The Count of Counting Counts
Posted: April 21st, 2009 | Filed under: Baseball | 52 Comments »
The following is my rather tedious effort to look at pitching counts and somehow become a smarter baseball fan. It is rambling, and there are quite a few numbers in it, and there’s a good chance that it will not interest you at all. This is what happens when spreadsheets are placed in the wrong hands. Proceed at your own risk.
When I was a young, an old baseball scout told me there were two tricks to looking like the smartest baseball fan in the crowd. The first, he said, was always watching the outfielder after a fly ball was hit. This will tell you how well the ball is really hit. If you see the outfielder racing back, yes, the ball was hit well. If you see him standing in place, no, the ball was not hit well.
This, he said, would keep you from screaming madly on routine fly balls and basically looking like a yutz.
The second thing he told me was this: Always watch the count. He said that the whole secret of baseball, all the mysteries, all the intrigue, all the hesitant swings, all the home run blasts, all the perfect pitches on the outside corner … all of them can be anticipated and appreciated by simply following the count*.
*I got a similar lesson from an old big league pitcher, Al Fitzmorris. He began his career as an outfielder and a hitter, and that didn’t work out too well for him. So he became a pitcher, and a successful one. He won 77 games in the big league with, as he calls it, limited stuff. But Fitz never stopped wanting to be a hitter, and years after his career ended he summed up his stalled hitting career this way: “I would have been fine if I could have started every at-bat with a 3-1 count.”
There are a few simple tricks that most baseball fans innately use when following the count. Everyone knows that a 3-1 count is good for hitters — “big pitch coming here,” is what announcers usually say — and everyone knows that pitchers have a big advantage when the count is 1-2 (“Got him in the hole”). Everyone knows that batters need a green light to swing 3-0, and pitchers don’t want to throw anything too good on 0-2, and that the runners may be going on a full count.
But, the scout told me, the deeper you delve into the count, the more you can learn about this great game. And so, I decided to delve deep. Real deep. Deeper than any sensible person would delve. I used the wonderful baseballreference.com to break down every count combination this decade — 2000-2008. And I tried (at times unsuccessfully) to see what I could learn from these combinations.
I should say, for those of you scoring at home, that there have been 1,690,302 action pitches this decade (not counting the start of the 2009 season). 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*.
*Which I understand are gone now … I never really played APBA Baseball so I don’t know. But I always liked the concept of an unusual plays chart. I want one of those for my life.
So here’s what follows: The count, the percentage of the time that count is the action pitch, and what batters hit on that pitch (the basic batting average/on-base percentage/slugging percentage). The numbers, as I say later, are a bit misleading, but they still can give you a good sense for how the game works.
Here we go:
First pitch.
Action pitch: 12.5% of the time — one out of every eight at-bats.
Batters hit: .338/.344/.547.
There is probably more grumbling at the park about batters swinging at the first pitch of an at-bat than any other time. The Kansas City Royals once had a prospective owner meet with manager Tony Muser and suggest to him that his batters needed to stop swinging at the first pitch. This, of course, is ludicrous. About one of out every five home runs hit in the game are hit on the first pitch of an at-bat.
Then again, about one out of every five double play grounders you will see are also hit on the first pitch. And that’s why there are so many complaints.
To get this out of the way: The offensive numbers listed above — all the numbers here — can be misleading because they only count balls that were HIT IN PLAY. Foul balls do not count. Swings and misses do not count (except when there are two strikes). Pitches that are called balls do not count (except when there are three balls). And so on. So the batting averages are naturally going to be much less with two strikes. And on-base percentages are going to be much higher with three balls.
Still, there are some cool things to see. There are real advantages, for instance, for batters who put the ball in play early in the count. Once they get two strikes on them, the averages go way down.
Here’s a statistic you can ponder if you like:
OPS for batters putting the first pitch in play: .891
OPS for batters who do not put first pitch in play: .739
So, for someone to decide to never swing at the first pitch … no, that’s probably not the best strategy.
0-1 Count.
Action pitch: 9% of the time — one out of every 11 at-bats.
Batters hit: .317/.326/.485
Well, this was a bit of a surprise to be: Batters would hit quite well when behind one strike. And this gets into what I was saying about hitting early in the count: As you can see, there is not a drastic difference if the batter is hitting behind 0-1 or ahead 1-0.
1-0 Count
Action pitch: 7.6% of the time — one out of every 13 at-bats.
Batters hit .339/.340/.563.
See? Batters do hit with a bit more power when ahead 1-0 — they are also a little bit less aggressive — but the point here is that the game does not shift dramatically to the pitcher until he gets two strikes on a batter. You always hear people say how important it is to get that first pitch strike … and it is hugely important. But getting that second strike is what turns an at-bat around.
0-2 Count
Action pitch: 7.7% of the time — one out of every 13 at-bats.
Batters hit: .162/.173/.236
Now, you can see the pitcher taking control. There are different philosophies about what to do with an 0-2 pitch. There are some pitching coaches and pitchers who think that this is absolutely the time to go for the strikeout pitch … the nasty slider tailing away, the split-fingered fastball in the dirt, the fastball up around the eyes. But there are others — and I tend to agree with this — who think that batters are so defensive at 0-2, that this is perfect time to go get them with a pitch over the plate (especially with pitch counts being SO important in today’s game).
This might be the most amazing statistic in this whole bit: Batters on an 0-2 count hit home runs once every 79 at-bats.*
* My favorite pitcher, Greg Maddux, gave up 11 home runs in his entire career 0-2 — that’s in more than 1,600 at-bats. No, he did not like wasting pitches. Here’s another good Maddux statistic: He only walked 45 batters in his entire career after getting ahead 0-2. Maddux had a 32-1 strikeout-to-walk ratio after he got ahead 0-2.
2-0 Count
Action pitch: 2.7% of the time — one out of every 37 at-bats.
Batters hit: 351/.351/.625
Well, this is a interesting situation … batters very, very rarely put the ball in play on a 2-0 count. In fact, batters don’t put the ball in play much more on a 2-0 count than they do on a 3-0 count. When they connect, though, they do connect hard … batters bang home runs once out of every 16 at-bats.
And you know, if I was a batting coach, I would want my batters to be a bit more aggressive on 2-0. Because here’s something else … pitchers only very, very rarely hit a batter with the count 2-0 (one out of every 179 or so at-bats). That tells me they are simply looking to get a pitch over the plate to get back into the count. It sure seems to me that 2-0 is an underutilized opportunity for hitters.
1-1 Count
Action pitch: 8.8% of the time — one out of every 11 at-bats.
Batters hit: .325/.330/.512
This is more or less a repeat of the first pitch. I have heard scouts and players say that, generally speaking, the most important pitch of the at-bat is the third pitch. And there does seem some truth to that. Most of the time (roughly 54% of the time), batters face a 1-1 count going into the third pitch. And the next pitch will, pretty often, determine the fate of the pitcher and the batter. Look what happens if a pitcher gets a strike:
1-2 Count
Action pitch: 13.6% of the time — one out of every 7 at-bats.
Batters hit: .177/.185/.263
Yes, pitchers are dominant in the 1-2 count. And this is actually the most common situation in baseball … a 1-2 count action pitch. And the batter is all but helpless. But when the third pitch is a ball …
2-1 Count
Action pitch: 5.6% of the time — one out of every 18 at-bats.
Batters hit:.337/.338/.554
Yeah, that’s a sizable difference. Batters hit 160 points better and slug twice as much when that third pitch is a ball rather than a strike. Batters may not know these numbers, but they instinctively know how much their chances go up when the count goes to 2-1.
Here’s a fun experiment: Next time you’re at a game, watch the batter’s reaction when the count goes from 1-1 to 1-2. They will, often, hit their bats with their hands or kick at the dirt or gripe at the umpire. I’ve never counted but I would bet that batters visibly reaction more than half the time.
3-0 Count
Action pitch 2.4% of the time — one of out every 42 at-bats
Batters hit: .390/.958/.780
A few facts about the 3-0 pitch.
1. Batters put the ball in play on 3-0 only about 7% of the time. In fact, over the course of a season, you will only see batters put the ball in play about 300 times on 3-0 … that’s about 10 times per team, per season. it does seem like in today’s game, lots of batters get the green light on 3-0, but the numbers say that you really don’t see them hit the ball on 3-0 very much.
2. When you DO see them hit it, there’s a good chance you will see them hit it a long way. Batters hit 3-0 homers roughly one out of every 10 at-bats.
3. Jim Thome, in his long career, has only put the ball in play 58 times on 3-0. He has hit SIXTEEN home runs. For the record, that’s one homer per every 3.6 at-bats.
4. Or how about Mike Piazza. In his whole career, he only put the ball in play FOUR TIMES on 3-0. That’s all. Four times. Apparently nobody was throwing Piazza a good pitch 3-0. And that was a good idea: Two of them were home runs.
5. Batters almost never get hit on the 3-0 pitch — one out of every 640 plate appearances.
6. Very, very few base runners try to steal on the 3-0 pitch, for obvious reasons, but those that do are ultra-successful — 89%. What’s interesting is that the 3-1 pitch — which has conventionally been called the perfect pitch to steal on — is anything but: Only 61% of base stealers are successful on 3-1.
2-2 Count
Action pitch: 12.9% of the time — one out of every 8 at-bats.
Batters hit: .194/.199/.299
You will often hear announcer say “He evened the count at 2-2.” But there is nothing really even about a 2-2 count. The pitcher is still firmly in control. If a pitcher consistently can make it so the action pitch is always 0-2, 1-2 or 2-2, he will do very nicely for himself and make quite a lot of money.
3-1 Count
Action pitch: 4.9% of the time — one out of every 20 at-bats
Batters hit: .355/.691/.638
Well, here is the ultimate hitter’s pitch … I’ve called a few home runs over the years, wowing friends and impressing strangers, but it’s really not that hard. When you see a good hitter at the plate (or a lousy pitcher on the mound) and a 3-1 count, go ahead, make the call. If you want to play the 3-1 homer game yourself … here are a few good players to consider:
Josh Hamilton: .611/.833/1.167
Chase Utley: .444/.713/.852
Adam Dunn: .392/.777/.908
Alfonso Soriano: .435/.694/.848
Jermaine Dye: .406/.656/.767
Full Count
Action pitch: 12.3% of the time — one out of every 8 at-bats
Batters hit: .229/.468/.381
A couple things interest me here. One, I find it interesting that one out of every eight or so at-bats goes to a full-count. That seems like a lot to me … that means you should see, seven to 10 full counts every single night. I wonder if that number has gone up through the years. I have no idea how to find out.
Also, it really is telling — again and again — that hitters really do swing defensively with two strikes. In total, with two strikes, batters hit .190/.257/.293.
I think that’s one of the takeaway as a baseball fan. Pitchers do not (and should not) give up many two strike hits. And they certainly should not give up two-strike extra base hits.
Another takeaway is that until the pitcher gets two strikes, the advantage* is with the hitter. Batters hit .334 and slug almost .600 when not facing two strikes.
*Of course, it’s tricky when you say the hitter has an advantage … even Ted Williams, the purest hitter who ever was, was retired more often than he reached base (his .482 on-base percentage is the best ever, but it means he also had a .518 out percentage). But this is something that’s tricky about baseball — pitchers always have an advantage. You will hear announces say all the time: “This just goes to show you that good pitching beats good hitting.” I have no idea what this means … it is literally true, but bad pitching also beats good hitting. Pitching beats hitting. That’s the game. That’s why even when a manager makes what seems an obviously dumb pitching move, it is still likely to work.**
**This “still likely to work” rule is not in effect when it comes to Royals manager Trey Hillman, who has now gone NINE DAYS since pitching Joakim Soria. NINE DAYS. Maybe he’s saving Soria for private functions, birthday parties, bar mitzvahs and so on.
And the last takeaway is this: I spent way, way too much time on this.
Loading ...
No you didn’t Joe. I enjoyed it quite a bit. It’s fascinating stuff, putting data with the 2-1 vs. 1-2 especially.
Oh, and circle me, or whatever.
Great post. Very interesting stuff.
This immediately brought to mind Craig Burley’s work for The Hardball Times back in 2004:
http://www.hardballtimes.com/main/article/the-importance-of-strike-one-part-one/
http://www.hardballtimes.com/main/article/the-importance-of-strike-one-part-two/
I love this stuff.
This really belongs in the managerial moves post, but i’m going to put it here:
First we’ll have Meche,
And then we’ll have Greinke
Then Kyle Davies,
Then two that are stinky
Then back will come Meche,
Who’s followed by Greinke
And followed, this time,
By three that are stinky.
[...] Travel Blog created an interesting post today on The Count of Counting CountsHere’s a short outline…old baseball scout told me there were two tricks to looking like the … Real deep. Deeper than any sensible person would delve. [...]
One of the most amazing statistics I’ve seen:
In 2007, Placido Polanco hit .402 (39 for 97) after being behind in the count 0-2.
[...] Joe Posnanski created an interesting post today on The Count of Counting CountsHere’s a short outlineThe following is my rather tedious effort to look at pitching counts and somehow become a smarter baseball fan. It is rambling, and there are quite a few numbers in it, and there’s a good chance that it will not interest you at all. This is what happens when spreadsheets are placed in the wrong hands. Proceed at your own risk. When I was a young, an old baseball scout told me there were two tricks to looking like the smartest baseball fan in the crowd. The first, he said, was always watchi [...]
I still maintain that Joey Gathright should be forced to take a strike, Wade Boggs-style, 95% of the time.
Let me get this straight, you are putting up a post on 1-2 vs 2-1 on the same day that the Royals tie a record by gidp? Did you do a team breakdown on that data? Would be interesting to see if teams that work the count (Red Sox, Yankees) do in those situations.
Fascinating. I can never get enough Poz.
Got me thinking — everyone assumes that the count is “stateless.” That is, that a 2-1 count is always the same. I wonder if there’s a difference in approach based on what the *last* count is. I’d be interested in:
1-1 to 1-2
1-1 to 2-1 (you mentioned these)
2-0 to 2-1 (is it a disproportionately big win for the pitcher)
2-2 to 3-2
3-1 to 3-2 (does your perception of “full count” change?)
etc.
Also interesting would be to understand what the most common pitch count sequences are. I.e., of all strikeouts, how many are 1-2-3, how many have one ball mixed in, etc. I wonder how common they all are?
… I’m about to say that we should build a Markov model and train it on the entire body of pitching data, but I’ll limit the nerdology.
Very interesting, I will definitely think of this the next time I watch a game. Thanks
The other set of numbers you can do, is the expected BA/OBP/SLP *before* each pitch, rather than the one when you know it’s the end of the PA. So the 0-0 numbers would cover all PA, the 1-0 numbers would cover all PA when the first pitch was a ball, and so on. This means you don’t have that massive jump in OBP when there are 3 balls. Neither set of numbers is better, they just illustrate different things.
My favorite story-
http://www.baseball-reference.com/boxes/CHN/CHN199707221.shtml
The Cubs coaches had reputedly seen similar numbers to Joe’s: they had noticed that Greg Maddux had worse numbers when batters put the ball in play early in the count. So, they told the Cubs hitters to get after him. Maddux pitched a complete game win, 76 pitches. They didn’t try that again…
I played on a fast pitch team that won a couple of state titles in the early 90s here in Missouri. One of the guys I played w/was always telling everybody to “take ’til you get a strike”. Drove me nuts. In about 1994 or so, Sports Illustrated ran an article on the effectiveness of hitting a good first pitch, which I made certain to give to him. It makes no sense to watch what might be the best pitch you’ll see go by just because that’s what someone taught you in Little League.
The stolen base on 3-0 vs. 3-1 stat you bring up is interesting, but I wonder if it just shows selection bias. Since 3-1 is considered the perfect pitch to steal on, there are lots of attempts by runners of various caliber. As you point out, very few base runners try to steal on the 3-0 pitch, so I wonder if the ultra-successful rate is due to the fact that this isn’t considered a good pitch to steal on, so the guys who do run are guys like Carlos Beltran, Tim Raines, or Willie Wilson, who are probably going to be successful no matter what count they run on.
I was surprised to note the .317 BA versus .326 OBP on first pitch action pitches. What is that? HBP? Are there more first pitch HBP than first pitch sacrifice bunts/flys? On 0-1 count the split is .339/.340, which is more what I would expect.
You’re right, DSoM. Almost no one should eschew swinging at the first pitch most of the time.
When I played Little League, I remember swinging at the first pitch quite often without being reprimanded. It was only the very worst hitters that were encouraged to take a strike as a general rule. (Of course, it was never good for anyone to swing at a first pitch out of the zone.) I think that is how my irrational Gathright prejudice originated.
Err, sort of answered my own question here. Sac bunts aren’t included in OBP calculations because the player is being asked to make an out, and since there were 1672 HBP vs. 1365 SF in 2008, the split holds. Guess the MLB is more bean happy than I thought…
“But there is nothing really even about a 2-2 count. The pitcher is still firmly in control.”
THANK YOU, Joe. This is a huge pet peeve of mine. The numbers being the same obviously does NOT make the count “even” – it’s like saying 2/4 = 2/3. I feel the same about 1-1, of course..
Pete shared this story:
“The Cubs coaches had reputedly seen similar numbers to Joe’s: they had noticed that Greg Maddux had worse numbers when batters put the ball in play early in the count. So, they told the Cubs hitters to get after him. Maddux pitched a complete game win, 76 pitches. They didn’t try that again…”
I think this story illustrates the problem with making inferences based on Joe’s analysis (as interesting as this post is). In particular, at bats are not experiments, and batters are not randomly assigned to swing or not swing on a given count. Thus, the same things that are likely to cause a batter to swing on a first pitch (e.g., the pitch looks hittable) are likely highly correlated with things that lead to batting success (e.g.,the pitch was hittable).
Assuming most of the difference in Avg and OBP before 3 ball counts is HBP, we see a real pattern emerge where pitchers are obviously trying to move the batter off the plate when they get ahead in the count, especially with 2 strikes. The HBP % is roughly .05 in even counts (0-0, 1-1, 2-2), neglible when the pitcher is behind (.01 when 1-0 and 2-1 and .00 when 2-0), and actually pretty high when ahead in the count (.09 when 0-1, .11 when 0-2, .08 when 1-2).
Joe,
I think those 3-2 stats are skewed because you obviously can’t have a count greater than 3-2, so it’s the only count that can occur more than once per plate appearance.
“Jim Thome, in his long career, has only put the ball in play 58 times on 3-0. He has hit SIXTEEN home runs. For the record, that’s one homer per every 3.6 at-bats.”
Any of those against Farnsworth?
If that were right, the same would be true on 0-2, 1-2 and 2-2.
I actually expected to see more 3-2 counts.
Joe, I don’t know if all the conclusions you drew were correct, but I sure love reading them.
“So, for someone to decide to never swing at the first pitch … no, that’s probably not the best strategy.”
Every year during Spring Training I tend to focus on one or two players to root for during the year. This year, the bulls eye landed squarely on Brett Gardner’s small frame. He’s the kind of speedy, scrappy player that I have always loved. I wanted him to win the CF job for the Yankees sooo much. And he did. I was overjoyed.
Now I just want him to swing at a first pitch every once in a while. It seems that every at bat I see he takes a fastball down the middle for a first strike. I can tell from BR that it isn’t true. He has made a swing on 5 first pitches out of 54 PA, but I can’t recall having seen one of them. He has fallen behind 0-1 56% of the time so far this season. For a guy that needs to get on base, he is in the hole way too much for my liking.
Still…did you see that catch he made on Giambi’s first AB last night? Sweet!
Also: I feel like the SB percentage’s being higher on 3-0 than on 3-1 has something to do with 3-1 pitches getting put into play more than twice as often as 3-0 pitches. If I’m going on 3-1 … well, I’m getting thrown out, because I’m fat. But if someone else is going on 3-1 — while I don’t know if the percentage of balls put in play that are walks is roughly the same for both counts — if someone’s going on 3-1, there’s a greater chance the batter makes contact, robbing that someone of a steal-attempt outcome. If someone runs on 3-0, they’re twice as likely to draw a throw, evening out the odds a little? Maybe? Someone run with this.
@John: It’s also worth noting that almost all intentional beanings are on the first pitch, which probably loads the HBP average there. Sac flies are probably more evenly spread out across the count.
Also: I second Bryan Adams (…) interest in the flow of a count. How does an at-bat change if the pitcher worked full from 3-0 vs a batter from 0-2, etc?
This is what makes the decision not to pitch Soria on Sunday all the more mystifying. Last night, he doesn’t want to bring him in for the eighth inning because he didn’t want it to be a 30 pitch affair that might make him unavailable today if a save situation arose.
But on Sunday, we had the freaking off day on Monday!!!! There is no possible excuse. And it ticks me off more as time goes by. I really need to think about other things.
@Chris: Any 2-strike count can occur a potentially infinite number of times per plate appearance, not just 3-2. Foul off an 0-2 pitch and the count is still 0-2.
Very, very interesting. I will think about that when I see a game.
I have a question: How is possible to have a OBP of .326 and a average of .317 in a 1-0 count?
Forget about my question of before… I forget the HBP
hitters don’t actually have a .958 OBP off of a 3-0 count, do they?! i know that usually that will be a “take,” leading to either a walk (success) or a 3-1 count (no action), but i would not have guessed that less than 5% of 3-0 counts result in swings that lead to an out, wow!
those thome and piazza stats are pretty wild too.
joe, did you come across any players who were particularly poor on 3-0 counts (relatively speaking, of course)?
As a long-time Orioles fan, it used to drive me nuts to watch Cal Ripken keep his bat on his shoulder for first pitch strikes down the middle of the plate. Towards the end of his career he adjusted and it seemed like he gained an extra hit a series by jumping on the occasional first pitch.
Weird, because my Little League coach always called 2-0 the “hitter’s pitch.”
Great stuff, Joe… Absolutely incredible to read for us stat-nerds…
Just to add to the discussion, I live in Fargo and get to see a lot of Twins broadcasts, and am absolutely appalled at how fundamentally bad Delmon Young is… Offensively he seems to have no clue what the strike zone is, and defensively he doesn’t seem to have a clear thought on what to do when he does get a ball hit to him (and the fact he’s DEATHLY afraid of the wall, IMHO)…
It would be interesting to do research on players who those in the game would categorize as disciplined hitters vs. those with no approach… Gwynn/Boggs vs. the Delmon Youngs of the world… Of course, the second category would have such a small sample of “late-in-the-at bat” counts that it might be immaterial!! LOL
Keep up the great work for those of us that DON’T have the time to delve into the numbers…
It seems likely that you would also see the hit and run on 3-1 counts more often than others, so you’ll have a few guys getting thrown out because the batter missed or took a strike. I don’t think you can get much from those numbers unless you gather the data from a select group of quality base stealers.
Chris (@21), 0-2. 1-2, and 2-2 can also occur more than once per PA since fouls on those counts don’t change them.
B.E. Earl, that *was* a sweet catch last night. If GGBG (Gritty Gutty Brett Gardner, as Peter Abraham calls him) gets his OBP up to .350, even .340, he’ll be a great addition.
I’ve learned you walk Josh Hamilton with a 3 – 1 count.
Interesting. Obviously stats are much worse with 2 strikes, because its the only time you can strike out. Would be curious to see the splits on balls put in play with 2 strikes (action pitch but not a strikeout)- do batters really hit that defensively or just strike out a lot?
Brett – You beat me to it. i think you’re right. 3-1 is when some doofus like, I dunno, Kevin Millar, decides it is a good idea to steal. It is not.
I think there is a logical error here that may be skewing peoples’ perceptions of how meaningful these numbers are.
Because these stats only allude to *action* pitches, before 2 strikes (or 3 balls) they can *only* refer balls in play. A ball in play is much more likely to result in a positive result for a batter. So for Joe, for instance, to point to this:
OPS for batters putting the first pitch in play: .891
OPS for batters who do not put first pitch in play: .739
as a refutation to the argument against swinging at the first pitch forgets that there is a difference between swinging at the first pitch and putting the first pitch in play. I would guess that a not-insignificant number of first pitch swings result in foul balls or missed swings (and thus no ball in play), and which therefore result in an 0-1 count.
Similarly, the stats enumerated by Joe here where there are 2 strikes are artificially skewed against the batter because they are limited to instances where the pitch is an “action” pitch. Because an action pitch can be a strikeout (which constitute a non-insignificant number of plays), the numbers will look worse because they don’t include plays where the pitch (except on full counts) was either a ball (thus a positive outcome for the batter) or a foul ball (a neutral outcome for the batter).
Great stuff as always, but you don’t really present any evidence that hitters swing defensively, as their lower two-strike averages could theoretically be all attributable to strikeouts. You work hard enough already, but it’d be great to see the averages of two-strike hitters on non-strikeouts.
I had a feeling something was amiss as I was reading this that I was going to try to put my finger on and then put into words. McKingford (#41) has, more or less, given it voice already but it makes me wonder what a better way to analyze the problem would be.
If we look at Joe’s analysis, we get the results of first pitch swings for action plays and we all know the results of first pitch swings for non-action plays is an 0-1 count. We know the possible outcomes for not swinging – the 0-1 count, a 1-0 count, a HBP.
What we need to know is the percentages of these different scenarios. What percent of the time is a first pitch swing resulting in an action play and what percent of the time is it resulting in an 0-1 count? What is the breakdown for not swinging?
If your first pitch swing results in an 0-1 count 50% (an arbitrary number) of the time, swinging at the first pitch has an immediate positive outcome only 17% of the time. The other 83% of the time, you’re in worse shape (even if only marginally so, to start). You give yourself no chance to get to those juicy 2-0 and 3-0 counts and limit your chances of seeing the mighty 3-1 count, taking away a lot of your future advantage.
If you can quantify the chances of not only an action play but the chances of each of the counts that can result from swinging or not swinging, you can sort of follow all the outcomes and branches and make a much more precise calculation of the value of swinging at a first pitch strike, laying off a first pitch strike or the decision to swing or lay off any particular count.
Even this analysis ignores things like the advantages to be gained in working a pitch count and that the numbers, as given above, may be as good as they are due to the current first pitch selectivity.
If you just want to look at numbers, though, ignoring non-action outcomes and the eventual action plays that result in future counts render these numbers, especially in early counts, meaningless.
McKingford said:
“I would guess that a not-insignificant number of first pitch swings result in foul balls or missed swings (and thus no ball in play), and which therefore result in an 0-1 count.”
I think you’ve been watching Vernon Wells, where he either watches a first pitch strike sail right through his wheelhouse, or he fouls off something that was heading a foot high and two feet wide.
Jimbo has had 16 HR on 3-0 counts, now go see how many of A-Rod’s 553 homers have come on a 3-0 count.
Sorry to follow up on my last point, but I was a bit rushed…
The data Joe presents really is interesting. I just don’t think it necessarily supports the conclusion Joe comes to – ie. that players with 2 strikes swing defensively (I’m not saying they *don’t*, just that this data can’t be the basis for that conclusion). Joe’s conclusion would be correct if the data here was confined to action plays where the ball is in play. But with this data, 2 strike action plays can be either a strikeout (a not improbable outcome) or a ball in play. Any action play with less than 2 strikes, by definition, excludes strikeouts – so we would expect much better outcomes (and thus higher Avg/obp/Slg).
Similarly, any action play with 3 balls includes the possibility of a walk (just as any action play with less than 3 balls excludes the chance of a walk).
As someone suggested earlier, it would be interesting to contrast this data with the outcome of plate appearances for each particular count. That is, the outcome of all plate appearances with an 0-1 count is likely different than the data presented here which is limited to 0-1 count plate appearances that end on the 0-1 pitch.
I’ve read so much material praising the strategy of taking the first pitch that I’m quite surprised 1 out of 8 plate appearances have their action pitches on the first pitch. That’s the sort of goodness you learn when you read Joe Posnanski. :p
Years ago, Tom Boswell ran an article in which he defended Kirby Puckett (one of my faves) and his prediliction to swinging away on the first pitch.
But Joe’s breakdown is clear and interesting–another in a string of line drives.
On a 3-0 count, all balls lead to walks. All strikes and fouls lead to some other count. And as we’ve seen, hitters in general tend to mash when they swing 3-0 AND put it in play.
In the last nine days, Soria has not pitched at all. Jonathan Broxton has two five out saves. Any idea whether I think Hillman or Torre is the better manager?
Much said about the 3-0 steal, but I think the main reason for such a high percentage was missed. Caught stealing is diminished because potential caught stealers are “saved” by the base on balls.
Great article. Will make my day at the yard even more fun!
Everyone should consult Stephen Jay Gould’s “Full House: The Spread of Excellence from Plato to Darwin”; here’s a link:
http://www.amazon.com/s/ref=nb_ss_gw?url=search-alias%3Dstripbooks&field-keywords=Full+House%3A+The+Spread+of+Excellence+from+Plato+to+Darwin&x=&y=
He talks about the lucky grad students who examined batting averages, etc., instead of logging paleozoic fossils. His love of baseball shines through.
I’d love to see some numbers on how *pitchers* perform on varying counts. In particular, I would love to see how other relievers stack up against Carlos Marmol after being behind 3-0 or 3-1. I’ve seen him get more strikeouts after starting out with three balls than any pitcher EVER.