Action Pitches


#1

I’ve come across another “old friend” in Action Pitches. This one’s a little weird, and because of that I never really gave it much of a look. But before I deleted it I thought I’d give it one last chance. When I did that and looked at it a bit harder and made some adjustments, I found something that might just have some interest after all.

The thought behind it is pretty simple and comes from an article I read that was written by Joe Posnanski. An action pitch is a pitch where something happened: Hit, walk, error, hit-by-pitch, sacrifice, strikeout, groundout, flyout, lineout .etc… It’s goes a bit further than just BIP.

When I ran it, I sat and looked at it for a while, and it was strange that what I saw seemed to be that the pitchers with the highest percentage of APs seemed to be the lesser skilled pitchers. I suppose it has something to do with most of the items being generally thought of as negative.

Anyhoo, I thought I’d expose ya’ll to it because you never know what will trigger a thought in someone else’s mind. :wink:

BTW, the reason I went with runs scored the way I did, was because I don’t track pitchers the way I track hitters. I know how runners go on that scored, but not how all runners got on. It’s a failure in my data that I’ll prolly fix one day, but not today! :wink:


#2

How is the Action Pitch % calculated?


#3

Hmmm…let’s look at that for a moment.
You say that an “Action pitch” is one where something happens. You also say that the pitchers who have the highest percentage of such pitches seem to be the less skilled. What I’m reading into this is that these are pitchers who throw an awful lot of pitches in the course of a game—have you factored in the number of foul balls hit by the batters? For example: someone comes to bat, runs the count to 3-and-2, and proceeds to hit a lot of foul balls—foul down the first base line, foul down the third base line, foul back to the screen, foul out of play in the first-base stands, foul out of play in the third-base stands, foul here and foul there. He’s making the pitcher work, is what he’s doing, and finally the pitcher either walks the batter or feeds him a cookie, a pitch that he can hit. I’ve seen that happen a lot. In fact, I remember one game many years ago, Yanks vs. Indians—I wish I could have seen that one, but the Yanks were playing in Cleveland. Here’s what happened, in the eighth inning.
The Yanks were at bat in the top of that inning, the score was tied, the bases were loaded with one out, and Phil Rizzuto was at bat. He ran the count full—and then he started fouling off one pitch after another. I was listening on the radio, and I couldn’t help chuckling because the Scooter was spraying them all over the ballpark and giving the fans lots of souvenirs. He fouled off 27 pitches in a row—and then I guess he got tired of all this, because on the 28th pitch he slammed one against the left-center-field wall, clearing the bases with a resounding stand-up triple. A moment later he scored on a wild pitch—and the Indians had to go to their bullpen.
Conversely, the pitchers who throw very few “action pitches” would seem to be the ones who don’t need a lot of pitches to get outs. They will throw 100 or fewer pitches—often 80 or 85 of them—and often it’s one pitch, one out. Ed Lopat used to do this a lot, and his “thing” was to let the batters get themselves out. These are pitchers with excellent control and command of their stuff, so they don’t have to work so hard at it. I seem to remember Mike Mussina as another such pitcher. And it doesn’t matter whether one is a fireballer or a finesse pitcher, or whether one goes for the quick strikeout or pitches to contact—if you don’t throw a lot of pitches, and you get the batters out, you’re doing a good job, and you come away with a nice W in the box score.
So I think that in calculating the number of “Action pitches” you need to factor in the foul balls. 8)


#4

dave, you’ve done me a favor by asking your question! I wondered why anyone would ask that, so I went into the code that was running and looked to see what was going on. JEEZ! I was running the wrong SQL statement!

Now that I have it corrected, all I’ve got to do is get the correct report loaded.

http://www.infosports.com/scorekeeper/images/actpit1.pdf

Now I can answer your question. :wink:

The number of action pitches is divided by the number of pitches.

Thanx again for not just accepting what’s up there! I have to admit that I’m embarrassed, but I’m not too proud to admit I made a mistake. Life’s too short to worry about such things. :oops:


#5

[quote=“Zita Carno”]Hmmm…let’s look at that for a moment.
You say that an “Action pitch” is one where something happens. You also say that the pitchers who have the highest percentage of such pitches seem to be the less skilled. What I’m reading into this is that these are pitchers who throw an awful lot of pitches in the course of a game—have you factored in the number of foul balls hit by the batters? [/quote]

No. I haven’t done that because the guy who dreamed up the metric didn’t include that as his definition of an action pitch. What action do you think there is on a foul ball where there’s no put out? I suppose you could say there’s action on it, but no more than a pitch that goes to the backstop with no runners on.

[quote]For example: someone comes to bat, runs the count to 3-and-2, and proceeds to hit a lot of foul balls—foul down the first base line, foul down the third base line, foul back to the screen, foul out of play in the first-base stands, foul out of play in the third-base stands, foul here and foul there. He’s making the pitcher work, is what he’s doing, and finally the pitcher either walks the batter or feeds him a cookie, a pitch that he can hit. I’ve seen that happen a lot.

Conversely, the pitchers who throw very few “action pitches” would seem to be the ones who don’t need a lot of pitches to get outs. They will throw 100 or fewer pitches—often 80 or 85 of them—and often it’s one pitch, one out. Ed Lopat used to do this a lot, and his “thing” was to let the batters get themselves out. These are pitchers with excellent control and command of their stuff, so they don’t have to work so hard at it. I seem to remember Mike Mussina as another such pitcher. And it doesn’t matter whether one is a fireballer or a finesse pitcher, or whether one goes for the quick strikeout or pitches to contact—if you don’t throw a lot of pitches, and you get the batters out, you’re doing a good job, and you come away with a nice W in the box score.
So I think that in calculating the number of “Action pitches” you need to factor in the foul balls. 8)[/quote]

What you say about there being PAs where there’s a lot of foul balls is true. But it still doesn’t meet the definition. Remember, I didn’t dream up this metric, some other fellow did. :wink:

Now if you’d like to come up with your own definition, I’ll do my level best to run it with the data I have. How would you like to see foul balls get factored in?

Do you want all foul balls, or foul balls with 2 strikes? I can’t tell which strikes are foul balls before there are 2 strikes, but I can after 2 strikes.


#6

Okay, now we’re getting into this a little deeper.
Suppose we have a batter who’s up there at the plate, and he’s taking, and the count gets to 0-and 2 or 1-and-2. He hasn’t yet swung at a pitch. So the next pitch comes in, and he goes after it, and he hits it foul down the first base line. That is when we would start the foul-ball count. Now suppose we have a hitter who will go after the first pitch no matter where it is. The pitcher throws, and the batter hits it down the right-field foul line into the seats—foul. That’s strike one, and we begin the foul-ball count with that pitch. So we have two different scenarios, two different spots where we’d start the foul-ball count, just as we would have two different kinds of pitcher, the one who throws a lot of pitches that get fouled off and the one who doesn’t.
One point to remember is that when a pitcher throws a lot of pitches that get fouled off it means that the batter is getting a good read on them and is awaiting a nice fat one in his comfort zone that he can hit. :slight_smile:


#7

[quote=“Zita Carno”]Okay, now we’re getting into this a little deeper.
Suppose we have a batter who’s up there at the plate, and he’s taking, and the count gets to 0-and 2 or 1-and-2. He hasn’t yet swung at a pitch. So the next pitch comes in, and he goes after it, and he hits it foul down the first base line. That is when we would start the foul-ball count. Now suppose we have a hitter who will go after the first pitch no matter where it is. The pitcher throws, and the batter hits it down the right-field foul line into the seats—foul. That’s strike one, and we begin the foul-ball count with that pitch. So we have two different scenarios, two different spots where we’d start the foul-ball count, just as we would have two different kinds of pitcher, the one who throws a lot of pitches that get fouled off and the one who doesn’t.
One point to remember is that when a pitcher throws a lot of pitches that get fouled off it means that the batter is getting a good read on them and is awaiting a nice fat one in his comfort zone that he can hit. :)[/quote]

Well, I could do the 1st one, but the 2nd is out of the realm of possibility, at least for me and my data. Sorry. But let me ask a few more questions to see if I can maybe still come up with a way to do this to at least get something to look at.

Maybe I’m misunderstanding, but it looks like your trigger is the 1st pitch fouled rather than anything else. The reason I say that is, there’s really no way there’s any hitter who literally swings at every 1st pitch. It may seem like some of them do that, but there’s no way its true. However, there definitely are hitters who don’t particularly like to allow a good pitch to go by without trying to hit it, no matter whether it’s the 1st pitch or the 6th. Usually its referred to as a patient hitter vs an aggressive one.

I do a metric on that which can be seen at http://www.infosports.com/scorekeeper/images/cbatting.pdf

I can also look at what each hitter and each pitcher does as well.

Page 70 http://www.infosports.com/scorekeeper/images/fbatting.pdf

page 65 http://www.infosports.com/scorekeeper/images/fpitching.pdf

I can kind of understand about it when it’s a foul after 2 strikes, but to say it’s a trigger at that point but only for 0-2 or 1-2 pitches and not 2-2 or 3-2 pitches has me a bit stumped.

But, I’m not quite sure at why the 1st foul ball is a trigger, and then only for that particular kind of hitter. What if a hitter who’s normally patient fouls off the 1st pitch? Why wouldn’t that start the count too?

I’m gonna do a little bit of “data mining” to see if its possible to spot batters who foul off a lot of pitches after 2 strikes. Be patient, sometimes it takes a couple of days to write the code and get it to work. :wink:


#8

Ok Zitaroo, I’ve jumped into the pit and come up with the 1st iteration of your foul ball theory, or at least what I have the capability to do.

http://www.infosports.com/scorekeeper/images/zita1.pdf

As you can see, that’s not in a “finished” form, but doing this stuff is usually a multi step process, and that’s not a bad step 1.

Now I have identified the outcome of all the at bats for all the hitters and pitchers, where there was at least 1 foul ball after there were at least 2 strikes on the hitter. I have to be honest and say I’m still not seeing how that can be described as an action pitch, but I’m pluggin’ along seein’ what kicks out. :wink:


#9

Here ya go Zita! Now I can tell which hitters have had ABs where they’ve fouled the ball, or pitchers where they threw pitches that were fouled, all with at least 2 strikes.
http://www.infosports.com/scorekeeper/images/zita2.pdf

Can you explain how knowing that helps me compare players or rate them? I’m really sorry but I’m not seeing it. :frowning:


#10

I think this is an interesting stat and I certainly don’t mean to be overly critical, but I am at a loss why you have a bunch of bad things added together with outs (a good thing) and subtract caught stealing (a good thing). Based on that formula a pitcher could have 100% Action Pitches and throw a perfect game by throwing 27 ground ball out pitches.

In fact in the stat sheet there are 2 pitchers Josh Roth and Mike Martel who threw 1 inning. Roth threw 8 pitches for a perfect inning and Martel threw 18 pitches with a walk, a HBP and a WP but he has the lower Action Pitch %. My suggestion would be to not not have a Action Pitch % but an Action Pitch/inning and not include the outs in the formula because every pitcher will get 3 per inning pitched. The other issue using the formula as is, is that it will favor a pitcher who throws a lot of pitches or 3-2 counts but doesn’t give up a lot hits, walks, etc.


#11

dave, “I” don’t have anything. :wink: This stat was totally conceived by a writer named Joe Posnanski, and all I did was take the formula he’d dreamed up, and run my data through it. But let me explain how I got to where I did.

Here’s exactly what he said. By action pitch, I mean pitch where something happened: Hit, walk, error, hit-by-pitch, sacrifice, strikeout, groundout, flyout, lineout and every other goofy thing you might see on the APBA unusual play charts*.

I took that and simply started pulling data. The only thing that really caused a problem was the outs. The way he stated it and the way I did it the 1st time, I included pick offs. I later got rid of that because I realized that I’d already accounted for POs when I computed the outs.

Another one I realized was a problem, was the stolen base attempts. At 1st I just counted them as action pitches, which they are. But later on I realized that wasn’t really valid because since CS was an out and I’d already counted the outs, it would be giving those pitchers who had had runners CS an unfair advantage, so I had to get rid of them.

In fact, I really think I should have included passed balls as well. I wouldn’t ordinarily, but since ROEs are included, I really feel passed balls should be too. Fortunately though, it really doesn’t make a great deal of difference for our team since we only allowed 24 PBs, so it wouldn’t make much of a difference.

However, here’s your chance to have an impact, at least on how I do the stat. If you agree that PB’s should be included, I’ll stick ‘em in and redo it. Aw heck, don’t bother. You know I already did it! :boingboing:

Hmmm. dave, I’m beginning to think you’re a numbers guy just like me, and I gotta tell ya that I think its great! To me, half the fun of looking at numbers is trying to determine if they simply make sense, and you’re obviously doing that. So, let me explain why I did things the way I did. BTW, in case you haven’t guessed, I really revel in explaining the numbers, at least how I see them. I figger, if you can’t explain ‘em, you darn sure shouldn’t be usin’ ‘em! :wink:

There’s a reason I include a lot of percentages. Its because people tend to understand them. I also use ratios, but for some reason, even though they’re really much the same as a percentage, they aren’t as easily comprehended. But, that doesn’t mean I have anything against your suggestion, but rather is that really the “best” way to display it?

That brings me to what you said about using innings as opposed to outs. Gotta confess something here. I have never been a fan of innings. I’ve always used them because they’re the traditional standard, but that doesn’t mean I like ‘em. :type: One reason I don’t like em’ is, when you deal in innings, there’s absolutely no way to get away from fractions of innings. When you work with outs, to me it seems that things just seem to be smoother. Its probably just preference, but that’s me.

As for this particular metric, the main reason I used outs was the way Mr. Posnanski described the criteria. When he talked about lineouts, flyouts, and groundouts, its pretty difficult to do those things in relation to innings. FI, if a P got 11 GO’s and 8 FO’s in a game, that would translate to 3 2/3 innings and 2 2/3 innings. Trust me, it can be done, but it sure looks bizzy when you try to show it. So, for me the answer was to simply use outs and be done with it.

However, that still doesn’t address your suggestion of action pitches per inning, so let’s see what we can do about that. How about a compromise? Having both the percentage and Action Pitches per Out is kind of a meeting of the minds. :inlove:

Now let’s address that last issue. No matter how hard one tries to eliminate numerical bias, every metric will always favor on group as opposed to another, but that’s what they’re supposed to do. Unless a metric is run for just one player, there will be an inevitable comparison, and that will lead to an inevitable ranking or queuing. So very often what ends up happening is, more and more factors are jammed into the metric to try to get them a unbiased as possible, and in doing that a couple things happen.

One is, the metric gets so busy, its difficult to understand for folks who aren’t numbers wonks and just want an answer. I’m a bad one for that because more often than not, I try to include the factors to show how the final numbers are computed. So in order to not “turn off” an observer, many times metrics are presented in their final form only, such as a BA or an ERA, and in this case action pitches.

That brings something else into play. Without seeing how the numbers are computed, its really impossible to compare players with any degree of validity. FI, is it really fair to compare a player who’s only thrown a couple of innings in 2 years to one who’s thrown a hundred?

So, in the end what I do is not try to make every metric “perfect”, but rather present many different things for consideration. An example is, if you go to pages 131, 133, and 140 of http://www.infosports.com/scorekeeper/images/cpitching.pdf you can see a whole lot more about not just 3-2 counts but any 3 ball count than you prolly want to know, and darn sure more than I could ever consider in conjunction with action pitches.

So in the end, I’m afraid that this particular metric is going to have to favor whomever it happens to favor, and that depends on how you view what an action pitch is supposed to show. To me its far different than pitching to contact, and certainly isn’t much of an indication of a pitcher’s success nearly as much as many other metrics. But it sure does show who would likely be more fun to was pitch a game. If you like a game where there’s a lot of thinking and going on, you’d prolly favor pitchers who throw a low percentage of AP’s. However, if you like a lot of action, the pitchers who throw a lot of AP’s are probably more to your liking.

Howz that dave? Don’t forget to take a look at the slightly new format here =è http://www.infosports.com/scorekeeper/images/actpit3.pdf