dynamics with the issue of selection at the bottom end removig laggards from the pool, you naturally wind up with a pretty tight distribution. Sure, pitchers have influence over the quality of balls hit against them. But the spread of that ability is very tight compared what we see in other skills (precisely because it is so crucial and thus a guy can't induce weak contact won't make it) and thet variance in the sustained level of ability is dwarfed by in-season variance.

This is actually a good example of a big point of misunderstanding with interpretation of stats in general - confusing an inability to prove the existence of a given skill level via a given sample with a denial that the performance may have been a result (at least partially) of having great skill. It's merely that a single season of, say .240, BABIP is much much more likely to be a result of luck than skill, so absent significant additional evidence, the safe bet is luck.

This pretty much sums up my thoughts on the matter. Using the word "luck" makes it sound like I should be able to step on a mound next season and throw 200 innings with a tidy .290BABIP.

dougdirt
11-24-2011, 09:49 AM
This pretty much sums up my thoughts on the matter. Using the word "luck" makes it sound like I should be able to step on a mound next season and throw 200 innings with a tidy .290BABIP.

It is worth noting that non-pitchers, who pitch, from 1993-part of 2011 (http://www.insidethebook.com/ee/index.php/site/comments/dips_nonpitchers_as_pitchers/) was..... .292. Not that you or I could go out and do that, but these professional players who are non pitchers aren't doing any worse than pitchers are once they allow the ball in play and keep it in the park.

PuffyPig
11-24-2011, 06:01 PM
It is worth noting that non-pitchers, who pitch, from 1993-part of 2011 (http://www.insidethebook.com/ee/index.php/site/comments/dips_nonpitchers_as_pitchers/) was..... .292. Not that you or I could go out and do that, but these professional players who are non pitchers aren't doing any worse than pitchers are once they allow the ball in play and keep it in the park.

I'd love to hear the explanation of the non-BABIP crowd on that one.

thatcoolguy_22
11-25-2011, 03:29 AM
Maybe its just that 3/10 balls you put in play will find a spot to be a base hit. 9 people trying to cover a large area of grass won't be perfect. Makes sense to me.

jojo
11-25-2011, 06:56 AM
Maybe its just that 3/10 balls you put in play will find a spot to be a base hit. 9 people trying to cover a large area of grass won't be perfect. Makes sense to me.

If the influence of everything around the the pitcher masks a true ability of the pitcher to control BABIP, then it's probably best cast as a luck metric.

It's interesting that it's relatively easy to detect an ability of the hitter to influence BABIP.

This whole issue has fascinated me from day 1 given the whole motivation for me to learn about saber metrics derived from the battle between pitcher and hitter.