Strikeouts, Walks and an Apparent Contradiction

[Linus]

Strikeouts, particularly the assessment of pitchers ability to strike out batters has garnered much attention in the last couple years.  So much so that some baseball pundits consider it the best indicator of a pitcher's effectiveness or potential.  While I think the strikeout is currently being overemphasized, I definitely agree with the premise that a ball not put in play for an out is far superior to one that is put in play, even if it should result in an out.

 

So, to my first apparent contradiction: if striking people out is so important as a pitcher, why shouldn't hitters be equally as concerned with preventing strikeouts?  Many observers say things like, strikeouts don't matter because an out is the same whether it is a strikeout or a flyball, etc.  Batters are striking out in record numbers with no apparent reservation or change in approach to prevent it.  This flies directly in the face of the modern pitching credo.

 

The same contradiction appears to be at play in reverse with walks.  We have people say they don't care if a pitcher walks too many batters if they have high strikeout rates, even to say that a walk is preferable to a hit because, after all, a walk is just one base and a hit can be more.

 

Contradiction number 2: why is the modern hitting philosophy to work counts with the hope of getting a walk, while driving up pitch counts.  If obtaining walks is such a crucial offensive strategy shouldn't there be equal emphasis on the pitching side dedicated to preventing walks?

 

I will await the wisdom of the Twins Daily posting community.......

[Mike Frasier Law]

I'm not sure I accept the premise. You're absolutely right that strikeouts have the same absolute value for a pitcher and for a batter. Similarly, walks have less value than any hit.

 

Here's where I disagree with your premise. Absent excellent ground-ball rates, allowing contact is bad for pitchers. That's because contact sometimes turns into hits, and hits score more runs than walks. So if allowing more walks is a consequence of also striking out a lot and having a low contact rate, the tradeoff may be worth it.

 

For batters, I don't think the goal is to work the count to take a walk. I think the goal is to force the pitcher to throw a pitch that the batter can drive. With strike zones shrinking as they have been, it means more and more walks for patient batters, and empowers them to really be patient and "wait for their pitch"

[drivlikejehu]

There is no contradiction. The difference lies in the ability of hitters and pitchers to influence the results of balls in play.

 

A hitter can choose to 'swing for the fences.' If the decrease in balls in play is more than offset because he is hitting the ball harder, then the strikeouts are a worthwhile trade-off. They aren't 'good,' but necessary to get the benefit.

 

A pitcher has no way of making this trade-off. Allowing more balls in play does not decrease the quality of contact. In fact, the evidence tends to show the opposite.

 

So far as supposed contradiction #2, again there is no contradiction - no one believes walks are a good thing for pitchers to allow. Command is universally considered to be crucial for pitchers.

[ashbury]

There is no contradiction. The difference lies in the ability of hitters and pitchers to influence the results of balls in play.

 

A hitter can choose to 'swing for the fences.' If the decrease in balls in play is more than offset because he is hitting the ball harder, then the strikeouts are a worthwhile trade-off. They aren't 'good,' but necessary to get the benefit.

 

A pitcher has no way of making this trade-off. Allowing more balls in play does not decrease the quality of contact. In fact, the evidence tends to show the opposite.

 

So far as supposed contradiction #2, again there is no contradiction - no one believes walks are a good thing for pitchers to allow. Command is universally considered to be crucial for pitchers.

I suspect that this is in line with the research that shows that few pitchers can control the Batting Average on Balls In Play (in the long run most come close to the league average around .300), whereas some batters seem to be able to routinely achieve higher BABIP than other batters.  Even though every run scored is a run given up, the jobs of batting and pitching are not precisely mirrors of each other.

[old nurse]

I suspect that this is in line with the research that shows that few pitchers can control the Batting Average on Balls In Play (in the long run most come close to the league average around .300), whereas some batters seem to be able to routinely achieve higher BABIP than other batters.  Even though every run scored is a run given up, the jobs of batting and pitching are not precisely mirrors of each other.

You are incorrect on BABIP for pitchers in the long run evens out.  While BABIP is variable for a pitcher season to season is variable, in larger sample sizes is stabilizes. Johan Santana was around .279 in the first decade of this century Over the same decade, Sidney Ponson was .310 There were many between them

[ashbury]

You are incorrect on BABIP for pitchers in the long run evens out.  While BABIP is variable for a pitcher season to season is variable, in larger sample sizes is stabilizes. Johan Santana was around .279 in the first decade of this century Over the same decade, Sidney Ponson was .310 There were many between them

OK.  But the range you can find among pitchers is quite a bit smaller than for batters, isn't it?  Agreed that a bit of variation from league average means that the principle is missing a detail or two that eventually may become better understood.  But for now...

[drivlikejehu]

You are incorrect on BABIP for pitchers in the long run evens out.  While BABIP is variable for a pitcher season to season is variable, in larger sample sizes is stabilizes. Johan Santana was around .279 in the first decade of this century Over the same decade, Sidney Ponson was .310 There were many between them

 

Right, the point here is that, whether it's someone like Santana or Ponson, more balls in play = worse results.

[old nurse]

Right, the point here is that, whether it's someone like Santana or Ponson, more balls in play = worse results.

My point was pitchers have some control in BABIP.

http://www.fangraphs.com/community/proejcting-babip-using-batted-ball-data/

http://www.baseballprospectus.com/article.php?articleid=2617

In terms of more balls in play=worse results, the correlation between runs scored and strikeouts is .50. There are other factors involved beyond the stikeout. For the same k rate in the group o pitchers with a k rate of .1-.2, the graph becomes fairly blob like. Also notice how few pitchers there are at the top of the k/pa line.

To answer the OP's question on why hitters are not afraid to strike out. An out is an out. Someone must have done a litle research and found the odds of getting a hit were higher if you swung your normal swing versus shortning up. It could also be hitting for most might be hard enough for one swing that trying to devlop a second might mess with the first. Pure speculation on my part.

[old nurse]

OK.  But the range you can find among pitchers is quite a bit smaller than for batters, isn't it?  Agreed that a bit of variation from league average means that the principle is missing a detail or two that eventually may become better understood.  But for now...

I gave you an incomplete range. Johan wasn't the lowest, Poson wasn't quit the highest. I also gave you from a list looking at pitchers who threw 1000 inings. If you look at this year at pitcher who threw at least 30 innings, Kevin Slowey was up at .387 Tony Sipp was low at .192 In looking sttas, if your BABIP is high as a pitcher chances are it is not poor luck. Remember Bruce Chen of .370 BABIP? Royals cut him after the Twins used him for batting practice.

[diehardtwinsfan]

 

I suspect that this is in line with the research that shows that few pitchers can control the Batting Average on Balls In Play (in the long run most come close to the league average around .300), whereas some batters seem to be able to routinely achieve higher BABIP than other batters.  Even though every run scored is a run given up, the jobs of batting and pitching are not precisely mirrors of each other.

 

Since we are talking contradictions, I'm going to simply state that I think this is a big one. If pitchers couldn't control babip, then I could go out there and throw and end up with an ERA roughly around 9. I think people are confusing a bell curve with luck. Hitters have control over their babip, but when it averages out, it averages out to approximately .300. Throw a bad pitcher out there though, and that babip will skyrocket.

[Mr. Brooks]

 

 

Since we are talking contradictions, I'm going to simply state that I think this is a big one. If pitchers couldn't control babip, then I could go out there and throw and end up with an ERA roughly around 9. I think people are confusing a bell curve with luck. Hitters have control over their babip, but when it averages out, it averages out to approximately .300. Throw a bad pitcher out there though, and that babip will skyrocket.

 

Not quite right.

You are forgetting the fact that HR's don't count in babip. If you tried pitching to MLB hitters, they would just take you over the fence most of the time.

On the times you actually were able to induce a ball in play, you would probably be somewhere around the .300 mark, over a large enough sample size. 

[Craig Arko]

Since we are talking contradictions, I'm going to simply state that I think this is a big one. If pitchers couldn't control babip, then I could go out there and throw and end up with an ERA roughly around 9. I think people are confusing a bell curve with luck. Hitters have control over their babip, but when it averages out, it averages out to approximately .300. Throw a bad pitcher out there though, and that babip will skyrocket.

It was never clear to me that a normal distribution (bell curve) was the correct model to use anyway. There are a lot of probability distributions found in nature.

 

http://en.m.wikipedia.org/wiki/List_of_probability_distributions

[ashbury]

Not quite right.

You are forgetting the fact that HR's don't count in babip. If you tried pitching to MLB hitters, they would just take you over the fence most of the time.

On the times you actually were able to induce a ball in play, you would probably be somewhere around the .300 mark, over a large enough sample size. 

They should hire Diehard (or me, who is even worse) to throw to the Home Run Derby contestants, then, because even with meatball pitches from their chosen thrower most of them don't hit a majority out.

 

As for bell curves mentioned elsewhere, every player in pro ball is pretty far to one side of that curve already; only the tail of the curve matters.  Studies that derive some sort of conclusion, such as control over BABIP, are from this part of the curve and may not have the slightest bearing on those of us in the middle or below.