Casey Blake rules.

First?

Pretty good article. Maybe there is hope for ESPN after all.

I was surprised to not see Scott Hatteberg show up in any of those lists. I don't think he has ever swung at a first pitch - EVER - nor has he had an AB of less than 6 pitches.

At the opposite end, surprised to not see Dmitri Young.

Good article. Presents some interesting data, doesn't try to do more with it than can be done. It doesn't reveal any fundamental truths about baseball, but it's a good read. If more ESPN articles were like this, I would stop by their site sometimes.

I mean, I know Bonds' walking situation is completely anomalous, and he does swing at almost anything that is a strike when he has the chance, but isn't comparing his pitches per PA in non-walk PA's against everyone else's overall P/PA kind of pointless? If you take out the walks, EVERYONE'S P/PA will go way down.

Maybe breaking it up by intentional/non-intentional would be helpful.

I was surprised to not see Scott Hatteberg show up in any of those lists. I don't think he has ever swung at a first pitch - EVER - nor has he had an AB of less than 6 pitches.

Hatteberg takes a lot of pitches, but he also fouls a lot of pitches off. I think that's why he's not on any of the lists.

Durazo would have certainly been on those lists last year. In 2004, he completely changed his approach, his walk rate plummeted, and he saw great results. Interesting...

I don't think Dmitri Young collected enough plate appearances last year. I was thinking of someone similar, Randall Simon; I would love to see some numbers for him.

As far as Bonds, I too am suprised they didn't address the IBB issue. Although there are instances of players being intentionally walked and swinging anyway (I remember Dante Bichette hitting a home run this way).

I got the impression that Schwarz was comparing Bonds' non-walk P/PA to everyone's non-walk P/PA. He says that Bonds' overall P/PA, at 3.93 is "in the middle of the pack", and then compares the 3.52 non-walk P/PA to "the average" 3.60. I think he means that 3.60 is the average for non-walks, but he phrased it poorly.

Alan does really good lit searches. I'm always impressed with how well researched his stuff is.

I did like this article - mostly because he came back to the correct (or what I think is correct) conclusion, but he did a good job of showing other things.

While he removed Bonds' walks, even if he hda removed just Bonds' IBBs, Bonds' mark would have dropped considerably. His avg with 4-pitch walks was 3.93. Without them, that goes down.

back much further than even Ted Williams' famous mantra, "Get a good pitch to hit."

That wasn't Williams. Well, it sort of was - Williams carried that torch, but Teddy Ballgame always credited that to Rogers Hornsby - passing along the keys to being a great hitter. Ted said that Hornsby gave him that advice in Spring Training in Daytona Beach in 1938. (That is on page 23 of "The Science of Hitting" By Ted Willaims and John Underwood; revised, 1986.)

It's also on page 63 of my paperback version of "My Turn at Bat" (1988 ed.)

Maybe breaking it up by intentional/non-intentional would be helpful.

That is exactly what I though as I read the walks/non-walks calculations. Intentionalal base-on-balls are totally different since, almost exclusively they cannot be more than 4 pitches. With Bonds, sometimes they try to throw a strike sometimes, get one over, and then nibble to 3-1 and then issue the IBB, but otherwise, they're all 4 pitches, so they can be excluded as a separate event entirely.

Too bad Bellhorn didn't get a mention. He fits the idea of this article perfectly--his patience helps him get a pitch he can hit.

I'm sure everyone here knows the story of his 2004 season. He managed to keep his job in April by drawing enough walks to keep his OBP around .400 even though his BA was under .200. As the season progressed, his walks went down, but he also started hitting better, and for more power.

He really struck the ST group as a guy who got the most out of his ability by trying to take everything not down the pipe. And that's good, because he doesn't appear very capable of hitting stuff that's not down the pipe.

I wish MLB would use Questec to show record things like swings at strikes vs. balls.

Besides Ruth as a pitcher, is Hornsby the best position-adjusted hitter?

Good article, good writer.

Besides Ruth as a pitcher, is Hornsby the best position-adjusted hitter?

I don't think we have the data needed to do positional adjustments for players before the '70s. At least, I've never seen it. I bet Ruth, even as an outfielder, still beats him, though.

I think the current crop of slugging shortstops could also qualify as top position adjusted hitters.

Except the very fact that there's a "Crop" of slugging shortstops suggests they're not espcially anomalous for their period, right?

Hatteberg takes a lot of pitches, but he also fouls a lot of pitches off. I think that's why he's not on any of the lists.

A fouled off pitch still counts as a P during the PA, right?

Yes, although I wonder how it is counted when there are two strikes?

As others have said, that's a really well-presented article. Nice work. Post #4 puts it well.

Schwarz points out, "Almost all hitters fare far better when swinging at the first pitch--not because it's the first pitch, but because it's one they've decided they like." Why is it that there is such a significant batting-average gap between the first pitch and subsequent pitches? If the second pitch is one the batter has decided he likes and swings at it, why is the batter considerably less likely to get a hit? It's just one pitch later. Does the one pitch that's already part of the count have such a great effect on the pitch selection and batter's anticipation?

Regarding "no linchpin from which it can sweep an arc of understanding" and "baking in the concept before the kiln fires," it's nice to see a sportswriter who use metaphors sparingly and without mixing them. They're much more effective that way.

Frank Thomas would have been second to Bonds with .206 BB/PA, and he was at 4.23 P/PA. And of course he has a great history of patience. Last three years average of 4.26 P/PA. I would nominate him as the second most patient hitter.

Interesting that Jermaine Dye's P/PA is so high. And although it's his career high it isn't really that far from ususal for him. He had been at 4.12 and 4.13 previous two years, but has never really walked a lot. Interesting.

Well if nothing else the White Sox will have a patient middle of the lineup (Konerko's got a decent P/PA also).

Why is it that there is such a significant batting-average gap between the first pitch and subsequent pitches? If the second pitch is one the batter has decided he likes and swings at it, why is the batter considerably less likely to get a hit? It's just one pitch later. Does the one pitch that's already part of the count have such a great effect on the pitch selection and batter's anticipation?

The first pitch probably doesn't have the highest batting average for all hitters, I would bet it's the hitter's counts like 3-1, 2-0, etc. Conversely, guys hitting a first pitch probably do much better than guys behind in the count and swinging more freely, i.e. 0-1, 1-2, 0-2, etc.

A fouled off pitch still counts as a P during the PA, right?

Yes, but it wouldn't count for percentage of pitches taken or percentage of strikes taken. Hatteberg saw 4 P/PA this year, which was below his career average.

I believe I've once seen a chart showing batting averages for each count, but maybe I'm confusing it with the chart I've seen for runs expected for each combination of baserunners and outs. Is the former chart out there anywhere?

Am I the only one just dumbfounded by the stat that Vlad swings at 92% of strikes thrown to him? Especially when you look at the other four names on the list -- they all suck.

Vlad takes 41.3% of pitches. I think the 8% is: Strikes called/Total Strikes

Well if nothing else the White Sox will have a patient middle of the lineup (Konerko's got a decent P/PA also).

The problem with Paulie is that he often watches the first two strikes go by.

Ted said that Hornsby gave him that advice in Spring Training in Daytona Beach in 1938. (That is on page 23 of "The Science of Hitting" By Ted Willaims and John Underwood; revised, 1986.)

I have the Rogers Hornsby article, "Wait for a Good Ball" in the 1931 "Play the Game...the Book of Sport" book by Mitchell V. Charnley.

From the piece...on taking batting practice.

And don't think that because nothing depends on it you are free to cut at bad balls. Unless the pitch is in the good zone, your bat shouldn't leave your shoulder.

Just anecdotal evidence, but I can remember Dye having some loooong at bats where he fouled off many pitches. Maybe this could explain his high pitches-seen, but low walk rate.

"Getting a good pitch to hit is not patience (or pitch discipline, as I prefer to call it)". It's a plate appearance strategy. Discipline is a specific skill of telling strikes from balls, which manifests itself in 1) when the batter takes a pitch, how often was it outside the strike zone, and 2) when the batter swings, how often was the ball in the strike zone.

So, a batter should honestly look at his ability to be disciplined (as defined above), plus his other skills, and create an optimal plate appearance strategy. You cannot just look at BB/PA, or pitches/PA to evaluate this. That is why Bonds, probably the most disciplined player (as defined above), sees only an avg # of pitches/PA. His ability to make contact when he swings is another part of that.

Since we don't have the proper data to measure true discipline, it is probably better to just stop putting labels on such "sub-skills", and simply look at outcome rates--such as K/PA, BB/PA, etc. A batter is good or bad based on the totality of his outcomes/PA. "Discipline" has an effect on every single outcome, and AFAIK does not have a significant predictive value beyond those outcome effects...

Vlad takes 41.3% of pitches. I think the 8% is: Strikes called/Total Strikes

Right. So assuming that Vlad swings only at pitches in the strike zone (which is the assumption that box scores use for ball and strike counts), he swings at 92% of the strikes thrown to him. Or actually, much, much higher, since the denominator doesn't include all the strikes that he puts into play. So, essentially, if you put the pitch in the strike zone against Vlad, you know that there is nearly a 100% chance he's going to take a hack at it. Which was my point.

That's what so amazing about Vlad, he'll swing at, and hit, balls way out of the zone-hard!

Vlad at the plate reminds me a lot of Clemente. With both of them, it sometimes almost seems as if the meatball right down the pipe is the pitch they can handle the least. Clemente's eyes would just light up with delight when he'd see a pitch come his way about six-to-eight inches off the plate away, especially low and away. He'd unleash a full-body slam on that thing and rocket it down the right field line. Hello triple.

Hello, Steve!

Position-adjusted Honus Wagner would be pretty fierce.

Besides Ruth as a pitcher, is Hornsby the best position-adjusted hitter?

I don't think we have the data needed to do positional adjustments for players before the '70s. At least, I've never seen it. I bet Ruth, even as an outfielder, still beats him, though.

Ruth as a DH would beat them all.

Does anyone have the stats on PA outcomes after a first pitch ball and a first pitch strike? It is very different, IRC. I've always thought the folks saying never to swing at first pitches have it only half right. Swinging at a first pitch ball is a really bad idea but taking a first pitch strike is also bad (depending, of course, on what the pitch is like).

Also, if you're a pitcher and you know that Jason Kendall swings at the first pitch only 3% of the time, shouldn't you make very sure you throw a strike? Does anyone know how common it is for pitchers to have such data?

If I were a pitcher, I would strive to throw a strike on the first pitch every time. If the batter swings and hits it, he might well end up with a very good batting average -- .300, perhaps -- but guess what, he's also got a .300 OBP, and I've got a 70% chance of getting an out with one pitch.

If he swings and misses, or swings and fouls it off, I've got an 0-1 count to work with. And if he takes it, I've also got an 0-1 count to work with.

Personally I don't care about BB/PA, P/PA, or Ks taken or whatever. I want batters who will swing hard at pitches they can handle and lay off the ones they can't (within reason).

Does it make sense to break it down into these components:

1) BABIP (including HRs) and SLGIP -- these tell us how much the batter punishes the ball when he makes contact.

2) Ks -- a high K total means the batter either has few pitches he can handle (lots of called strikes and fouls) and/or he's swinging at a lot of pitches he can't hit at all.

3) BBs -- he at least has enough sense to lay off balls.

Maybe 2 and 3 go together in K/BB ratio.

Those are pretty much what I think of as the three components of plate discipline. I do not want to see batters taking pitches they can handle. I think the value of a lead-off man taking lots of pitches so everybody can see what the pitcher's throwing are overblown. I think the short-term value of tiring out a starter is much lower than many people think....plus I think you'll tire him out more by smacking a pitch somewhere, forcing him to face another batter, than letting a good hittable pitch go by.

I remember a couple years ago, someone here looked at, essentially, SLGcontact or maybe ISOcontact -- what did the player do when they made contact with the ball (i.e. resulted in a BIP including HRs). If memory serves, Jim Thome won that contest.

Anyway, we can think of prototypes of hitters using the above three criteria:

Barry Bonds: kills the ball when he makes contact; apparently handles most strikes and doesn't swing at pitches outside the strike zone. That leads to high BA/SLG, low Ks, high BBs. Or: crushes the ball when he swings, goes down to 1B when he doesn't.

Jim Thome/Adam Dunn: kills the ball when he makes contact; sacrifices ability to make contact to achieve that, but doesn't swing at balls outside the zone. That leads to good BA, great ISO, high Ks, high BBs.

Vlad Guerrero: kills the ball when he makes contact, though probably less than people think.* Is clearly able to handle most any strike (and some balls) so doesn't need to sacrifice ability to make contact to achieve that. Since he's almost guaranteed to see a pitch he can handle in a PA: good BA and ISO, low/medium Ks, low/medium BBs.

Alfonso Soriano (when he was good): kills the ball when he makes contact (less than the above guys obviously); sacrifices ability to make contact to achieve that and/or swings at balls outside the zone. Good BA, good ISO, high Ks, low BBs.

Ichiro: doesn't kill the ball when he makes contact, though hits for decent average.** But can "handle" almost any strike (and some balls). High BA, low K, low BB.

I don't know which of those guys is most "patient" (we're pretty sure it's not Ichiro), but I don't really care. I do know that Bonds (or Sheffield in his prime) are the guys I want on my team. It's a coin flip between the Dunn/Thome and Vlad types for second place.

* If we look at contact only (2004 BA/SLG/OPS):

Vlad (538 contacts): 424/753/1177
Dunn (373 contacts): 405/866/1271
Thome(364 contacts): 382/810/1192
Thome 2001 (for fun) 449/962/1411

Note that 1411 is still lower than Bonds overall OPS last year! But Bonds 2004 wasn't all that strong on this measure: 407/913/1320

Also for kicks:

Jose Hernandez 2004: 407/760/1167
Jose Hernandez 2002: 401/745/1146
Jose Hernandez career: 365/610/975

** Same basic principle. Since Ichiro puts the ball in play so much, even last year his contact BA was 409, which is obviously good but not as far out as his 372 BA overall. For his career his BAcontact is "just" 373. That's not much better than Jose Hernandez.

So back to Jose Hernandez. Here's a neat one:

365/610/975 Hernandez career
358/630/988 Soriano career

Yet there's an 80 point difference in career OPS. Amazingly enough, the difference is that Soriano has better K rates (per PA):

.276 Hernandez
.196 Soriano

Hernandez actually has a higher BB rate which makes up some of the difference. (there are lots of other differences too)

But here are the ones that will blow everybody away:

365/610/975 Jose Hernandez career
339/600/939 Gary Sheffield career
372/675/1047 Vlad Guerrero career
372/678/1050 Jeff Bagwell career
370/680/1050 Frank Thomas career
356/724/1080 Barry Bonds career

So next time someone wonders what the value of controlling the strike zone is, tell them that it's all of the difference between Jose Hernandez and Gary Sheffield and most of the difference between Hernandez and Guerrero, Bagwell, Thomas, and Bonds.

I think the short-term value of tiring out a starter is much lower than many people think.

I think so too. Or at least benefit of wearing out an opposing team's pitching staff is reaped by the entire league, not just your team.

Isn't the Bagwell, Thomas thing kind of creepy. Born on the same day, very similar stats, had their best year in the strike year, most similar to each other in BB-ref. Here once again exact same numbers on OPS on balls in play (including HR). eerie.

---"Personally I don't care about BB/PA, P/PA, or Ks taken or whatever. I want batters who will swing hard at pitches they can handle and lay off the ones they can't (within reason)."

Fine, but where are the stats to evaluate that? That is similar to the definition of discipline that I proposed a few posts ago. But I realize that such discipline is simply one component that goes into the end products of K/PA, BB/PA, HR/PA, etc. If you succeed at identifying all of the other "sub-abilities" which go into, say, BB/PA, and combine them in an appropriate fashion, you will end up with.....BB/PA.

I know all about how much each outcome, such as BB, regresses to the mean, given the # of PAs, etc. So, if you include that, I don't see why we should get all gaga over non-outcome "abilities" such as discipline, or contact, etc. Such emphasis is usually emphasized by people who are trying to "sell" their fantasy baseball evaluations.

Fine, but where are the stats to evaluate that?

Walt defined them - BABIP and SLGBIP.

That's an aspect to being "patient" at the plate. The goal to patience is to get a good pitch to hit. As Bonds has the highest average and the highest SLG, then I say he hits *the right* pitch. A batter shouldn't be rewarded (in the sense that he is rated "more patient") because he mishits balls he meant to hit.

Repoz, very nice. Can you make me a copy of that?

My wife actually helped me when I was struggling at the plate. She said "Just because it's a strike doesn't mean you have to swing at it."

Good tip - wait for a strike you can drive, don't just protect the zone.

---"Fine, but where are the stats to evaluate that?
Walt defined them - BABIP and SLGBIP."

Well, then let Walt demonstrate to us that evaluation by BABIP and SLGBIP produce better projections than evaluations by BB/PA, etc.

Furthermore, there is no reason to believe that BABIP really reflects a batter's ability to "swing hard at pitches he can handle and lay off pitches he can't". It might simply reflect the batter's natural pop.

Look at a pitch which is either 1 inch into, or out of, the strike zone. With a difference of only 2 inches, a batter who can "handle" one can probably handle the other--but the difference is that if he takes both pitches, there will be a significant difference in the count (assuming, of course, accurate umpiring), with a concommitment difference in the linear weights of that PA going forward. So, to me there is a difference between a pitch a certain batter can handle which is in the K zone, and which is not in the K zone.

Bottom line--baseball wins are dependent on outcomes, not on skills. We know what the outcomes are, but we don't really don't what all the skills are. We know how to weight and regress the outcomes. We don't really know how to define and regress the underlying skills.

So, spare me all of the fawning over Joe Blow's "patience". AFAIK, it does not translate into better future performance than an "undisciplined" hitter of similar overall production--given an adequate sample of PAs, of course.

Actually, you got me, Alphonse. I think BB/PA and SLG tells me a players abilities.

I'm also surprised that Mark Bellhorn doesn't show up anywhere in here. Seems like he gets at least one ball an at bat, which would lead to a lot of 4.00 pitch counts what with the Ks and Walks.

Guess the balls he puts in play go into play pretty early.

See #12, Alex.

Am I missing something or will players who swing and miss a lot have, as a result, misleadingly high P/PA? Conversely, will batters who make contact a fair deal will end their AB's sooner and have misleadingly low P/PA?

Take this Josh Phelps AB: swinging strike, swinging strike, ball, called strike. On it's surface, the 4-pitch PA makes him "seem" patient, but he was trying to end the AB in both pitches #1 and #2, and who knows if those pitches were even in the strike zone.

Take this Barry Bonds AB: ball, ball, swing (ball in play). By pitch-count alone, he was less "patient" than Phelps (3 pitches to 4). However, he didn't try to end his AB until pitch #3, and then, quite probably, only on a pitch in the strike zone.

I'm wondering if guys like Dye and Blake, not big walkers, have misleadingly high P/PA counts because they swing and miss a lot, and as a result extend their PA's.

Well, I wasn't trying to answer anything and I wasn't trying to predict anything. I was just speculating, seeing if we could make some progress towards finding better ways to talk about and measure what we talk about and try to measure.

Actually of course I was just trying to amaze folks -- who here would have thought that Jose Herandez had better results on BIP than Sheffield? Who would have thought he could rival Bonds (with a higher BABIP)? It doesn't always work, but counter-intuitive results often lead us to a better understanding of underlying processes.

I will add that there's a difference between prediction and explanation. Actuaries don't care why certain things increase your risk of dying, they just care that they are predictive because all they care about is reducing their financial risk.

Doctors of course do care about why certain things increase the risk and so try to explain them. Explanation is, of course, largely an academic exercise, but I'm an academic.

What good could my speculation serve? Well none if it leads nowhere which is the most likely outcome.

Here's what I thought was interesting in my little Jose Hernandez exploration -- there's not a huge amount of variation among these guys in terms of what they do with BIP. Barry Bonds has a 300 point edge in career OPS but only a 100 point edge in career OPSBIP. Sheffield has a 200 point edge in OPS; Guerrero has a 200 point edge in OPS.

It's speculative but it suggests that controlling the strike zone is more important than even us statheads had thought. It also suggests that there are different ways of controlling the strike zone (Vlad vs. the others) that produce pretty similar results (Barry excepted). It suggests that teaching young players to control the strike zone may be even more important than we thought.

Then there's Thome, who's at 404/810/1214 for his career, beating all those folks handily. Unlike Hernandez, he also walks a lot, but still his Ks bring him back to that impressive pack.

For kicks, here's Bellhorn: 363/620/983.

Or maybe it's just me. I always thought of Sheffield as one of the best hitters in the game. Not thinking about the walks and only some about the Ks, here's a guy who ropes line drives all over the field. I think he's got the quickest bat I've ever seen (pre Barry 2001).

But he doesn't outhit Jose Hernandez or Mark Bellhorn when they make contact. He and Bellhorn have the same walk rate too. What differentiates him from Bellhorn is his contact rate -- i.e. he apparently can handle (or chooses to hit) more strikes than Bellhorn.

I wonder if it has lessons for us on age effects. Why don't young players with old age skills age well? Does their OPSBIP drop or is it their contact rate? Do players like Vlad maintain their OPSBIP and contact rate?

I doubt anything here is revelatory. It's just a different way of parsing up the same numbers. But again when I see Hernandez matching Sheffield, I take notice. So maybe parsing up those same numbers will help us see something we haven't seen before.

It's a little like DIPS if you will. Not in DIPS finding about pitchers' BABIP, but in that it challenges the notion that high-K pitchers and keeping the ball out of play are the best option. What DIPS says is that it's a pitcher's combination of K, BB, and HR (and indirectly the BIP rate) that have the biggest impact on his performance ... but there are lots of different ways to get there.

The numbers in this post, in comparing say Sheff, Guerrero and Thome suggest that good hitting is a combination of OPSBIP, contact rate, and K/BB ratio. No, that's no different really than BA/OBP/SLG ... anymore than K/BB/HR is really any different than WHIP and ERA (or CERA at least).