Games: Minnesota Twins, April 29: Judging the Royals with Lee Judge

Games » Minnesota Twins

Apr29

Freshly squeezed

Pitchers use three weapons to attack hitters: location, movement and velocity. Larry Vanover, the home-plate umpire for Sunday’s game, took one of those weapons — location — away from Royals starter Bruce Chen.

One of the theories about Chen is that he needs to have a loose strike zone to succeed. Vanover had a very tight zone and forced Bruce into the middle of that zone by calling borderline pitches balls. But give Vanover credit. He was consistent. He missed calls when Nate Adcock was pitching, too.

Actually, I saw Vanover’s tight zone help the Royals on occasion, Brayan Pena probably should have have been called out on strikes in the ninth inning a couple pitches before he was, but a tight zone probably hurts Chen more than other pitchers.

Speaking of Nate Adcock, what he did was a big deal. Nate saved the bullpen by throwing 5 and a third innings in relief. That might make difference in the upcoming series against the Tigers.

Game notes

In the bottom of the second, Alex Gordon made another great diving catch. Last week, I asked Alex the key to making that kind of catch. He laughed and said, “A bad jump.” Actually, Alex said he tries to catch the ball off to the side, which allows him to watch the ball into the glove.

In the top of the third, Jarrod Dyson drove the ball to the warning track, but hitting the ball in the air is not what Dyson wants to do. He needs to hit the ball on the line or lower.

In this game, Dyson had trouble at the wall. He missed what looked to be catchable ball (easy for me to say) when he went up against the wall and missed it. Some outfielders can block out the pending collision with the wall, some can’t. I haven’t talked to Jarrod and haven’t seen him play the wall enough to know whether Sunday’s play was an aberration.

The Royals have been trying to jump-start their offense by giving hitters the green light to hit on a 3-0 count. That hasn’t worked at times, but it did Sunday in the fourth inning. Moustakas got a 3-0 green light, banged a ball into center field and got a hit and an RBI out of the deal.

Jeff Francoeur got thrown out at third base on the play, but there was one out, and that’s the time you push the envelope on getting to third. If you want to fault Frenchy for being overaggressive, do it over the fact that Twins center fielder Denard Span was left-handed and didn’t have to make a pivot to throw the ball, but it took a perfect throw to get Francoeur out, and most of the time that bet pays off.

Mike Moustakas made an error in the fifth inning while transferring the ball from his glove hand to his throwing hand. Moose sometimes has the tendency to flip the ball into his throwing hand, and that can backfire.

In the top of the sixth on a 3-1 count to Billy Butler, Twins catcher Joe Mauer tapped his glove on the ground, signaling for a ball down. The pitch was up, and Billy singled. As some of you have pointed out, Brayan Pena he doesn’t like to tap his glove on the ground as a signal to miss down. Other catchers use that signal.

What have you done for me lately?

Apparently, there now now Internet campaigns dedicated to getting Angels manager Mike Scoiscia and Twins manager Ron Gardenhire fired. These guys are considered two of the best managers in the game, but because of bad starts, some fans want them sacrificed.

But that’s nothing. Royals manager Ned Yost told me that when he coached for Atlanta every time the Braves went through a bad stretch, fans would call for the firing of Bobby Cox. A future Hall of Fame manager.

Firing Scioscia or Gardenhire would not make things better—they would make things very much worse, but as former Mets general manager Steve Phillips said, sometimes you fire the manager to “give the fans hope.” If a team has a problem and the responsible individual can be identified, by all means make a change. But firing someone who isn’t at fault—just so fans who don’t know much about baseball—can feel better doesn’t make much sense to me.

Here’s a fan who hopes baseball teams would be smarter than that.

A reader asked me a question; here’s the incredibly lengthy response

Part One: Question and Answer

Do I have any statistical evidence to back up my assertion that a fast runner on first base helps the batter at the plate? The reader wanted to know whether the Royals’ quantitative-research analysts had given me any information on the subject. The answer is no, but if they have that information, I don’t think they’d share it. Why let other teams know whether something is or isn’t working?

I could say that players, coaches and managers believe a base-stealer helps the hitter, but that is like saying everyone knows the world is flat — in 1491. I wouldn’t expect anyone who is into metrics to be overly impressed by that. I hope everyone who is into metrics returns the favor and does not expect me to be overly impressed by news of a study that “proves” a base-stealer does not help the hitter.

One of the studies the reader cited was from “Baseball Between the Numbers,” and the author was Jonah Keri. I have that book on my shelf, so I read the two articles by Jonah Keri, “What’s the Matter with RBI? … and Other Traditional Statistics” and “Is Wayne Huizenga a Genius?” Neither article had anything to do with base-stealing. Maybe there is an article in the 434-page book that deals with the subject, but if so, I didn’t find it Thursday night.

My skepticism about studies that “prove” something started early. When this website first came online, I was attacked for saying that Rick Ankiel could really go get a baseball. That was my opinion after watching Ankiel consistently start from shallow center field, get great jumps and smoothly catch balls on the warning track.

An outraged reader said the “range factor” numbers proved Mitch Maier was better. A little research proved that range factor (putouts and assists divided by the number of innings played) did not account for the fact that Ankiel had been playing center field for the Cardinals, a team with much better pitching than the Royals, and did not have the same opportunities as Maier.

I’ve learned to be skeptical of defensive metrics that don’t factor in where a fielder starts. Or systems that allow an observer to “eyeball” the type of ball put in play and the sector it wound up in. Or a study that proved the Royals were playing Eric Hosmer out of position, but apparently—it’s what I was told—didn’t factor in the “no-doubles” defense that had him positioned on the line. Or another study, concerning relievers, that didn’t factor in the human element and reactions to pressure.

All of these metrics tell you something, but they don’t tell you everything. In one way or another, all these metrics leave something out. (Of course, you can say the same thing about the “eyeball” test. So it’s good to remember that both approaches tell you something, but not everything.)

But back to the question: If I have no statistics to back up my belief that a base-stealer helps the hitter, why believe it?

Let’s start with what we know. We know it takes longer to deliver a breaking pitch to home plate than a fastball. I timed a pitcher the other night and he was getting his fastball to the plate in 1.2 seconds (you can’t run on that), and he got his off-speed stuff to the plate in just under 1.5 seconds (you can run on that).

We know that teams look for breaking-ball counts to run in. It seems logical to conclude that the pitcher will be encouraged to throw more fastballs. Many catchers have been accused of calling too many fastballs when a fast runner is on base. When Yogi Berra was brought into work with then-Yankees catcher Jose Posada, that was one of his first criticisms. Posada called for too many fastballs in an effort to throw out runners.

We know that a pitcher who uses a slide step tends to lose velocity and movement (not every pitcher, but most of them). And we know the slide step can cause the ball to stay up in the zone. I was told that Royals pitcher Luke Hochevar had trouble throwing strikes out of a slide step and would often fall behind in the count. We know that’s not good.

I haven’t counted how many big hits I’ve seen since I began paying attention to a pitcher’s delivery times and compared them with the hitter’s average when the pitcher isn’t in a slide step, but I’ve seen a lot of big hits when the pitcher is trying to be quick to the plate.

We know that varying rhythm and release points is difficult for a pitcher. Royals first-base coach Doug Sisson urged me to watch the quality of the pitch after three consecutive pickoff attempts. Once again, I haven’t counted them, but I’ve seen an awful lot of bad pitches in that situation.

I could go on (I think I already have) and talk about defensive positioning and catchers rushing throws, which screws up their framing, but if you’re not convinced, that’s OK. Let’s leave it at this. I — and a lot of ballplayers — believe that a base-stealer helps the hitter. While I haven’t conducted a study, the evidence I have seen supports this belief. Until I see a study that makes sense to me and disproves that notion, I will go on believing it. Feel free to disagree.

Part Two: Subjectivity

I had my math buddy (I try to have a friend in every area of expertise so I don’t have to know anything myself) Mike Keefe look at some advanced metrics and tell me what he thought. Mike is a political cartoonist and won the Pulitzer Prize last year (which I’m kinda P.O.’d about). He also has a math background (a dissertation short of a doctorate) and taught statistics and probability at the college level.

What Mike said was interesting. The models he looked at resulted in a scientific-looking number, but he cautioned me that there still was subjectivity involved. Even math requires variables. Deciding what variables are included is subjective. Deciding how to weight those variables is also subjective. The more variables a model includes, the more chance there is for inaccuracy.

Mike said he would trust simple models — batting average, on-base percentage — more than complicated models. And to truly understand the complicated models, it helps to have a background in statistical correlation and regression analysis. (I don’t, and I suspect that a great many people who quote these studies don’t either. And if you’re quoting studies you don’t understand, then you’re doing the same thing I am: trusting the expertise of others.)

It seems clear that when I talk to the people who do have educational backgrounds in this area, looking at one number and deciding a player’s worth — or lack of it — is a mistake. It’s all part of a larger picture … and that picture is, to some degree, subjective.

Part Three: There are only so many hours in a day

The reader who asked the question seems to be a nice guy who is genuinely interested in the answer. He does not seem dogmatic or out to score points to prove his intellectual superiority. I think he, and many other readers, are truly interested in understanding baseball, and many of them do so through metrics.

But I don’t always have time to do this.

I can write short responses to specific questions, but writing term papers in response to statistical studies is awfully time-consuming. When the Royals are in town, I’m putting in 18-hour days on a regular basis. When this site started, I tried to give an individual response to each reader, but the site is growing (thanks for that) and it’s proving impossible to keep up that practice.

There are individuals who want to conduct lengthy arguments over things they read here. In one sense, that’s fair. I’ve made a statement, shouldn’t I back it up? In another sense, it is like going to Oklahoma Joe’s for lunch every day (which actually sounds like an excellent idea) and wanting to have a lengthy argument about why they’re not serving Chinese food.

It ain’t what they do.

Some readers have expressed frustration that I do not see baseball from their points of view and want to know why sabermetrics can’t be part of this site. What would be the point? There are plenty of sites that already take that point of view. There is no shortage of blogs that think Royals general manager Dayton Moore is an idiot and Yuniesky Betancourt ought to be playing in a softball league. From the very beginning, the point of this site was to bring fans the player’s points of view.

If what the players and coaches have to say interests you, great. I can help with that. But doing so takes up most of my day. I spend hours on the field trying to develop an understanding of the philosophies and strategies that make ballplayers, coaches and managers do what they do. Then I try to dissect the game using what I have learned from the participants.

I don’t fault anyone for seeing things from another point of view, and I enjoy the fact that what I have written often starts a debate among the readers. Please keep doing that. Feel free to call me a moron (it wouldn’t be the first time), but it’s become impossible for me to respond to every comment or engage in lengthy exchanges over statistical studies. I’m not saying you shouldn’t have that debate. I’m saying I don’t always have time to join in.

And I realize what I just said about not always having time to engage in lengthy debates probably will set off a lengthy debate. Good luck with that.

Comments

Thayne Griffin
1 year ago

How about how Moose’s HR today? Before he came to bat, Marquis had only thrown four or five pitches (I don’t remember which) to get the first two outs of the inning. It looked like Moose was set on taking a few pitches to try and help out the team, and he ended up getting a solo HR. Great was to reward him for trying to put the team first.
Jim Fetterolf
1 year ago

Moose is quietly putting together a very professional, high-quality year; good bat, good glove, has his head in the game, and is showing veteran level maturity. Fangraphs has him at 1.1 fWAR with a lot of love for his defense. Interestingly, fangraphs actually rates Moose higher than Lee’s Polk system, at least as of last night.

Good catch on the pitches, Thayne.
Randy Westfahl
1 year ago

I use the old system where I watch them and then I give the the RW of good or bad. Right now Moustakis is very good. It is a simple method and much easier to understand :)).
Jim Fetterolf
1 year ago

The RW is probably as good as the LJ or JF. What so-called advanced metrics are trying to do is get a set of numbers to “see” closer to the same thing that a trained, experienced eye does. Most Hot Stove disagreements with metrics comes from metrics still unable to quantify an outstanding or bone-headed play.
Jim Fetterolf
1 year ago

For those interested in the charts, this is Dan Brooks excellent site’s effort on today’s Bruce Chen outing.

http://www.brooksbaseball.net/pfxVB/pfx.php?month=04&day=29&year=2012&game=gid20120429kcamlbminmlb1/&prevGame=gid20120429kcamlbminmlb1/&prevDate=0429&pitchSel=136600.xml

Third graph down gives Chen’s pitch plot. For comparison, this is his previous start against Toronto. He got a fatter strike zone that game, third graph down:

http://www.brooksbaseball.net/pfxVB/pfx.php?month=04&day=23&year=2012&game=gid20120423tormlbkcamlb1/&prevGame=gid20120423tormlbkcamlb1/&prevDate=0423&pitchSel=136600.xml

PitchFX is a great tool.
Joel Kallem
1 year ago

Lee, your lengthy defense of your approach is unnecessary. Those of us who came to this site, came for exactly the main reason you cite - it gives us insight into the minds of the players and coaches as they play the game. The grid is interesting and sometimes sparks some conversation, but the “gems” you share are the real meat we come for. I’m glad we are not a metrics emphasized site. Having worked with numbers all my life, I would share the old saw that figures can lie. Being a good CPA might mean you can come up with all the “right” numbers, but doesn’t necessarily mean you can make the right decisions by interpreting those numbers in a way to maximize potential whether you are talking about business or a ball player.
Scott E. Hassler
1 year ago

I have a dream, that one day, Sabrmetricians and traditionalists, shall one day live in harmony, united by a common passion for the American past time. There are few absolutes. Ttruth is rarely white or black, but generally found somewhere in the shades of grey. I like metrics, I also like other approaches. The horror.
Jim Fetterolf
1 year ago

Joel, Scott, agree with you both. One thing that makes this site and community unique is the opportunity to discuss and integrate all aspects of the game, we can be above rigid ideologies and find what works to help us enjoy and understand the game and seem to be able to do it in a civilized and thoughtful way. AM radio this isn’t, thank G*d:)

Have a great week guys.
Sean Fischbach
1 year ago

Lee;

What your saying is one answer to strategy in baseball is not always the right one?

WOW, Who knew!!! ;)
Gaines Arnold
1 year ago

I disagree that this is in any way a stats friendly site, and BTW your second point from your supposedly ABD-math expert friend is actually wrong Lee. But, I won’t go into how the the vaguaries of variables are made miniscule because that would require a dissertation and is not appropriate for this site.

The reason I come back here, personally, is the same reason I watch news reports on both CNN and Fox News. I have a conservative point of view, so I agree, mostly, with Fox News reporting, but, because they are slightly biased (I say that tongue-in-cheek), I also watch other points of view (BTW I have seen some of your political cartoons, and lets just say I enjoy the baseball stuff). All that is to say that you offer something different from the stats sites which is a good thing.

Although I enjoy the stats-oriented stuff more, I like to temper that with a little fluff from the minds of those who “see” the game more than statistically analyze it. And that is what this site is about for me.
Lee Judge
1 year ago

Well, first off, thanks to anyone who comes to this site for any reason. I hope it supplies a point of view that’s helpful in understanding baseball.

It’s certainly not the only point of view, but, at least to me, it’s a point of view that wasn’t often heard by the public. It’s an attempt to bring the dugout and clubhouse conversations to fans.

I don’t know if what I wrote was a defense of what I’m doing as much as an explanation: it’s becoming impossible for me to give individual responses and conduct lengthy debates—at least on a regular basis.

Some days I’ve got time, but increasingly, some days I don’t. That’s a direct result of more readers and more comments—and I’m happy to have both.

I just didn’t want to alter the way I’ve been doing things without explaining the change. I still check in every day—mulitiple times—read everything and respond when I have time. (I’m waiting on an editor right now.)

I think I’ve softened somewhat on my view of metrics since starting this thing: I’ve come to believe you’ve got to sort through the information the numbers give you and do the same with the information the players give you. I can ask three players the same question and get three different answers.

I think the more you learn the less sure you are that there is one right answer.

OK, back to my day job.
Mark Harkins
1 year ago

Lee, As an out-of-town fan who rarely gets his eyeballs on anything Royals other than the highlights on MLB.com, your site is the first place I go after checking the Box score. OK, so Chen had a bad outing. The Star reports (for all I like about what they do) are not going to say the strike zone shrunk. You provide great nuggets of info like that about what really happened that you can’t get anywhere else. Then you go in-depth on an issue (like throws home from right field) that make me think and provide more insight into the game. Great stuff.

Keep doing what you’re doing man… and thanks!
Nathan Coltrane
1 year ago

Rex was hammering the HP umpire on squeezing Chen on the strike zone. Your comments are right on the noze to the guys in the TV booth. It was a good analysis that makes a viewer and reader see what is behind just the balls and strikes.
Lee Judge
1 year ago

BTW: The website is currently convinced that Nate Adcock (the new #47) is Victor Marte (the old# 47). We have someone in IT arguing with the website now.
Lee Judge
1 year ago

Mark and Nathan: Thanks. So much is left out of a boxscore. It says Chen had a bad day, but not much about why.

The more these guys tell me, the more interesting the game becomes.
Shane Lorimer
1 year ago

I use mlb.tv (not premium) so I’m stuck with the home team announcers, and even they commented on the tight strike zone Chen was getting. They made some jokes about the Ump being pissed that he was throwing to first so often, and they had a fly away day on Sunday too. This may well be true, but I don’t think anyone is going to come up with an advanced statistical study to prove UmpGrump’s impact on pitcher performance.

Statistics are like a bikini. What they reveal is suggestive, but what they conceal is vital. - Aaron Levenstein
Jared Dull
1 year ago

I’m wondering if the ending to Friday night’s game had anything to do with the tight strike zone. I would imagine Gardenhire had a little something to say to the crew about his two superstars getting punched out on borderline calls to end the game, especially being at Target Field. I know umpires and referees in all sports aren’t supposed to try to even things out with “makeup” calls, but they are human. I’m sure it happens whether they are conscious of it or not.
Jim Fetterolf
1 year ago

“I disagree that this is in any way a stats friendly site”

This site is quite stat friendly, being somewhat based on the Polk system, it’s just not a sabremetric oriented site. Lee attempts to quantify elements of a game that other systems don’t.

I did a brief comparison the other night between fWAR and LeeWAR for some of the players and it was a little closer than I expected. After Friday’s game Moose was at f1.0 and L 0.84. Alex had f0.9, L 0.83, Esky f0.9, L1.04, Billy f0.6, L1.04, as some examples. The biggest disagreement was on Hosmer, f0.2, L 0.95. My methodology is to move the decimal point two spaces to the left on total Grit points to derive LWAR.

Given the similarities of values, I then tried to identify why they differed and offer a couple of reasons; outstanding plays and mental mistakes. Polk also docks for batters striking out and striking out looking, but rewards for RBIs, long at-bats, sacs,and runs scored.

In conclusion, the major differences among fangraphs, BR, and Lee are on what should be weighted and how much.
Lee Judge
1 year ago

Jim: You’re starting to scare me. And, Shane, someone who should know mentioned the same thing: umpires don’t appreciate 10 pick off throws on getaway day.
Brendan Woodbury
1 year ago

Lee -

I appreciate the response. Obviously, I’d hoped for more information, but I understand that the Royals might not want to give it out, and I appreciate you taking the time to find out what you could and write it all out.

I’m a little confused by your repeated use of the word “prove” in the discussion above. I don’t think I ever used it in my comments raising questions about the aggressiveness on the bases, and I understand that outside of controlled experiments, proof is elusive.

I was just looking for evidence. I was frustrated in how many outs the Royals were running into on the bases and I was trying to understand why they were doing that.

My instinct, like yours, was that aggressive base stealing probably helps the batter some. Originally, when I started looking, I was trying to figure out if the help to the batter was outweighed by the harm caused by all the extra outs.

However, every study I’ve found says my instinct is wrong. Although there are individual cases where a distracted pitcher grooves a fastball, in aggregate, batters don’t do any better with dangerous base stealers on base. The authors have looked at the data and been unable to find a difference.

I think the easiest study to understand on this topic is here: http://www.hardballtimes.com/main/article/base-stealer-intangibles-part-1/

I am glad that the Royals have good guys working the numbers and looking for an edge, and I hope that their work pays dividends, but I’m not going to just assume that they know what they’re doing until they start winning more games. Frankly, the beginning of this season (where we’re not putting up nearly as many runs or wins as you’d expect given our hitting) gives me less confidence, not more.

I know I don’t know more than the Royals about everything in baseball. I may not even know more about this anything, including this. But the franchise has shown on the field that it doesn’t know as much as the teams it’s competing against, and I hope it’s open to outside studies like the ones I’ve found that suggest that trading outs for bases is not often an effective strategy, even if it appeals to our instincts.
Greg Tatro
1 year ago

“Pitchers use three weapons to attack hitters: location, movement and velocity.”

And that brings to me to the curious case of Bartolo Colon. How has he been successful so far this year? Certainly Oakland as a pitcher’s part helps, but…

All he does is throw strikes…he threw 38 straight against the Angels a few starts ago. He has only walked 6 so far in under 43 innings. I saw a stat that 83% of his pitches are fastballs. I think the point is the pitchers who try to be too cute with hitters usually get into trouble.

Pitcher’s biggest weapons are throwing strikes and using the defense behind them.
Brendan Woodbury
1 year ago

Lee -

I don’t question your buddy’s knowledge of math, but the idea that batting average is a simple stat is too much for me to swallow.

Batting average is a stat that divides hits (times where the batter reaches base after putting the ball in play, except when the oficial scorer judges that the batter would have been out under one of two alternate circumstances [fielder’s choice or defensive error])) by at-bats (plate appearances not resulting in a sacrifice [which is also a judgment call by the official scorer], catcher’s interference, base on balls or hit by pitch).

That’s the simple statistic we should stick to?

I completely understand people who say that they’ve been working with batting average long enough that it feels comfortable to them, and this is a game after all, so who wants to make it a bunch of work? That’s an excellent reason for trusting batting average and ignoring newer statistics.

I can not understand the theory that newer hitting statistics, like wOBA, are more complicated and therefore less trustworthy than batting average.
Jim Kissane
1 year ago

To the stat guys trying to figure out if a fast runner on first helps or hurts the batter: Good luck. As Lee points out, statistical modeling is an applied science, not an exact one. Too many variables can put the result beyond the reach of the data, at least in any conclusive, objective way.

The Royals stole more bases last year, but they also hit better than they have so far this season. At its simplest, better hitting is going to produce more runs as well as the opportunity to steal more bases. I don’t know that you can necessarily correlate the two because team speed is utilized in different ways - either actively to produce runs or passively to avoid outs. There is lots of gray in there, too - for example a hit-and-run can produce a stolen base, even if that wasn’t the original intent.

Beyond the raw data, the type of hitter comes into play. I grew up watching Lou Brock steal bases, and for much of his career he had two hitters ideally suited for batting in the 2-hole behind him - Curt Flood (who should be in the HOF, but that’s another post), and Ted Sizemore (who adapted to the role after Flood was traded). Both were outstanding contact hitters who could hit to all fields and who could foul off pitches to protect Brock. The 1-2 combination (and Brock would be the first to point out that it took more than just him) could demoralize a team in the first inning.

The presence of a base-stealing threat (a better term than simply a fast runner, since there ae plenty of fast runners who are not base-stealing threats, especially in the American League) is going to affect the hitter, but whether the effect is positive or negative is going to depend very much on the hitter. Patient hitters have a better time coping with the pitcher’s distraction by the baserunner. The throws over to first and the long intervals of holding the ball can drive impatient or immature hitters completely nuts.

Pitch selection also becomes skewed. On a team with an active running game, a batter may have to swing at a bad pitch either as part of a hit-and-run or simply to protect a runner who got a bad jump. On the plus side, however, batters will generally see more straight pitches because they take less time to reach the plate and they’re easier for the catcher to handle (and a distracted pitcher throwing fastballs is never a bad thing when you’re batting). A contact hitter at the plate can help a base-stealer as much as the base-stealer helps him; a guy who strikes out a lot is probably going to strike out more and thus is going to limit the running game.

I think team speed can be a deadly weapon, because a leadoff walk can become a two-base hit, and a right-side grounder puts a runner on third, and all of those scenarios can present advantages to hitters, or at least put the pitcher and the defense on edge. Still, that’s a synergy and I don’t know that sabermetricians have the tools to quantify that. Also, whether a team’s intent is to advance baserunners or to avoid outs, success in those opportunities is going to be hard to credit solely to either the runner or the batter. But my impression is that it helps the team in general, even if I don’t have the data to back it up.

Whether there is a cause-effect relationship or whether it’s simply a matter of using your resources to your best advantage, it’s somewhat like situational ethics. When it works, you look like a genius; when it doesn’t, you look like you’re depriving a village somewhere of a perfectly good idiot.
Jim Fetterolf
1 year ago

For those interested in wOBA, a useful stat:

“Weighted On-Base Average combines all the different aspects of hitting into one metric, weighting each of them in proportion to their actual run value. While batting average, on-base percentage, and slugging percentage fall short in accuracy and scope, wOBA measures and captures offensive value more accurately and comprehensively.

The wOBA formula for the 2011 season was:

wOBA = (0.69×uBB + 0.72×HBP + 0.89×1B + 1.26×2B + 1.60×3B + 2.08×HR + 0.25×SB -0.50×CS) / PA

These weights change on a yearly basis,”

http://www.fangraphs.com/library/index.php/offense/woba/
Brendan Woodbury
1 year ago

Jim K -

You raise a bunch of good theoretical concerns about study design, but I can’t tell that they’re specific to the actual studies I linked to, which I think address all of them. I very much encourage you to read the study I linked to above: “Base Stealer Intangbiles (Part 1)” on the Hardball Times website.

With regard to Curt Flood, a hitter you believe is ideally suited to benefiting from a base stealing threat on first, I looked up his career splits on baseball reference.

Flood was a very good hitter. His career stats were .293 AVG; .342 OBP. Like (almost) all hitters, he was significantly better with runners on base. His batting average jumped 14 points to .307 and his on-base percentage jumped 15 points to .357. Whether it’s distraction or difficulty pitching out of the stretch or limited pitch selection, pitchers struggle with men on base.

However the theory here is that there’s an extra advantage to a having a base stealing threat and Flood is well suited to take advantage of that distraction. If it’s true, we’d expect him to do a little better in a base stealing situation (runner on first, second and third open) than he does overall with men on base.

At least in Flood’s case, we do not see the results predicted by the theory. His batting average drops by 4 points and his on-base percentage drops by 13 points.

It may be that the defensive is put on edge, but that edge helps them. Or the batter gets distracted or something else. I don’t think statistics can answer those questions. But even if they can’t explain the mechanism, they should be able to give us strong answers about the results. The analysis I present above can be refined (e.g. isolating only the times Brock was on first and Flood was hitting 2nd), but the results I found make it unlikely that Flood’s results were better with a base stealing threat on first.
Lee Judge
1 year ago

Brendan: I cited a several cases where people told me a single study “proved” something—if you felt lumped in with that, my mistake.

As I said before, I think you’re open-minded and are just trying to figure things out. (Aren’t we all?)

My friend’s point about statistics had to do with variables. From a math point of view, in the end, batting average is pretty simple: there might be variables in what’s considered a hit and what’s considered an at-bat, but someone else works that out for you and you just have to divide one number by another.

I thought his point was well-taken about deciding which variable to use: what if somewhere along the way someone had decided hard-hit outs should be included or forcing an error through speed?

Selecting the variables is subjective. I think remembering that there’s a subjective element to everything we do—why study this element and not that one?—is good. It keeps us from being dogmatic.
Lee Judge
1 year ago

Brendan: I’ll read the study you suggested as soon as I get a chance.

Jim K: In my opinion your right, there are an incredible number of factors to consider in any given situation.

I don’t think that means we should throw our hands in the air and give up, but it does mean we ought to be slow to jump to an all-inclusive conclusion.
Lee Judge
1 year ago

Jim F: Hard to believe that stat hasn’t caught on with school kids.
Brendan Woodbury
1 year ago

Lee -

I’m totally lost on the complexity point now. If I understand you, you’re drawing a distinction between distinctions that are drawn by somebody else (the official scorer or the stringer categorizing the data) and distinctions drawn by the person making the model? What is the significance of that?

Even if there subjectiveness in everything we do, some things are more subjective than others, and we can and should discriminate on that basis. Runs allowed per nine innings is a less subjective stat than ERA. On-base percentage is a less subjective stat than batting average. wOBA is a less subjective stat than slugging percentage.

I agree with your friend’s basic idea that we need to be aware of potential sources of errors. They can add up as we increase the number of variables and the amount of subjectivity in each variable. However, I can not see how this favors a stat like batting average.

P.S. In the battle between old stats and new, I’m not sure the school kids are on the side you think they are. ;-)
Jim Wilson
1 year ago

Lee, I thought your description of what this site is all about was fantastic. I suggest you post it somewhere permanent on the site for easy reference so you don’t have to repeat it. I don’t think this site is very friendly to sabermetrics (Jim F., please don’t argue with this, it’s how people feel and it’s personal to them) but I’m starting to come around to your view that that is just fine. I can go other places for that and it is really easy (as your suggest) to ignore the Polk/Judge points system (which I do). I don’t let it upset me that you like it (and others may to) — there’s plenty of other good information here. Thank you very much for that!

Having said that, you can’t ignore numbers completely (and I know you know that) so . . .

I have to agree with Brendan on batting average. It is a “complicated” stat that does involve variables/subjectivity. It’s not just dividing one number by another. BA may seem simple because we’re used to seeing it but familiarity doesn’t make it good or simple. Your math friend may not have understood how BA is actually calculated. The other problem with BA is that it has a poor correlation to runs scored. That’s not surprising since it eliminates a number of situations in which a runner gets on base (avoids an out) and ultimately scores. OBP is a very simple statistic — times on base divided by plate appearances. And OBP correlates well to runs scored. As Jim F. has told us (and will again I’m sure), OBP is not perfect by any means but it is simple, meaningful and readily available in my box score.
Jim Fetterolf
1 year ago

“Runs allowed per nine innings is a less subjective stat than ERA. On-base percentage is a less subjective stat than batting average. wOBA is a less subjective stat than slugging percentage.”

I actually agree with this statement and prefer RA, as does Rany, and wOBA. OBP is not subjective but is heavily contextual, which is why I don’t care for it.

“I agree with your friend’s basic idea that we need to be aware of potential sources of errors.”

The error I see most often is historical stats used for predictive purposes without accounting for inflection points, “Chen’s a 4.48 pitcher” is a refrain I hear often. Next most common error I see is a probability, like wOBA, used for calculating a player’s value out of context. I like Runs and RBIs, so think runs scored and driven in, RSD, is a more useful statistical number within a team and like aSLG, TB+BB+HBP+net steal/PA to give self-created bases/PA, a useful number.

“P.S. In the battle between old stats and new”

The so-called “new” are rapidly aging in the face of PitchFX, FieldFX, BsR, and others, all attempts to get closer to quantifying what a trained and experienced eye can see. wOBA was created to address the inadequacies of SLG and OBP, BsR to get a more valid fielding number than the sophomoric UZR, itself an upgrade over the archaic fielding percentage.
Jim Wilson
1 year ago

On stolen bases: I have to believe that every team has a rule of thumb success rate that they want to achieve and while that number may or may not be based on statistical evidence, if it’s not it’s probably not that far off. I also have to believe that the Royals current rate of 64% is below (well below, I hope) the team’s expected rate. So in-game strategy and performance were not matching up with expectations. I was unhappy because our success rate implies that we were attempting steals that we should not have been. The approach appears to have been toned down some and, hopefully over time, we will continue to see the success rate improve.
Jim Fetterolf
1 year ago

JW, 75% is the number I most often hear for SB success rate needed. Royals haven’t been doing well this year, but part of that comes from opposing pitchers getting quicker to the plate, which creates other effects, as we saw with Luke’s slide-step last year.
Jim Wilson
1 year ago

Me too on the 75%. Unfortunately, because I think that’s low. But it depends on the team, the park, etc. You explanation of why our rate is low dosen’t help me — they have to take that into account.
Lee Judge
1 year ago

Got this from Curtis Ruder, he spells out the problem better than I can:

“http://www.pankin.com/sabr32.pdf

This is a study that I think is pretty good. It is a bit dated; I think it includes data from 1980-2001 in its study. The good news is that while it is dated, it is broad enough in different types of eras that I think it would likely stand up to scrutiny today.

Anyway, his conclusion is that the best base stealers provide a boost in batting average by 10-30 batting average points, 0-20 OBP points, and 30-60 slugging percentage points.

Roughly, that is a bit less than the equivalent of starting with a 1-0 count.

He also found that both the strikeout percentage and the walk percentage were higher when the best runners were on base, consistent with the following hitters taking some pitches to allow the runner opportunities to steal. I do not know his methodology in determining who the best base stealers are – in the powerpoint, he cites Rickey Henderson, Vince Coleman, and Willie Wilson in particular. Those guys clearly are elite, but that doesn’t tell me much about where he draws the line. Studies that divide all players into a couple of pots – guys who stole 20 or more bases, and guys who didn’t is one I have seen – have by and large shown no statistically significant improvement in batter’s performance. So this one is interesting in that it has a different result, but if it only applies to three or four guys at a time in the league, then it is of limited value. It is great to hit behind Jose Reyes, but maybe that’s about it. But if it is great to hit behind Alcides Escobar as well, then that is another story entirely.

There is a reason that the metrics people find this question particularly interesting, and that it is that the results generally run counter to the conventional wisdom on the question. For all the reasons you cited last night, batters should hit better with good runners on base. And yet, by and large, we have been unable to isolate an improvement on batting average, on base percentage, or slugging percentage.

To take another example, we all think batters do better when ahead in the count and worse when behind in the count. And the data match the expectations basically exactly. Outcomes on 3-0 counts are best for the batters, and outcomes on 0-2 are worst. The conventional wisdom matches the analytical results, which is good, but it is also boring. Imaging trying to present a paper entitled “Batters hit better when ahead in the count.” But by and large, even though we would expect hitters to see more fastballs, for pitchers to be less effective because of slide steps, and the like, we have never been able to isolate an advantage for hitters based on the quality of the runner. The advantage from having a runner on does not depend on the caliber of runner. The advantage is basically the same whether that guy is Butler or Escobar. And that is so counter-intuitive that it has become a passionate question.”

Brendan, I read the piece you suggested. So either professional ballplayers are wrong and there is no advantage or the studies are missing something.

I don’t think it’s impossible for ballplayers to be wrong: Frank White said when he played you weren’t up there to walk, you were up there to hit. Clearly, baseball’s attitude toward the walk was incorrect for a long time.

The difference here—at least for me—is that it only makes sense that the walk is valuable (if it’s bad for pitchers it’s got to be good for hitters), but a base stealer not improving the hitter’s odds—for the reasons mentioned above—seems counterintuitive.
Lee Judge
1 year ago

OK, game’s once again postponed. So I get another night off that I’ll have to pay for in the future.

But before I go:

I think Jim’s right about the desired level of stolen base success being 75%..that sounds familiar.

As I understand it, the problem has been guys pushing it in an effort to “make something happen” during the losing streak.

Wrong reaction: it should be the same well-thought out process every game. The runners are given “keys” (physical movements by the pitchers they “key” on—and apparently some runners have been taking off even though they’ve had trouble reading the keys.

So the question becomes is whether to blame the philosophy or the execution of the philosophy.
Jim Fetterolf
1 year ago

“You explanation of why our rate is low dosen’t help me — they have to take that into account.”

I’m sure they are, Doug Sisson is the ultimate stopwatch nerd. Just a matter of adjustments, but it has been suggested that even if the rate has fallen, a pitcher taking 1.2 seconds to the plate is going to distort mechanics and require more fastballs, which changes the dynamics for the hitter. Still early and there’s some new guys, I think it will even out in the low 70s by the end of the year. Not every aggressive out is a mistake, sometimes the other team just does everything on the play right.
Jim Fetterolf
1 year ago

For those interested in the prophets of sabremetrics and The Book, this is from Tom Tango’s site looking at Ron Polk’s system:

http://www.insidethebook.com/ee/index.php/site/comments/ron_polk/

Lee, or anyone, any idea when Coach Polk formalized the system? I forget.
Lee Judge
1 year ago

Jim: I got the Polk book in 1990 or so, had to be before that.
Loose Seal
1 year ago

Click, James. “What if Rickey Henderson Had Pete Incaviglia’s Legs”: Baseball Between the Numbers. Page 116-117.

“Another supposed advantage of the stolen base, as any announcer will tell you, is that base-stealers distract the pitcher, alter the positioning of the opposing team, put pressure on the defense, and do everything but part the defense like the Red Sea for the batters behind them. Such theories do stem from logical observations of the action on the field. With a man on first, pitchers throw over to first, requiring the first baseman to play much closer to the bag and the batter than he normally would. The shortstop and second baseman may play closer to second in anticipation of a throw from the catcher, or they may break toward the bag on a hit-and-run, allowing a few extra singles to dribble past. But the idea of putting pressure on the defense isn’t based on simply any runner on first; only players who are threats to steal supposedly have that effect. Thus, the stolen base creates scoring opportunities by its mere threat. That’s the theory anyway.

“To test this theory, we can start by breaking runner on first in steal situations into five groups based on the frequency of their stolen-base attempts. Then we compare the performance of the batters behind them to their expected performance given their overall stat lines. (It’s crucial to adjust for the expected performance of the subsequent batters, as base-stealers tend to be in the highest spots in the lineup so that a disproportionate number of subsequent batters will be the #3, #4, and #5 hitters.)

“In 2004, batters who came to the plate with the top 20 percent of base-stealers on base—as determined by their stolen-base attempts rate—saw their on-base average plus slugging average (OPS) increase by 34 points over expected performance. On the other hand, batters coming to the plate with the slowest runners on first saw an increase of just 4 points. Over the past five years, batters at the plate with baserunners most likely to steal on first improved their OPS by an average of 24 points, the next group 27, then 17, 20, and 13 for the slowest. So it’s true that runners on first who are more likely to steal improved the performance of the batter at the plate, but the difference between the most and least aggressive base stealers is marginal: about 11 points of OPS, with data popints bouncing around between the five groups of base-stealers. Even adding the minor and inconsistent improvement in subsequent batter performance, base-stealers rarely contribute to a team’s offensive performance in amounts that merit attention they receive.”
Lee Judge
1 year ago

Here’s another thought (and probably why you should take a walk once in a while—things occur to you when your mind wanders):

What if we’re looking at the wrong end of the spectrum?

Everyone seems to agree that a “base runner” helps the hitter, but can find no consistent evidence that a “base stealer” helps a hitter even more.

But the effect we’re looking for (the pitcher and defense altering their behavior in ways that help the hitter) isn’t brought about by a stolen base, it’s brought about by the perceived threat of a stolen base.

What if pitchers perceive anyone who ever stole a base and is currently not on crutches a threat? What if, in their minds, almost everybody is a base stealer?

If that’s true (and I don’t know that it is) then the separation we’re looking for would be at the other end of the spectrum: the players that are not a perceived threat to steal a base.

That would mean the pitcher would be free to go about his business—full leg kick, not going over, throwing whatever pitches he liked—without altering his behavior.

Ned Yost told me Greg Maddux pretty much refused to do anything differently no matter who was on base. Maddux felt that if he executed his pitches, they weren’t going to score anyway. So it might interesting to see if Maddux fared better than other pitchers when he had a runner on.

Don’t know if this idea holds water, but it’s a thought.
Loose Seal
1 year ago

Lee, It took me 5 minutes to locate the passage (above) from Baseball Between the Numbers, that you state you could not find.

You’re doing your readers a disservice by claiming (or implying) that evidence which disproves your opinions/beliefs does not exist, or that it “doesn’t make sense” to you.

If you want to engage in an honest discussion of these stats, then we can learn from each other. But if you want to present your beliefs, and then will ignore the opposing side’s evidence, “not find” the evidence cited, or conclude that it needs “to make sense” to your preconceptions before you believe it, then we’re talking about science vs. faith, and those two don’t mix.
Lee Judge
1 year ago

Loose Seal (may I call you Loose?): Thanks. That must’ve been the article Brendan was suggesting, but it was by Click not Keri.
Lee Judge
1 year ago

Loose: I was directed to an article in “Baseball Between the Numbers” by Jonah Keri. On one of the few off nights I’ll have in the next six months I sat down and read both articles by Keri and wrote a response.

The article was actually by James Click and the book is 434 pages long, so I missed it. I never said the article didn’t exist, I said I didn’t find it. I never said it didn’t make sense to me—how could I? I never read it.

(And in fact the article seems to say there is an advantage to having a base stealer on, it’s just not as much as some people think.)

If you’ve read all the comments above I don’t know how you could conclude there’s not an honest discussion going on.

Frankly, complaints about intellectual honesty from someone who went to the trouble to create a fake name and Facebook account so they could comment anonymously seems wildly hypocritical.
Brendan Woodbury
1 year ago

Lee -

The first study I linked to (http://cyrilmorong.com/Havoc.htm) specifically compared the two extremes — the ten highest and lowest base stealing teams out of about 300 seasons, so it ought to capture the two extremes.

As for Maddux, he shows the typical degradation with runners on (AVG up 15-20 points, OBP up 20-25, SLG up 30-40)

Overall: .250AVG/.291OBP/.358SLG Runners: .265AVG/.317OBP/.381SLG

However, Maddux gets worse again in a typical steal situation (runner on first, other bases empty): Stl Sit: .288AVG/.314OBP/.416SLG

Maddux might not have thought he was doing anything differently, but the results say otherwise. Maddux was a great pitcher overall, but with a man just on first, he was not great and not even good. He was certainly doing something different to allow batters to hit 38 points better against him in a steal situation.

This is actually a great test case for the theories that people build up while being involved in the game day to day versus what the data show is actually happening. It’s a great case to show the value that data offer.
Brendan Woodbury
1 year ago

Lee -

Sorry for creating the confusion. Keri is listed as the author of the book at my local library, which is where I found it (and on the Amazon page). I missed that the article had a different author.
Jim Fetterolf
1 year ago

Brendan, good work. Glad you’ve joined us.
Loose Seal
1 year ago

Lee,

Despite your ad hominem (ad-seal-em?) attack, my point holds true. Just read your own comment above mine. You state:

“Everyone seems to agree that a “base runner” helps the hitter, but can find no consistent evidence that a “base stealer” helps a hitter even more…If that’s true (and I don’t know that it is)…That would mean the pitcher would be free to go about his business—full leg kick, not going over, throwing whatever pitches he liked—without altering his behavior…Don’t know if this idea holds water, but it’s a thought.”

Your entire premise is based on a maybe- there’s-evidence-maybe-there’s-not attitude. My point is that there is in fact, a very good study with evidence that a base-runner’s speed is of extremely minor importance.

Is the evidence perfect? No.

Is it evidence? Yes.

Did you imply in your article that this evidence did not exist, and then continue implying that in the comments? Yes.

Is it your job to research your articles more thoroughly so that a commenter (with an Arrested Development reference as a username) can’t discover your mistake within 5 minutes?
Jim Wilson
1 year ago

I’m not surprised by the result. Even with a slow baserunner, the pitcher and first baseman are trying to limit the runners lead so the first baseman is closer to the bag, the pitcher is distracted and throwing out of the stretch, defensive shifts are limited, etc. Having a legitimate base stealer increases the advantage but not much because most of the advantages already exist. Again this argues against the necessity to aggressively steal.

Brendan’s point is very good. Humans’ observations and recollections are generally not very accurate. That’s why data is important. I have an essay coming on that and how I disagree with Jim F.’s statement on several occasions that advanced metrics are an attempt to “see” what a trained and experience eye sees (or something to that effect). I would also like to revisit the 75% success rate. My point about it being too low is that I think the numbers show that to steal you should have at least a 75% chance of success — on each attempt. Of course, in many cases the chances are higher than that so within a sufficient sample size the actual success ratio should be higher than 75% if there is always at least a 75% chance of success. Getz’s SB% in 2009 and 2010 was 93% and 88%, respectively, but has dropped to 75% the last two years. More to come.
Jim Fetterolf
1 year ago

“Despite your ad hominem (ad-seal-em?) attack”

Lee was relatively polite, having less experience with posters who hide behind screen names on the chat room sites. Questioning Lee’s honesty is rude and uncalled for, but a trait among the folks you run with. I remember you from last time and the other site, but do compliment your courage in showing up alone. That’s a positive step.

In general, if you have information, offer it and it will likely be discussed. Get personnel and you’ll probably be ignored.
Jim Wilson
1 year ago

Loose Seal, thanks for the contribution. I hope you will continue to particpate in the discussions on this board. I have read your posts elsewhere and enjoyed them. I think most, if not all, of us on this site fully appreciate Lee’s strengths, weaknesses and biases. We all have them, including you. Personal attacks are counterproductive to the goal of sharing information and learning. Lee makes an incredible effort to respond to all of the comments on this board, which is very unique. We may not always agree with what he has to say but he has a passion for baseball that we share and a passion for more knowledge.
Loose Seal
1 year ago

Jim F., Lee said that I was being hypocritical and then attempted to discount my points by pointing out that my facebook account is anonymous (something which you do as well). The “information” I provide is largely about the way Lee collects and presents facts to his readers. When Lee incorrectly asserts that there is not evidence to support his opponents’ points, I will point this assertion is wrong. There are many hard-working baseball writers whose work is being disrespected by such false assertions. It is not that I am calling him dishonest, it is that I question whether Lee is qualified to discuss advanced stats on his site, as I have seen him repeatedly misrepresent information that is readily available on the internet.

Jim W., thanks for your compliments. I do appreciate Lee’s strengths, such as tidbits like Hoch’s mentality with runners on base.

Lee, I’m not questioning your honesty, but I question the way you present opposing viewpoints in a public forum such as your stories. I simply think that you should research your assertions about sabremetrics more thoroughly. Or maybe just stick to the on-field details like Hoch’s mentality and the signs that baserunners use to get good jumps. There’s a ton of value in that kind of reporting, and it’s clearly something that no amount of statistics could ever uncover.
Jim Fetterolf
1 year ago

“I disagree with Jim F.’s statement on several occasions that advanced metrics are an attempt to “see” what a trained and experience eye sees”

I look forward to this, but would offer both FieldFX and PitchFX as examples of attempting to generate data that often will reach the same conclusions an old catcher or pitching coach reach on experience.

We discussed Dewan’s DsR, for instance, so I did some reading on it and what I noticed is that it has more bins than UZR and recognizes that 1st basemen hold runners on and that middle-infielders move on a run-and-hit. The next step, what FieldFX will do, is to quantify range relative to the fielder’s placement rather than an arbitrary grid imposed on a map of the field, as well as speed of the ball in play and a fairly precise vector, basically what the human eye does:)

On hitting, I’m probably more interested in metrics than some might think. I like wOBA, for instance, did some reading on it, respect the way the values are derived, and find it a useful tool.

Good posts, JW.
Terry Payne
1 year ago

Ron Gardenhire and Mike Sciosia can manage for me any day of the week. Not that I don’t like Ned, but if those teams are stupid enough to fire either RG or MS, I’d snap them up in a New York nansecond.
Brendan Woodbury
1 year ago

Jim F -

2 quick things:

1) I don’t know why teams would spend all of this money on FieldFX and PitchFX if they merely expected the data to confirm what old players and coaches already know. I think Lee’s correspondent Curtis Ruder is much closer to the truth when he says SABR researchers tend to focus on areas where data conflict with conventional wisdom (i.e. the knowledge of old players and coaches).

2) You really should read the UZR primer on fangraphs. I posted one section on here last week to clarify that it does exclude data where the defense is in a shift. I’ll post another section here to clarify that it does take into account when the first baseman is holding the runner on (and other defensive alignments dictated by the base runners):

“The base runner and outs adjustments are a proxy for infield defensive alignment. With a runner on first and less than 2 outs, UZR assumes that the SS and 2B are playing in “double play position” which is typically closer to second base and a little shallower. With a runner on first and no one on second, it is assumed that the first baseman is holding the runner. With a runner on first or second and no outs (or 1 out and a pitcher at bat), the third baseman often has to play up in anticipation of a bunt.”

http://www.fangraphs.com/blogs/index.php/the-fangraphs-uzr-primer/
Brendan Woodbury
1 year ago

Lee -

I feel like I was responsible for the confusion about Baseball Between the Numbers and the resulting spat, and I’m sorry for that, but I do think one item in your original post was unfair and reflects a concern that Loose S. raises.

You wrote: “Or a study that proved the Royals were playing Eric Hosmer out of position, but apparently—it’s what I was told—didn’t factor in the “no-doubles” defense that had him positioned on the line.”

Whoever told you that was 100% wrong.

In that section (Part One: Question and Answer) you make a good point: many proponents of Sabremetrics overstate the results. I would argue that a large dataset should carry far more weight than years of anecdotal observations, but I certainly concede your point that the best the data can do are to provide suppport for a position, not prove it.

However, there’s a flaw that often shows up in the opponents of Sabremetrics that I think deserves more coverage if you’re going to call out the proponents for overstating their conclusions. There’s a natural response when a study produces a confounding result to speculate flaws in the study design.. Jim K did this above regarding the base stealing. I’d encourage anyone who speculates about such things to go see if the study accounted for these flaws, but, even if they don’t, there’s no real harm done by speculation.

However, sometimes speculation turns into factual claims as it did with what you were told on the Hosmer defense story. I hesitate to call it a “lie” because that suggests intent, but it’s completely untrue. The actual study closes by making the following points: * Hosmer reached 5 ground balls (all up the line) that most other 1B do not.
* He failed to reach 29 ground balls (mostly in the hole btw 1B and 2B) that most other 1B do field.
* The only way that trading 5 for 29 is anything other than harmful is if the 5 he stopped would all have been triples with the bases loaded and the 29 would all have been bases-empty singles. And of course, they were not. * So positioning Hosmer close to the line appears to be preventing a few extra-base hits (“no doubles”) but allowing a lot more singles. So the net result is negative.

If this had been a comment about some random study, I wouldn’t have said anything (although I may have gone and checked the study to see if the flaw did exist). But it’s not. For one, the study focused on the player most believe will be the franchise’s cornerstone for the next six years. More important to me is that it was done by a KU student who obviously put in a bunch of time and effort. He got a bunch of recognition in the SABR community, but I think he also deserves recognition in the Royals community. He certainly doesn’t deserve to have his study misrepresented as having flaws it didn’t commit.

Everyone who cares about the Royals really ought to take a look at the work Connor did. I know the Royals have. Jin Wong, the head of the Royals stat team, confirmed that he read the study when it came out. It wouldn’t surprise me if you see Hosmer positioned significantly differently this year as a result.

http://www.royalsreview.com/2011/11/10/2538524/eric-hosmers-defense-part-2-positioning
Lee Judge
1 year ago

Brendan:

Here’s the part that jumps out at me:

“The only way that trading 5 for 29 is anything other than harmful is if the 5 he stopped would all have been triples with the bases loaded and the 29 would all have been bases-empty singles. And of course, they were not. * So positioning Hosmer close to the line appears to be preventing a few extra-base hits (“no doubles”) but allowing a lot more singles. So the net result is negative.”

In general I think that’s true, but in a specific game situation (up by one, bottom of the ninth) maybe not.
Jim Wilson
1 year ago

LOL! Lee, you are predictable!
Lee Judge
1 year ago

OK, time for one more post and then I’ve got a long day in front of me:

The one part of this experience I haven’t enjoyed is the bickering, the one-upsmanship and generally nasty attitude that the internet often fosters. Too often, I’ve responded in kind.

I love being out on the field and learning something new, I love bringing that experience back to the readers and I love the exchange of ideas that information generates.

As I said in the original post, I’m trying to bring you a point of view (the participants’) about the inner workings of the game. A point of view that isn’t often heard.

I tend to agree with that point of view, but by no means is it sacred. The point about memory is well-taken (I’m married to a psychologist and have read more than my share of material on how the human mind works). Kevin Seitzer keeps his set of stats—long ab’s, hard hit outs, moving runners over—for that very reason. He wants numbers to back up his impressions.

But if I say a base stealer helps the hitter and you’re thinking, “I know of three studies that says it doesn’t” I don’t expect you to accept what I’ve said. Or disagree silently.

It does a service for everybody—including me—to bring that information to the site. After this exchange I still think a base stealer helps a hitter, but the evidence suggests maybe not as much as I thought.

That evidence will temper any new statement I make on the subject. That’s a good thing.

I think these discussions are fine—wait—hell, I think these discussions are great. But I think they can be conducted without rancor. As I’ve admitted in the past, sometimes the fault has been mine and I’ll try to modify that behavior.

If we can all do the same, the site’s reputation for outstanding reader participation will continue to grow.
Lee Judge
1 year ago

Geez, Jim, I prefer to think of myself as consistent.
Jim Wilson
1 year ago

That’s what I meant! Consistent!
Brendan Woodbury
1 year ago

Lee -

OK, let’s assume your hypothetical and all 34 of these cases took place with the Royals up by one in the bottom of the ninth. In that situation, you’d concede six singles to avoid a double?
Jim Wilson
1 year ago

The diffence between 5 and 29 is about 40 points of BA and OBP. What would you pay in the open market for that (a player with 40 points higher BA & OBP) — $5 million a year? Of course it depends on 40 points higher than what? But it’s a lot of value.
Jim Wilson
1 year ago

Exactly my point. Based on the numbers, there is no situation where it’s a good trade.
Brian Robinson
1 year ago

On a lighter note … Although I never publicly criticized Rex Hudler about his color commentating I did scratch my head over the hiring early on. Hudler seemed too intense for such long season. It didn’t help that the Royals were losing and Hudler’s attempts at glossy that over sucked (in a word). However Rex took a respite and came back with a new, more subtle approach. I like it. Rex is starting to grow on me.
Brian Robinson
1 year ago

dang, I need to check before I click … should have typed ‘too intense for such a long season’ and ‘Hudler’s attemps at glossing that over’
Lee Judge
1 year ago

Brendan: It would depend on the situtation: six harmless singles spread out over several innings vs. a double with a fast runner on first in a tie ball game in the late innings?

Yeah, I’d make that trade in a heartbeat.

Cramming all the hypotheticals into one inning doesn’t work: you would stop playing no doubles once there was a man in scoring position.

There is a good case to be made for never playing no doubles (guarding the lines). I’ve met baseball old-timers who think it’s the baseball equivalent of the “prevent” defense. Go into it too soon and you’ll get nibbled to death.

Some people think you should never use it: why guard against the most unlikely thing?

By the way, no sweat on the author mix-up. Innocent mistake. Now I’ll have to go read the entire article.
Sean Fite
1 year ago

Lee, keep doing what you’ve been doing. Those of us who appreciate the hard work you put in for our enjoyment and education will continue to enjoy this blog.
Brendan Woodbury
1 year ago

Lee -

Yeah, sorry if I wasn’t clear on that. Obviously six singles in a row is worse than any double could be. But spread out over a number of games, I still think six singles is likely to end up with more men on/past second base than a single double is. Are you counting on being able to strand more than five of the six runners at first?
Jim Fetterolf
1 year ago

““Or a study that proved the Royals were playing Eric Hosmer out of position, but apparently—it’s what I was told—didn’t factor in the “no-doubles” defense that had him positioned on the line.””

Connor’s a sharp guy, familiar with many of his efforts, but the hole in his effort was that positioning defenders is a result of how the team is going to pitch a particular batter to try to take advantage of the batter’s skills and tendencies, among other reasons including man-on-first, tendency to bunt, and even a fielder’s abilities, some move better right than left. Watch how the Royals play Captain Jeter when he comes to town this weekend as an example. As pointed out last year, if Jeter pulls a ball the pitcher made a mistake.

One question on the 29 “singles” for Brendan et al: were they actually singles or were they just balls Hosmer didn’t reach that were in his pie slices? How many of them did the 2nd baseman get, if any? Been a lot of talk recently about shifts and UZR and DsR, so-called advanced metrics, have a hard time quantifying them.
Jim Kissane
1 year ago

Sorry if this is a non-sequitur, but I got interrupted by the evening job and the day job.

Throw up my hands and give up? Lee, I’ve been doing this since I was in junior high, trying to figure out whether Bob Gibson was worth the $125,000 he signed for after St. Louis won the 1967 World Series (didn’t have Win-Shares back then; heck, my notepad really was a notepad!). So, no, I’m not throwing in the towel by any stretch.

My point is that base-running, like baseball, is a team sport. I don’t really care if Curt Flood’s BA drops four points with Lou Brock on first, as long as the result produces runs. The individual stats are less important than run production, and Brock was a run factory. I wish I still had some of those notes from way back, because they were eye-opening. The 1967 Cardinals were a very good team, but they were at their best with Brock on base. Conversely, my recollection is that when Brock didn’t reach base, they were sub-.500. I think that’s telling.

And that was what I meant when I pointed to the 2-hole batters - that they didn’t necessarily benefit from Brock’s presence on first (in fact, as Brendan points out, demonstrably the opposite), but that Brock certainly benefitted from them. And the results benefitted the team. Ultimately, that’s what matters most.

Brendan makes some very good points, most notably that the data need to support the strategy. Pete Palmer counts the run value of a steal at .22, while the run value of a caught-stealing is -.38, almost double the opposite way. I think that’s accurate - losing runners on the bases kills rallies.

Consequently, I don’t think these Royals necessarily have to resort to base-stealing to produce runs because, in general, it’s not worth the risk. Still, there are ways to take advantage of team speed to facilitate scoring, or at least to mitigate the risk of, say, a double play (also a rally-killer). Those are the situational ethics I referred to.

The key metric of success in that regard is run production. The individual statistics - at least the ones mentioned here - don’t capture that. John Walsh’s analysis credits base-stealing “disruption” with fewer than two runs per season, but he focuses only on the next batter (Stealer on first, second base open, less than two out). Batting average increased 12.5% in those instances (BA increased 9.4% with any runner on first in those situations), but the rest of the inning doesn’t doesn’t fit Walsh’s statistical model, so we don’t get an analysis of the overall team results. Nor did his research factor in the score. I would like to see the data for a score differential of three or less, especially a differential of 0 or 1 when I would suspect the disruption factor (even if self-imposed by the pitcher and defense, as it is with the “no-doubles” defense) to increase.

So there’s still some work to do, and much of it gets into the “texture” of each situation. Those variables are going to be somewhat - if not very - subjective, and they are going to be increasingly difficult to isolate. But give up? Shoot, we’re just getting started!

And Lee, if I haven’t said so before, I love the work you’re doing here. It promotes a great deal of intelligent and fair-minded discussion. It’s the best blog in the Star, by far, and with the best contributors - Brendan, Jim F, Jim W and all, thanks for keeping it real.
Brendan Woodbury
1 year ago

Jim F -

Connor mentioned at the beginning of his article that he’s using the UZR database.

Since UZR throws out plays where a shift is on, there are no shifts included.

Since UZR does not penalize a defender on plays where another fielder makes the play, we know that all 29 of the plays in question (balls Hosmer didn’t get but 1B field most of the time) went for hits. Connor does note high in the article that there were actually 44 balls that Hosmer didn’t get to but most 1B do, but 15 of these were fielded by the 2B, so they don’t count against Hosmer. To answer your question directly, all 29 went for hits, but I don’t know if they all went for singles.

And I’m not sure it matters whether Hosmer is positioned right but the pitcher put the pitch in the wrong spot or Hosmer is positioned wrong. The team needs to adjust. If the pitchers can’t hit their spots well enough to force balls up the line to Hosmer, they need to move him back off the line.
Lee Judge
1 year ago

Jim: Thank you, the past two years and counting have been one of the most interesting periods of my life.

I’ve learned more baseball in the past two years than the previous twenty. Most of it from the pros, some of it from readers like you.

I think you hit on a very important point: context. What’s true in general might not be true in a specific situation or game.

You might hate bunting in general, but if Chris Carpenter has been shoving it for six innings in a 1-0 playoff game and you get the leadoff hitter on first in the 7th, and your eight-hole hitter is on deck with a pinch hitter to follow, bunting might be your best alternative.

(And that’s another point: it’s not enough to disdain a strategy or player—you also need to have a better alternative.)

But to get back to the general vs. specific—that’s part of my answer to Brendan. Generally, I’d rather give up one double than six singles. Specifically, how much does a double versus a single hurt me in this situation?

As usual, once you start exploring the situation, there’s no “one size fits all” answer.
Brendan Woodbury
1 year ago

Jim K -

I think you’re exactly right that the key metric is run production. My initial source of frustration with the Royals’ small-ball efforts was that we weren’t scoring as many runs as a team with our basic offensive stats typically does. In an effort to “manufacture” runs, we’re actually destroying runs.
Jim Fetterolf
1 year ago

“we know that all 29 of the plays in question (balls Hosmer didn’t get but 1B field most of the time) went for hits”

Checked baseball reference and Hosmer, surprisingly perhaps, has a very average range factor in his career so far, although gets dinged on Rtot and other stats, including BsR, which only gives him a little more love than it gives the Moose.

“The team needs to adjust”

And does, nearly daily. One of the curses of a young team or one with new players is that it takes a while to get a “statistically significant sample size” to base most everything on. Last year among the pitchers were Kyle Davies and Jeff Francis, this year Jon Sanchez and Luis Mendoza, so new tendencies to be adjusted to. Last year Hoz had Mike Aviles and Johnny Giavotella next to him quite a bit, this year Yuni, with his known left-range problems. All that a study can usually do is give a snap-shot or an average or aggregate, it isn’t always good at showing trends or ranges within a season, which is much more important within an individual game.

“there were actually 44 balls that Hosmer didn’t get to but most 1B do, but 15 of these were fielded by the 2B, so they don’t count against Hosmer.”

But the study can’t quantify how many of those balls Gio, Aviles, Betemit, and Getz should have gotten to in a shift. That 15 plus the 5 down the line became outs, 20 out of 49. That looks a little better than 5 of 34. May explain why Hosmer’s RF is average. Also reminds us of the difficulty of trying to isolate individual performance in a team game. May also suggest why only one of last year’s four 2nd basemen are still on the team.
KC Guy
1 year ago

One thing I’m noticing in the studies cited so far that are attempting to figure out the affect a runner on first has on the batter’s numbers is that it seems like they’re using a much too general approach to try to analyze a very specific situation. From what I can tell, the general idea has been to look at situation in which there is a runner on 1st, with 2nd & 3rd base open, then comparing the outcomes vs. a batter’s overall numbers.

This seems like it is too broad an approach, due to a number of variables. The main thing is that it is not situation specific – just because a runner is on first, there are many times it would not be considered a stolen base situation, the biggest being when one team has a large lead (especially in the late innings). In these cases, not only is the pitcher not concentrating on keeping the runner close, the first baseman likely isn’t even holding the runner, meaning the batter is now hitting against a normal defense and a pitcher focused directly on the at-bat. However, this situation would be included in the “stolen base situation” results of the studies, which would skew the numbers back towards the general overall stats.

Also, what happens when the runner on 1st actually steals second (or gets thrown out)? Is it still counted as a “runner on 1st/stolen base situation” at bat in the calculations, since that is how it started? At that point, the end result of the at-bat would occur with the 1st baseman in his regular position, one or both of the middle infielders at normal depth, and the pitcher focused more on the hitter, since the threat of a steal of 3rd is much less than a steal of 2nd.

There are other variables that would need to be factored in as well. A left-handed “reader” who is decently fast to the plate would negate a lot of the advantage of the stolen base threat. Likewise that pitcher wouldn’t have to worry nearly as much about the slide-step and being quicker to the plate, especially when compared to a right-hander who is normally slow to the plate. However, both of those pitchers are counted equally in the studies I have seen so far. Batters are also at the mercy of a hit-and-run – they may simply be asked to put the ball in play to protect/advance the runner, which could count either for or against their stats (and the “stolen base situation” numbers), but should almost be considered part of an entirely different type of play.

Of course, it’s always possible that some of these situations were accounted for in the studies and I managed to skim over it, but then again, I imagine there are many more variables that haven’t been considered as well..
Brendan Woodbury
1 year ago

Jim F -

UZR excludes shifts. I don’t understand why you keep bring it up. We’re only looking at plays where a shift was not on.

The study shows that there are 44 ground balls that Hosmer didn’t get to but are fielded by most 1B in the same situation as him (same situation means whether the batter is right- or left-handed, whether the batter is fast or slow, whether the ball is hit hard, medium, or softly, and whether there’s a runner at first being held on). The study shows that there are 5 ground balls that Hosmer fielded but most 1B in the same situation did not get to.

If I understand you correctly, you’re saying that the 2B should have bailed him out of more of the 44 balls he didn’t get to (and that’s why they’re not here anymore?). I don’t understand why the more obvious solution isn’t repositioning Hosmer. He’s faster than almost every other 1B. Put him in position to make plays instead of relying on a 2B to have superhuman range.

Finally, range factor is not going to be a useful measure for a 1B’s ground-ball defense. It’s inflated by all the putouts he records on throws he fields from other infielders. Any 1B on a team with groundball pitchers (especially ones that don’t get a lot of strikeouts) is going to have an inflated range factor.
Brendan Woodbury
1 year ago

KC Guy -

If the effect is meaningful, it ought to show up even if the study captures some situations where it would be absent.

It’s good to think about limitations of the studies, but it can be dangerous to just keep repeating studies, excluding data until we find the result we expected to find.

Some studies find no difference at all with a good basestealer on first (versus other runners-on situations); some find a very small advantage to the hitter. If you ran enough studies with enough different constraints, you’d probably end up with at least one that showed a big advantage. But I’m not sure you would have learned anything from that.
KC Guy
1 year ago

I tend to disagree — what if the effect is significant, but it’s being watered down because there are too many variables being included that hide the actual difference? It seems like the studies are trying to find a very general answer to what are very specific situations. If you want to know if having a base stealer on first really makes a difference, you have to be able to isolate the specific times you have a base stealer on first — in a legitimate base stealing situation — and analyze the results from there. Which I don’t really see being the case with these studies so far, since the data would have to be very specific, and be adjusted for the things that actually happened during the at bat that might have changed the situation. In the end, it seems like they’re just coming up with some very broad numbers that don’t conclusively prove anything, which isn’t very useful come gametime..
Jim Fetterolf
1 year ago

“If I understand you correctly, you’re saying that the 2B should have bailed him out of more of the 44 balls he didn’t get to (and that’s why they’re not here anymore?).”

Less “should have” than could have. We had three very bad 2Bs last year that aren’t with the team this year, so that might be a factor.

“UZR excludes shifts. I don’t understand why you keep bring it up.”

Probably because “shift”, from UZR’s pov, hasn’t been defined. Did UZR’s pie slice for 2B move a step or two toward 1st when Hoz moved to the line? Is it any adjustment from a pure straight-up alignment, Hosmer setting up two steps left of straight-up, or does it require the SS to be on the right side of second? The small “s” shift is in effect nearly every play, greater “shifts” in bunt situations or with the infield in or Dyson at the plate, smaller shifts based on the pitcher-hitter match-up or game situation. There are vertical shifts constantly based on power of the hitter. That is what FieldFX will be able to “see”, the pie grids replaced by zero-point starting placement of the fielders.

“If you ran enough studies with enough different constraints, you’d probably end up with at least one that showed a big advantage.”

Absolutely true. My experience, whether in politics, economics, or baseball, is studies can be crafted to support nearly any possible thesis and bias can be easily and often subtly introduced by choice of constraints. Connor’s use of UZR rather than the more finely sliced BsR is a choice that could influence the results of the study.

Good posts, Brendan. We’re not really on different sides so much as coming from different directions to the same point.
MattandJulie Herbster
1 year ago

I’m concerned about Jarrod Dyson’s ability to make the hard catch. By my count, there have been 3 balls now that were tough catches that he should have made. The first two were during his first call-up after the Cain injury. The third was one is the one you identified from saturday’s game. I hope this is not a trend. Dyson has the great speed to get to the ball. Now I hope he has the guts and concentration to finish off the catch.

Matt
Jim Kissane
1 year ago

That’s why I love baseball. You can take as much or as little of it as you like. With the casual fan, you sometimes don’t get much more complicated than the score. To the serious fan, every pitch is a contest with its own contexts and nuances. For the best treatment of baseball in literature, read Rolfe Humphries’ poem “Polo Grounds” - it’s exquisite.

I think it’s fun to discuss the games and strategies, from obvious things like fundamentals to the itty-bitty details like Larry Vanover’s strike zone. I also appreciate the links to serious studies. This old dog likes to learn new tricks.

Still, sabermetrics merely identifies trends, using past data in an attempt to identify probabilities and the best course of action. None of it is evidence, it’s simply history. And as they say in the stock market, past performance is no guarantee of future results.

I think Hoz is the best fielding first baseman in baseball. But he plays where he’s positioned - those instructions come from the bench. I would expect his placement to vary depending on a lot of factors: hitter’s tendencies, inning and, yes, who’s playing second base next to him.

The “no-double” defense is effective - heck, it cost KC a win when Esky scorched one down the third base line with two on against Detroit and Cabrera turned it into a game-ending double play (#7 in the losing streak). As has been mentioned, it has to be used correctly (i.e., not in the 6th inning), just as the prevent defense has to be used correctly in football (almost without exception, it is mis-used). Catchers can help, too, by not calling for a pitch a lefty can pull (not worried about our left-side infielders. At ALL.)
Brendan Woodbury
1 year ago

Jim F -

I know BsR as a baserunning metric on Fangraphs that measures how well players take an extra base in situations other than steal situations (i.e. does a player go from 1st to 2nd, 1st to 3rd to 1st to home on a single). You seem to be referencing a defensive statistic. Where does it come from?
Brendan Woodbury
1 year ago

Jim K -

There is a world of difference between “past performance is no guarantee of future results” and “none of it is evidence, it’s simply history.”

The world is not completely random, and yesterday’s best players are far more likely to be today’s best players than yesterday’s worst. Nobody can know the future — there are no guarantees — but we can make certain predictions with confidence.

No doubles defense does work at preventing some doubles. Your example with Cabrera was good, and we know that Hosmer got to five balls down the line that most 1B don’t get to.

However, that does not settle the question of whether no-double defense is effective. There’s a trade-off, and the evidence suggests that, at least for Hosmer last year, the Royals got the short end of the trade. There are basically no circumstances where an extra base hit up the line is as valuable as six singles.
Eric Blatt
1 year ago

One thing I don’t understand here is the argument over whether having a speedy runner on 1st gives the hitter a “big” advantage.

The conventional wisdom guys have pointed out some reasons why one might expect that a base-stealing threat might tilt the pitcher-hitter match-up slightly in favor of the hitter.

Meanwhile, the stats guys have cited the Click study which concludes that a good base-stealer on 1st boosts the hitter’s OPS by about 30 points over what a slow runner on 1st would contribute.

Isn’t this a case then where conventional wisdom and stats more or less align? Maybe conventional wisdom would have held the advantage gained to be a bit larger, but nobody was expecting Jarrod Dyson on 1st to turn Chris Getz into Alex Gordon at the plate.

Baseball is a game of spending great effort to eek out narrow advantages. A 30 point OPS boost turns a .275 hitter into a .280 hitter. The effect of a speedy runner on 1st isn’t worth the attention that the play-by-play guys tend to give it, but it’s not something to ignore either.
Jim Fetterolf
1 year ago

“I know BsR as a baserunning metric”

Bad brain, should have been DRS, Dewan’s thing that I’ve been reading up on.

“yesterday’s best players are far more likely to be today’s best players than yesterday’s worst”

All else being equal, injuries being an obvious inequality, Mike Sweeney and Roy Oswalt two who come to mind, another inequality being a player changing conditioning, swing, or approach, as the Worst Outfield Ever did last year or Master Chen over earlier years. Inflection points have something to do with how history is weighted, Gio a recent example, hot hitter before the hip injury, dud afterwards, now seems to be moving past his rehab and creating a new, positive inflection to his career trend.
Brendan Woodbury
1 year ago

Eric -

I’m not reading the Click study the same way as you. The 30 point spread comes from the smaller sample size (a single year). The longer term study shows an 11-point spread, and I think that’s the spread we would expect to see going forward.

The bigger question is whether it’s worth attempting more stolen bases at the cost of a lower success rate because of the distraction it provides. If the distraction only shows up as an 11-point OPS difference for the best base stealers, it suggests that lower SB% for more distraction is a bad tradeoff to make.
Brendan Woodbury
1 year ago

Jim F -

Doesn’t DRS have Hosmer as a worse defender last year than UZR does?
Eric Blatt
1 year ago

Brendan -

You’re right that the 5 year sample had the gap at 11 points rather than 30. But the data looks to me like even the 5 year sample may not have been large enough to fully smooth out the distribution.

If one really wanted to precisely measure the impact of this “distraction” effect, they’d have to control for more variables than Click controlled for. But I think what we really care about here is the order of magnitude of the effect, which Click’s study gives us.

We know that the impact is small but statistically significant. It’s probably better to talk about the battle between the pitcher and the baserunner as it relates to the likelihood of a successful steal than talking about how it helps the batter, but announcers (and writers!) have space to fill.

A good rule of thumb for commenters would be to cover the big things well before talking about the little things. A stolen base (or pick off) is a big thing, and it’s worth going into the details. A 2% advantage for the batter might not be worth as much detail.
Jim Fetterolf
1 year ago

Brendan, DRS had him at -9 last year, UZR/150-8.3. This year -1 and -5.8, respectively. My eyeballs tell me that both grid-based systems aren’t real good yet, but DRS has made some adjustments to make it better by using more bins and a couple of adjustments.
Lee Judge
1 year ago

Man, I gotta tell you that I just love it when you guys go on with without me. It tells me that this is becoming a clearing space for ideas, comments and conversations.

Pretty neat.
Brendan Woodbury
1 year ago

Jim F -

I don’t know enough to have an opinion between DRS and UZR (or Total Zone or other similar systems); however, make sure you’re comparing like concepts. DRS should be compared to UZR, not to UZR/150 (which is the UZR projected out for 150 games played. In Hosmer’s 127 games at first last year, his UZR was -7.0 and his DRS was -9. In his 21 games this year, his UZR is -0.8 and his DRS is -9.

For whatever reason, DRS seems to consistently see his defense as a little worse than UZR. The fact that he’s negative at all I think indicates continuing problems for the Royals’ positioning of him,