The Dick then added this...

[Now you send in a post with a profanity. What's wrong with you? All I wanted to know is if Cecil Cooper should have bunted. The bottom line is you don't know if he should have or not. Just say, ''I don't know.'' Everyone knows there are eight thousand possibilities. It's the same way at Whataburger, but I've got enough sense to say ''No. 1 with cheese.'' It's ''yes'' or ''no'' on bunting. Did you expect Cecil Cooer to convene a geek convention and hold a six-hour roundtable. Good Lord.--Richard]

I think I might be with Richard in this case, especially after reading the post and seeing that last (above) response....

Of course MGL knows better than Cecil Cooper whether the bunt was called for or not, but his post was inflammatory and he basically baited Justice into his responses.

"I thought it was a great post that would be much appreciated."

Enough said there.

I think posting 1500 words on the value (or lack thereof) of the sacrifice is a bit much, regardless of how right MGL may be.

I sort of have to side with Justice here as well. I mean, really, I tried to focus through MGL's post but it was pretty dense. Good information, for sure, but if I have to pick a side I tend to the "Yes or no" camp on this one. Paralysis of analysis and all that.

Of course MGL knows better than Cecil Cooper whether the bunt was called for or not

Maybe he does and maybe he doesn't. You don't have to be a statistical guru to have good instincts and baseball smarts. I've seen lots of people that are really talented at something do their job incredibly well and not be able to really explain in sophisticated detail how or why they did it. Doesn't make them any less competent.

Of course MGL knows better than Cecil Cooper whether the bunt was called for or not,

maybe he should of mentioned whether the bunt was called for or not somewhere in that treatise

This is the classic left brain/right brain battle.

I make a living writing, and have a twin who is an auditor.
I can teach this freaking course.

The two sides don't understand each other, at all (well, we do, but we grew up together. That helps a lot).

Both sides are right here, in a way. It's a matter of walking the mile in the other guy's shoes.

MGL comes across as pretty abrasive quite often, but not in his post on that blog. Justice is just an ####### who thinks of his blog like an editor-mandated MySpace page. The last thing he wants is to be drawn into a lengthy discussion about baseball strategy.

I really don't know why MGL couldn't have just taken a stab at answering the question at the end of the post. It's not that hard to figure (estimate) out [no, I haven't done it, while not hard it is time consuming], and if you're going to write a treatise promoting your book the least you could do is actually answer the question. That said, Justice is completely worthless so I don't see how anyone can be on his side in this. You can think they were both jerks(I do, and have thought MGL was a jerk for as long as I've read his posts), but there's just nothing good about Richard Justice in this post or any others.

Unless Cooper has read the book (The Book), there is no way he would know whether a bunt in that or any other potential situation is correct, and there is certainly no way any fan would know.

Of course MGL knows better than Cecil Cooper whether the bunt was called for or not

Managers haven't studied the issue to the extent that MGL has, true, but most of them know the percentages well enough so that they don't bunt all that often with even decent hitters. Take out the pitchers from the distribution and I think that you'll find that there really aren't THAT many situations where managers bunt with hitters who really shouldn't be bunting. Further, as MGL knows (and said in the linked entry), if the other team knows that you WON'T bunt, then they can safely ignore that possibility when setting their defense, so you have to at least keep the "threat" of the bunt in the mix.

That'a a lengthy way of saying that MGL "doesn't" know better than Cecil Cooper whether or not the bunt is appropriate "in that specific situation". The manager has to look at what the defense is doing against his team, how good (or poor) a bunter is batting, and what other strategic options he has, and weigh all of those factors appropriately.

-- MWE

Unless Cooper has read the book (The Book), there is no way he would know whether a bunt in that or any other potential situation is correct, and there is certainly no way any fan would know.

Yeah, I thought that was incredibly arrogant as well.

Justice is a first-rate columnist who seems to turn into Mr Hyde on his blog. Knowing that, I might not attempt to start a deep academic investigation of the bunt on said blog :)

I've never really understood why MGL is so arrogant about this stuff. For example, any 2nd year economics PhD student could do any of the stuff he does. That doesn't mean it's not interesting or valuable work-someone has to do it-but I can't see acting like this stuff is an impossible task that no one could understand without reading "The Book" when you could pay a grad student 5k for the exact same results.

A pox on both houses. :-) OK, not a pox, this is pretty trivial. Some of this is just standard misunderstanding of blog posts -- i.e. a face-to-face conversation would have gone a bit smoother. But yes, MGL did, unfortunately, come across as he often does. One, his post wasn't that clearly written; two, there was lots of "there's no way anyone could know the answer unless they've spent a week analyzing the data" and "opinion is worthless"; three, there was at least one shout out to "nonsensical logic" which, if I were Justice, I might think was directed at me.

Justice was no prize either. He could have just said "look, I'm interested in an answer as to whether Cooper should have bunted in that specific situation (and why) not a treatise on the sac bunt." The added bit about stat geeks was unnecessary, although again, face-to-face, that might have come out a bit more like "you guys know a lot of stuff but you sometimes have a really hard time explaining it clearly." That's absolutely true ... and unfortunately, MGL is sometimes a classic example (as am I).

What concerns me more is how does MGL ever think anyone, including him, is ever going to have all that information? He talks about 7 different outcomes from a sac bunt attempt. How many batters (even including pitchers) have anywhere near enough sac bunt attempts to answer that with any accuracy? Mark Loretta, who I think was the batter in question, has just 56 successful sac bunts in his long career. MLG says about 60% of sac bunt attempts end in that, so let's say he's had 100 sac bunt attempts in his career. Even assuming the chances of those outcomes are constant over time (i.e. no age/injury effects) and constant across grass/artificial turf, etc. NOBODY has enough data to know if Loretta is a "good" bunter or not. And, sure, in that situation, I assume that Cooper has as good a guess as anybody.

I haven't read the Book so I don't want to lodge a formal criticism, but I simply don't believe there's ever enough information for a statistical analyst to know whether a given bunt situation is correct or not. There are about 20 variables involved and almost none of them are measured with any accuracy. The number of simplifying assumptions are staggering. It seems to make about as much sense as predicting the outcome based on whether the next pitch is a knee high inside fastball, a knee high outside fastball, a high fastball, a cut fastball, a breaking ball or a change except you probably have more data in the pitch outcome case.

In practical decision terms, it probably comes down to these factors -- (1) is the batter a good bunter (managerial opinion); (2) is the 3B playing in (fact ... or opinion if it's "far enough in"); (3) is the 3B a good fielder (measured in the general sense with a lot of error; measured in terms of "is he good charging on bunt plays" probably by opinion); (4) is the batter fast (measured with relatively little error); (5) base/out situation (fact).

I think the game theory stuff is mostly silly. Yes, if you NEVER bunt, teams won't play in on you. Has there ever been a team that NEVER bunted? No. The same silly argument comes up in base-stealing. If you never steal a base, they won't hold you on. Except that of course everyone knows that as soon as they stop holding you on, you start stealing. So no team would ever make the decision to never hold any runners from team X on because they know that would be moronic. There are some reasonable hypotheses around that you won't throw over as often or you'll have your 1B break off the bag sooner, etc. but I don't think anyone has looked at that yet.

Unless Cooper has read the book (The Book), there is no way he would know whether a bunt in that or any other potential situation is correct.

MGL is the way, and the truth, and the life; no-one comes to the True Wisdom, but through Him.

mgl complaining about a post's tone? I think the irony meter just went tilt.

MGL is, frankly, an intolerable, raging ***hole more often than not. The way he often conducts himself online is a discredit to the sabermetric community, and for the sake of all of us he really ought to just STFU and go back to numbercrunching. And apologize to Richard Justice for crapping a humorless, arrogant lecture out onto his blog instead of answering the ####### question.

Seriously, his smarts are impressive but his personality makes him a remarkably irksome presence in my opinion.

Edit: I just read his full post on Justice's blog. My god! Sit down MGL, nobody is interested in reading a treatise. It's every bit as bad as Justice said: a preeningly self-congratulatory & condescending & INTERMINABLE tract when a simple sentence or short paragraph would have been nice. And Justice is right: he does come across as the sort of man who never has never known the touch of a woman, if you know what I mean.

his personality makes him an unwelcome presence in my opinion.


Well then, next time, don't let him in your opinion. He can bloody well stand outside your opinion, like the rest of us.

mgl complaining about a post's tone? I think the irony meter just went tilt.

No kidding.

MGL is, frankly, an intolerable, raging ***hole.

I tend to agree, which is a shame, because he's done a lot of interesting, novel work in areas of stat analysis that hadn't been well-studied before. He's sort of like the would-be elected official with great policy ideas and a condescending, holier-than-thou demeanor in speaking to the voters (resulting in his more likeable, less intelligent opponent winning the election--not that I'm drawing any parallels with real-life examples here...).

I'm carrying no water whatsoever for the Richard Justices of the world, mind you (tools come from all angles, perspectives, and ideological stripes), but I've always thought MGL was a godawful spokesman for the stathead community, which, to win converts, needs its ideas to be presented by someone who doesn't completely fit the stereotype of a people skills-challenged, condescending know-it-all.

NOBODY has enough data to know if Loretta is a "good" bunter or not.

While I may be totally wrong on this, I would think that how a batter bunts in practice situations (Eg, BP) is more indicative of how they will perform in a game situation than how he swings a bat. In other words, the difference in difficulty between bunting an 80 MPH BP fastball and a 95 MPH fastball is not as monumental as the difference between hitting them.

What concerns me more is how does MGL ever think anyone, including him, is ever going to have all that information?

I quite agree with Walt that this is the key point. The reason MGL's post didn't conclude with "So Cooper should/shouldn't have told Loretta to bunt" is because it's impossible to ever draw such a conclusion based on MGL's method. He says so himself:

There is ABSOLUTELY no way (in heck) that it is possible to figure out if and when a sac bunt attempt is correct until and unless you know exactly how often the batter is going to [do one of seven things for which there will always be totally inadequate sample size]

So then what good does it do to put those unknown probabilities and quantities in if it makes the problem unsolveable? It's simply not a practical approach if your guide to decision-making never actually yields an answer.

Agreed that MGL sounded, if not inflammatory, certainly awkward and arrogant. Justice's reaction is strange though. Until MGL shows up Justice is siding with stathead common wisdom. His defense to the first people that disagree with him is, more or less, is: ask a stathead, he'll tell you I'm right. So it's confusing when he lashes out with the anti-stats stuff.

By the way, I did try to read through MGL's post. And the thought that crept into my head was, "You're taking the fun out of the game." That is one fo the most infuriating charges ever levelled at statheads, and in my opinion usually baseless. But I actually turned into Murray Chass for a few moments today. I'm not sure how many times I've had that moment before, but a 1,500 word analysis of the sac bunt in which the conclusion apparently is something like "It's too complicated for you or anyone else to evaluate with any certainty" did it to me.

His defense to the first people that disagree with him is, more or less, is: ask a stathead, he'll tell you I'm right. So it's confusing when he lashes out with the anti-stats stuff.

Well, I can see Justice being annoyed that the first stathead doesn't say "Yup, Justice is right", but rather "Buy my book! And by the way, fans are too dumb to know the answer anyway."

"Yup, Justice is right", but rather "Buy my book! And by the way, fans are too dumb to know the answer anyway."

Stand back, ladies and gentlemen! This matter requires the services of...a trained sabermetrician!

Did you guys read the whole thread? Justice made fun of anyone who said that Cooper should have called for a bunt. He was like a renaissance doctor making fun of a Roman Empire physician: "you idiot, it's quite clear why he's sick! his humours are out of order. bleed him a little - that will see him right, as anyone who has studied modern medicine knows." It's the certainty that he's right, despite not actually researching the subject himself (other than by reading second or third-hand accounts of earlier research, apparently), that seems to be why justice bothers MGL.

unfortunately, MGL is not a gifted writer, and he tends to come across as arrogant and condescending. so while i believe he's 100% correct in criticizing Justice's attitude in this post, he doesn't do himself any favors with the way he does so.

I need to re-read the chapter in The Book again, but I wouldn't be surprised if it wasn't out of date. Unless the percentages therein are nash equilibria, if the opposing manager knows you bunt with the frequency prescribed in The Book, he can change the way he arranges his defense in such a way as to make the stated percentages not optimal any more.

Also, despite his research, there is no way he can categorically state something like "the pitcher should never square away before the pitch is thrown" - something like that changes the likelihood of getting a bunt down in fair territory, and thus alters the entire analysis. Say a pitcher strikes out 90% of the time that he attempts a sacrifice but doesn't square away first, but has a normal success/failure pattern otherwise. However, because pitchers know they suck at bunting without squaring away early, they do so 99% of the time they are asked to sacrifice; that causes the defense to behave a certain way, etc etc. Along comes MGL saying that pitchers should never square away because it brings the third baseman in too close, and his historic data proves it. All of a sudden, they start striking out 90% of the time.

That's an extreme example, but it's certainly at least a little bit plausible. It's what makes the certainty of MGL's pronouncements so distasteful to many of us, i imagine.

Stand back, ladies and gentlemen! This matter requires the services of...a trained sabermetrician!

Man that would be great. I could charge like $100 an hour and go on service calls. Maybe set up an apprentice system where I can get underlings to do all the work and I still get a sizable percentage. Ah, dreams...

Man that would be great. I could charge like $100 an hour and go on service calls. Maybe set up an apprentice system where I can get underlings to do all the work and I still get a sizable percentage.

saber-amway!

Man that would be great. I could charge like $100 an hour and go on service calls. Maybe set up an apprentice system where I can get underlings to do all the work and I still get a sizable percentage. Ah, dreams...

Or you could get your apprentices to start writing thrillers featuring sabermetricians. The Regression, The LOOGY, The Small Sample Size Murders

Now that we've all pissed on MGL a bit, let me defend him too.

I don't think he's a raging #######. He's blunt. He has little tolerance for fools ... and his definition of fools may be a bit wide. He doesn't write that clearly a lot of the time.

But ... leaving aside that the folks who've actually met him seem to find him a nice guy, he does seem to be genuinely trying to be helpful. There may be a certain arrogance to a 1500-word post (says the king of 1500-word posts), but he's just trying to give the best answer he knows and get as much "truth" out there as he can. And, god love 'im, he goes where the numbers say to go, I don't think I've ever seen him even put forward the "my projections say Jay Bruce is gonna suck this year but I think he'll beat those numbers" sort of CYA argument.

As to statistics ... 2nd year econ PhD student? Hell, any undergrad econ/stats student should be able to do that stuff, it's pretty standard regression. If memory serves, when Bradbury was comparing fixed and random-effects models for the Mazzone effect, MGL didn't even know what he was talking about. He likes to toss around "Bayesian" although that's not really what he's doing.

But that's fine. To my knowledge, nobody in sabermetrics is doing anything beyond entry-level econometrics. The value they bring is not statistical sophistication, it's a hugely valuable understanding of and dedication to the data and measuring/adjusting things as best they can. Generally speaking, that's a hell of a lot more important than tweaking the model to get a slightly better standard error estimate.

#32 - Walt:

Yes, yes, yes. All well and good. But then I read stuff like...

If you really want a thorough analysis of the bunt, get The Book.
[...]
I will leave you with this:
[...]
Unless Cooper has read the book (The Book), there is no way he would know whether a bunt in that or any other potential situation is correct, and there is certainly no way any fan would know. Opinion is what you think of your wife’s earrings. The answer to these questions cannot be garnered from opinion; in fact, opinions are irrelevant
[...]
Try incorporating all of those things in your head.
[...]
I may have given away too much, and now you don’t have to read The Book!

Please tell all of your friends and relatives that:

1) NO ONE, but someone who has (competently) studied or read about sac bunt attempts and all of its statistical intracacies, can possibly know or figure out when a sac bunt attempt is correct or not, other than in obvious situations, so everyone can take their “opinions” and well, you know…

...and I want to donkeypunch the man. Repeatedly. If that isn't your instinctive reaction too (okay, minus the violent hyperbole), then you may well have the same problem relating to the real world that he does.

Oh and incidentally, The Book? It's only okay. Decent concepts, shoddy execution.

I don't think he's a raging #######. He's blunt. He has little tolerance for fools ... and his definition of fools may be a bit wide. He doesn't write that clearly a lot of the time.

That does make you a raging ####### in my opinion.

Want to be blunt? Fine. Honesty's a good thing, admirable even. Too often things get suger coated.

Want little tolerance for fools? Great. If you sat there trying to listen to everyone's arguments, even the stupid ones, you'd be there all day.

Can't write clearly? That's cool. Not everyone writes like Albert Camus.

But you can't be all three. You can't be unable to fully express yourself, then become indignant and even a little insulting when people don't understand you. You can't be blunt, then have (as you note) a nearly complete inability to tolerate opposing points of view. You can't be blunt and be unable to write well enough that you don't annoy people.

I've read probably hundreds of articles and/or posts by MGL, and at least three-quarters of the time he comes off like a complete wanker. Sure, he's done great work in the realm of quantifying fielding, but much of it is mitigated by his obnoxious and arrogant persona. I hate this argument, because it's often used by people who don't even want to make the effort to understand sabermetrics and deeper philosophical thinking about the way baseball is run, played and managed, but MGL isn't finding opposition towards new ways to combat AIDS or global warming or global terrorism. It's baseball, so cut the crap and stop being such an unsympathetic dickbag.

How arrogant do you have to be to call your book, "The Book."

and could somebody please hook those guys up with a little html. thats the most third grade looking website . .. ... ever!

#32, read #16.

#33 -- I don't think MGL has an inability to tolerate opposing views ... as long as those views are based on strong data and analysis. And we'll have to differ as to our definitions of "raging ########."

But we all differ. MGL has contributed a ton to sabermetrics and I'm more than willing to put up with his obnoxiousness in exchange. To others, it's not worth the tradeoff.

I deal with people like this all the time. I am occasionally a person like this. Either you take the value they have to offer and ignore the other stuff most of the time ... or you don't.

Camus wrote clearly?

How arrogant do you have to be to call your book, "The Book."

None at all. You just have to know a catchy title that people will remember and cite when you find one. I suspect that John Dewan became an instant hit even with mainstream writers because his was called the Fielding Bible. (And I recall Tango posting "damn, I wish I'd thought of that title.")

And, crap, I can't remember his name but an economist wrote a book called The Death of Economics. He got a lot of heat from economists about that. In either the paperback edition of that book or a subsequent book, he noted that clearly economists knew nothing about marketing. Yes, he could have called the book something like "the challenges posed to standard economic theory by nonlinearity" but that, oddly, he thought The Death of Economics had a bit more ring to it and might sell a few more books.

How arrogant do you have to be to call your book, "The Book."

About as arrogant as this guy

Sounds to me that Justice is doing the same type of analysis as Pete Palmer. With a few very specific exceptions, base for an out with a position player is a bad trade.

However it's pretty close to mandatory with a good bunter who has a substantial chance of reaching, and it's a really bad idea if the batter isn't a good bunter.

For the rest, even if we had perfect information on the considerations Walt mentions in #16 (and we never will of course) -- well we're talking about something that will have a very small impact on your probability of winning. Best I can tell you're better off worrying about batting order and that's of trivial importance.

Again, best I can tell (and I've checked as closely as I can) any time that you're talking base for an out as the consolation prize a bunt is going to be reasonable. And I think reasonable is as close as we can get.

Walt/31: good post.

***

Eso/32: I don't mind, at all, any criticisms of The Book. But, it is unfair to make a one-line review as you do, without any kind of support to your argument. MGL, in his post, however inflammatory some may take it, at least provides the reader with a good deal of information to consider.

So, please, expand on the "shoddy execution" statement. I'll check back in later to agree with you or explain otherwise.

***

Gambling/34: Exactly what is it about the website that you find third-grade? I'm pretty sure the BBWAA website is worse.

In any case, I prefer constructive criticism to just pure attacks, especially since the look of the website wasn't even an issue here! It seems that this was an opportunity to make a negative statement, so you grab for something. (Btw: MGL had nothing to do with the design.)

MGL offered information in his post. If you make such a negative post, at least leave us with something useful that we can move forward on.

***

Walt is an example of one that has consistently made good posts in this thread, offering at least a devil's advocate view, so that any reader can consider both sides somewhat fairly.

***

The reaction to MGL reminds me of the reaction to Clemens and Pedro. Great value for a long time, but once you get on the bad side of your supporters, man they will pile on you, and forget about everything else.

I'm not necessarily defending MGL, but at least stick to the specifics, rather than making general condemnations.

***

By the way, I don't think MGL really cares if we sell any copies of The Book. If I were to let him, he'd probably give them all away. In fact, every dollar he earned from The Book was donated to Retrosheet.

Kevin, I'd say the reason to go through the numbers is so that others can:

a) check your assumptions

b) use the work for other situations

c) (most important) see if you're just blowing smoke.

Anybody can offer up an opinion on the matter at hand. MGL's work shows that he understands the issues at hand.

And I'm sure this is probably what infuriated Justice too. Criminy, what's the point of all the numbers crunching if you can't draw any judgements from them?

Yes, but it's not just the failure to draw a conclusion that's annoying (though it IS annoying); it's the failure to draw a conclusion after completely belittling people for even TRYING to draw a conclusion that MGL himself ends up admitting cannot be drawn with anything resembling certainty. To spend 1,500 words saying "you ignoramuses can't possibly understand the issue unless you've read my book" only to say, "having said all that, everything's inconclusive, and I don't really have an answer either" (where not having an answer isn't really acceptable, because the decision has to be made [almost instantaneously] whether to bunt or not), strikes me as serious asshattery--the equivalent of launching into a self-righteous closing argument diatribe against a defendant, then beating a hasty retreat from the courtroom just as the "not guilty" verdict is being read.

Retro, I don't agree that not having an answer is unacceptable. There are situations where you don't have a provably correct answer.

Having said that I'd suggest that a better format for MGL's comment would have been something like:

Too close to call. Math follows for the two or three people who are interested. (Don Malcolm has had the pleasure of editing my work and would probably find the idea of my offering editorial suggestions deeply amusing)

and could somebody please hook those guys up with a little html. thats the most third grade looking website . .. ... ever!

It's simple, elegant, effective, AND PORTABLE, works across multiple platforms, unlike the huge numbers of websites that are "hooked" up with a "little" html, that will not work if the reader is using a browser or setup that the designers of the site did not test with / for.

This is just my personal/professional experience, but the most contentious relationships I've seen have almost always had their root in poor communication, much along the lines of what we've just witnessed between Justice and MGL. And to me the failing is clearly on MGL here. This one statement sums it up best:

Wow. If you are serious with that statement, you are a real #######. I'd rather be incoherent.

I'm sure MGL knows it's not a zero-sum game. But I've been getting the impression over time that MGL thinks he's not incoherent, and that anyone who doesn't get what he's saying is a fool not worth his time, except for scorn. I generally get what MGL is saying, but I have to admit he doesn't ever make it easy.

The key thing to remember in communication is that the person giving the information has a responsibility to make sure the message was received as intended. MGL takes no such responsibility here (and in other cases in the past), even though it's clear the message wasn't received as intended. Justice may be an #######, but it's also clear he didn't get an answer. He got a lot of information. It's like he ordered a pizza, and instead of a pizza box he received a shipping container packed full of ingredients, and a note somewhere therein saying "there is no pizza".

Let me try to improve his answer:

Q: Should Cecil Cooper have had Mark Loretta attempt a bunt in that situation?

A: We'll never really know for certain.

Objectively - taking out all matters of opinion and judgment, and using the historical record - we can't get there. We can tell from the data that context greatly affects the outcome: the score, the number of runners on base, and the inning; how good the hitter is at bunting; how good the opposing team is at fielding bunts; whether the hitter has enough speed to have a decent chance of beating the throw; and so on. (Read The Book: Playing The Percentages In Baseball for a detailed look at this.)

In this specific case we know there were runners on 1st and 2nd, with no out, in the 6th inning, down 2 runs. But many things we don't really "know" in an objective sense. We don't "know" how well Loretta is at bunting - he hasn't bunted enough for us to really judge that with statistical certainty. We don't "know" how well the defense was prepared for the bunt. We don't "know" - before the pitch was thrown - whether the pitch was conducive to a good bunt, or a double to the gap, or a GIDP. For these things we need subjective opinion and judgment.

In the end, the quality of the answer depends on the knowledge of the data, and the quality of one's opinion to bring us where the data can't. Me? I believe opinion is irrelevant, and had I been in Cecil Cooper's shoes Mark Loretta would still have been standing there waiting for instruction while I waited for the accumulation of several hundred years of data needed to answer the question to my satisfaction. YMMV.

That's still longer than it probably should be, but it's a start.

MGL is a real nice guy, despite how he comes across in some postings. The problem in this case is that his first post, while containing great information, was completely overwhelming and probably well over the heads of most of the audience.

I'm kind of skeptical when it comes to the practical value of bunt or not bunt analysis. A manager has a limited time to make the call, I don't think MLB would appreciate any delay that takes place while an analyst feeds all the relevant data into a computer.

And for some of the information, a manager is going to have a better grip than any compilation of data - if Cecil Cooper wants to know how good a bunter someone is, he can observe a Mark Loretta bunt more times in practice than retrosheet has of him bunting in his entire career.

if Cecil Cooper wants to know how good a bunter someone is, he can observe a Mark Loretta bunt more times in practice than retrosheet has of him bunting in his entire career

This is an excellent point.

village/44: nice post. If we were all as lucid as your alternate post, we'd have no issue. Or wars.

***

AROM/45 is right: the lack of knowledge that we have as analysts would ideally be filled by those scouts (or managers or fans) with it. A model is completely dependent on its inputs. MGL has fantastic model on the bunt (i.e., he's thought of everything that you'd want, when considering whether to bunt or not).

What he doesn't have is all the necessary data to establish the breakeven point for specific instances. He's got the breakeven points for generic pitchers and fast runners and slow runners and great hitters and bad hitters, and good bunters, etc. But, Mark Loretta specifically, in that situation against that pitcher, with that fielding alignment, with expectations or not of Loretta bunting or not bunting (game theory)? We don't know. We can't know.

This would be similar to those who want Vlad to swing less. You can create the perfect model as to when to swing or not, but you still need to know what happens if you ask someone who usually swings at something to be more patient. And how will pitchers react to that, knowing this. The model can still work, but we are completely reliant on reasonable data. (See Clay's excellent article in BP08 for a good what-if domino effect.)

Overall, we can come to a good conclusion as to what to do on 0-2 and 3-0 counts. But, for specific instances, we really have no idea at all if a batter should swing or take (and how hard/quick to swing), if the pitch is anywhere close to the strike zone. If someone ends up swinging on 3-0 counts 50% of the time (league average is under 10%), and he ends up hitting league average in those counts, then we know he made a poor choice... overall. But, we don't know, for each and every instance, whether it was a good call.

The only way to know, for a specific instance, is if you give the benefit of the doubt on the unknowns, and it STILL comes out as an easy call (like I did with the "When to walk Barry Bonds"). In some cases, it's a very easy call (either way), and in many cases, it's too close to call, because of the unknowns.

I like the answer in 44. One more important point since it's a 1st and 2nd situation.

We don't know how often this particular set of fielders will try for a play at third that's just not on. One of the best possible outcomes is a fielder's choice with no outs recorded.

if Cecil Cooper wants to know how good a bunter someone is, he can observe a Mark Loretta bunt more times in practice than retrosheet has of him bunting in his entire career.

I don't think you could get more precise than a letter grade from watching Loretta in practice. Good enough for my purposes, but not for people who want to know the one true answer.

Who the hell is Richard Justice?

He wrote "Cocaine"!

Newsblog: The Book Blog: MGL: Who the hell is Richard Justice?
(49 - 11:33am, Apr 04)
Last: The District Attorney


I'm getting really interested in these types of sidebar looks. One yesterday was "Who do you want up in the clutch...backlasher."

Today we learn that Richard Justice is the District Attorney.

In what may be a rare occasion, I largely agree with Davis.

First and foremost, I disagree with people that say MGL is "incoherent" I don't think there is much problem in understanding anything he writes. Second, I don't think he is globally caustic. I've seen him involved in plenty of conversations (some of them with me) where there is no vitroil at all.

He has a pretty sizable ego; its certainly not bigger than people like James or some of the self-proclaimed sabermetricians that you find on this board. I don't think putting a :) after your insult makes it any less of an insult, and cleverly using humor doesn't also make it less of an insult. It just seems that many people here view someone who insults with snark as being cool, and those that insult directly are assoles.

I don't know if he has a grudge, but I wouldn't blame him for having a small one. He has really put in some hours pouring over data. He has made some real discoveries, and I think he has made a living being a REAL analyst for much of his career. Meanwhile, you have people claiming to be "sabermetricians", some of which can't do hardly any mathematics beyond addition and subtraction. Others that con other people into doing their work. Even worse, you have some that publish results based on a paucity of data and make overreaching conclusions based on such simple findings.

If the goal is to make people think that sabermetrics is the next noble profession that everyone's mom wants for their kids, then maybe the community does need to police mgl. Of course, there are some much bigger offenders that should be policed, including columnists that drop the word "Idiot" at least once a month.

However, if the goal is to make sabermetics a real science with tangible results and useful in the business of baseball, then mgl (and tango) are not your problems. They are your champions. I'd worry about policing other people that can't back up their statements with their data.

Today we learn that Richard Justice is the District Attorney.

Someone go over to his blog and tell him to keep his son away from the porn mogul's daughter.

The key thing to remember in communication is that the person giving the information has a responsibility to make sure the message was received as intended.

Simply not true. MGL is not a professor, grammar school teacher, advocate, parent, babysitter, or apologist when he is posting. He does not have a duty, and if he did, he isn't and shouldn't take that duty to reach an audience completely unfamiliar with analysis.

In fact, taking that duty would excerbate the problems that others claim --- the length of his post.

If you don't like his style, that is one item, but you can't parlay that into a bunch of Catch-22 criticisms:

(1) (a) Posts too long compared with (b) Arrogant with short, opiniated answers

(2) (a) Duty to explain with (b) writing a treatise

As is plainly seen in this thread, much of these types of criticisms are aesthetic criticisms.

If you want to know what the man means, and YOU are unable to do it from the post then maybe you should go ask somebody or do a little rsearch. If you still don't know, you can certainly ask a question and I bet you get an answer.

Of course if you get in his face, or make your question personnel, then I bet you catch his ire.

Who do you want up in the clutch...backlasher."

WOOOOOOOO

The key thing to remember in communication is that the person giving the information has a responsibility to make sure the message was received as intended.

That is true, at least according to a Communications professor I once had. Obviously you can't be sure that someone understands, but you're supposed to do your best to make sure they get what you are trying to convey, which is, after all, the point of communicating with other people.

That would obviously mean tailoring your message to the audience (at least to an extent), so maybe dumbing down one's message/language on Richard Justice's blog would be wise.

I reread the Justice thread. The original MGL comment (at 1:22) reads much tamer the second time through (to me anyway). There's alot of great comments in that post. It's about as an informative a 1500 word post on the sac bunt as you can expect. (I guess we all did a Word count on that?) Do we have a disagreement here from anyone?

Justice's followup line to that was I thought perfect, whether intended to be sarcastic or jovial. MGL's response to that was fine. The post following that was from a reader who clarified who MGL was.

Until here, we're ok, I think.

My problem, as I suspect mgl's, is when Justice says: "That's the problem in dealing with the stat geeks. They've got so much stuff in their brains they can't get a coherent thought out."

But, MGL's post was very coherent. Village/44 captured it in this thread.

If someone were to come to my blog, and post an intelligent post like MGL, I wouldn't then go ahead and insult him! How does that make any sense? You throw a party, invite comments, get smart content, and then... insult? Welcome to the world of Bill O'Reilly.

There's a difference between complaining about the tone of a post and the content of a post. While you can take exception to mgl's tone, his content was spot-on. Justice's later response to which mgl took exception was not a complaint on Justice's tone, but his content (specifically, the insult).

There really wasn't a reason for mgl to retaliate to Justice's level, but perhaps he felt insulted enough that having contributed so much to his blog, that a bit more respect was deserved.

The followup posts by the readers simply reaffirm that his post was well-received by some.

Take out the pitchers from the distribution and I think that you'll find that there really aren't THAT many situations where managers bunt with hitters who really shouldn't be bunting.

How quickly we forget.

I don't think you could get more precise than a letter grade from watching Loretta in practice. Good enough for my purposes, but not for people who want to know the one true answer.

Actually, a letter grade would suffice for purposes of making a model. If assigned by knowledgeable people, it would probably be better than a rating someone could come up with by merely looking at Loretta's 100 or so game sac attempts over a 15 year stretch.

When you construct a model, too much granularity will destroy your sample size, so you need to make estimates. Having 200 different speed ratings, the kind that might give a 199 to Carlos Gomez and a 198 to Ichiro, is overkill. I know MGL just uses a 1-5 grade for baserunner speed in his sims. I know because he's written about it a few times, and that rating scale brings back a lot of good memories for an old microleague guy.

My problem, as I suspect mgl's, is when Justice says: "That's the problem in dealing with the stat geeks. They've got so much stuff in their brains they can't get a coherent thought out."

That's kind of rude of Justice, but he's pretty much right. MGL's post was almost stereotypical. Justice asked for a glass of milk and MGL was telling him how to breed dairy cows.

Simply not true. MGL is not a professor, grammar school teacher, advocate, parent, babysitter, or apologist when he is posting. He does not have a duty, and if he did, he isn't and shouldn't take that duty to reach an audience completely unfamiliar with analysis.

This is pretty ironic coming from you, BL, given the way you attack Voros and the DIPS crew for presenting their work.

Roy/58: it may be true that coherency of stat geek is not their forte. But, that it may be true to him doesn't mean he gets to say it! As you say, it was rude. So, we agree here.

As for MGL giving too much information, that was his right, to the benefit of whoever read his post. MGL wasn't necessarily posting only for Justice's benefit. Again, is that a problem? It's not like they were in a dialogue (at that point). MGL gave solid information.

That it was not appreciated doesn't mean that one gets the green light for rudeness and insult.

***

Frida/59: I will suspect that Backlasher will make the same distinction that I make, regarding content and tone. "Tone" is fairly subjective (unless its downright nasty), but "Content" is what we truly debate. My guess is that BL will take exception to the content/theory of DIPS as he sees it presented.

Simply not true. MGL is not a professor, grammar school teacher, advocate, parent, babysitter, or apologist when he is posting. He does not have a duty, and if he did, he isn't and shouldn't take that duty to reach an audience completely unfamiliar with analysis.

In fact, taking that duty would excerbate the problems that others claim --- the length of his post.

1. If he cares about getting his point across, he has a duty to himself to get it across.

This doesn't absolve the receiver of the message. If the receiver of the message cares about receiving the message, he has a duty to himself to make sure he received it as intended. To the extent both parties want the communication to take place successfully, they have the duty to themselves and each other to make it work. But I was addressing the perspective of the sender, not the receiver; and for the sender's perspective it's not successful unless the sender has done what is necessary to get the point across.

If we're talking about any of the roles you cited, then there's explicit responsibility to others as well. I suppose it could be argued that MGL adopted for himself the role of teacher, and thereby had adopted that kind of responsibility to others; but I don't know that, and I'm not going down that route.

2. It doesn't strike me that he was trying to reach an audience "completely unfamiliar with analysis". Rather, it strikes me that he was trying to reach an audience, period. He doesn't seem keen (here or in other instances) on trying to improve his answers when his point is not received by what would be considered the target audience. He seems content with the notion that if someone got his point, there's no problem with his communication abilities. By that measure, any success with a sacrifice bunt justifies all sacrifice bunts.

All of the above is how it seems to me; I could be wrong. I'm certainly not about to suggest any of this has to do with ego, or arrogance, or whatever. I don't know the guy.

3. It wouldn't exacerbate the length of his post. I think I've demonstrated pretty well that his message could be conveyed in a more coherent and much less verbose manner.

village/61: your message was more coherent, but less informative. Sometimes that's good, sometimes not. For a blog post, your way is almost always better. Basically, MGL made an "A-" post, which you can turn into an "A+" using village/41's way (for a blog post). It wasn't a "D" or "F" post by any means.

Regardless, up until Justice made his rude insult, I see no issue here. Do you?

Actually, a letter grade would suffice for purposes of making a model... I know MGL just uses a 1-5 grade for baserunner speed in his sims.

MGL is a bright guy so I'm sure he has some kind of good model there. But let's not then pretend his model seriously relates to what he claims are the seven key points:

1) make out and move the runners over.
2) make out and not move the runners over.
3) hit into a DP.
4) strike out.
5) walk.
5) Get to two strikes, hit away, and do one of the above or below, plus get an extra base hit.
6) reach on a single.
7) reach on an error.

For example, is Mark Loretta a "good" bunter? If we were to take these seven key points seriously we would have to have at least three different letter grades for Loretta's bunting ability (putting the bunt in play before getting to 2 strikes, moving the runners over when getting the bunt in play, avoiding the double play when you put the bunt in play). Lets be serious here. I can't help but feel that the real purpose of MGL putting down these points is to say "Look how smart I am, as I have thought about this issue and you haven't" - which is an attitude that comes across more than once in his work. The thing is that it is to a large extent justified - MGL really is cleverer than most people, certainly far cleverer than me - but it's still going to piss people off. Justice is well within his rights to retort "Everyone knows there are eight thousand possibilities."

What's more, once you start moving away from a robust stastistical analysis of the problem, you have to demonstrate that your model works better than other people's models. For example, in the comments on the linked thread, MGL states

If you (a manager) use(s) [my model], you will gain 90% of what you would gain if you did a thorough analysis of every strategic choice in a game, AND you would do significantly better than the average manager can and does do.

If you're going to make a claim like that, please supply some evidence.

I think it possible to come up with a reasonable guess of whether or not someone should bunt. I did this in analyzing Angel bunts in 2006. It's an estimate, of course, but I think reasonable. And it couldn't have been done without The Book, btw, both MGL's bunt chapter in it and Tango's writings on the associated blog.

I think talking about the actual issue of bunting and strategy is a lot more interesting than ######## about MGL's personality or whatever. Sure, there are times when he comes across as less-than-ideal, but that's true of all of us.

This is pretty ironic coming from you, BL, given the way you attack Voros and the DIPS crew for presenting their work.

See Tango's comment and reread my post. First, mgl's content is usually pretty solid. Second, when you disagree with what he says, he will engage you in conversation if you talk about the points he made.

If he cares about getting his point across, he has a duty to himself to get it across...I suppose it could be argued that MGL adopted for himself the role of teacher, and thereby had adopted that kind of responsibility to others;

No and no. You seem to be taking yourself, or some other member of a group that might read his posts and presuming that he has accepted some duty to himself (or that person, or that reprentative audience). I doubt he gives a flying #### about that audience. He isn't writing for Sports Illustrated. He likely cares about another audience or is talking to another specific person when you are reading those posts.

Right now and in my previous post, I am talking to you. Anyone that can follow the conversation is able to join it, but I don't take a duty to explain it to anyone beyond you. That is the nature of conversation, and this blogging is an outreach of conversation.

Rather, it strikes me that he was trying to reach an audience, period.

And he is always able to reach an audience, period.

by what would be considered the target audience.

That is the problem. You are trying to define a target audience for him.

He seems content with the notion that if someone got his point, there's no problem with his communication abilities.

That is making a pretty broad read on someone's objectives. That is especially true, when if the audience is people like Tango, Emeigh, etc. I doubt they have a problem with being reached by MGL.

By that measure, any success with a sacrifice bunt justifies all sacrifice bunts.


That is logically incorrect, and rhetorically unfair. First, the plausibility of a statement is not the same as the truth of a statement, and mgl has not defined it. Second, you set up that zinger by imparting motive on someone that does not exist.

It wouldn't exacerbate the length of his post. I think I've demonstrated pretty well that his message could be conveyed in a more coherent and much less verbose manner.


If you are his editor, you might get it more consise to the target audience that you have chosen, but I doubt the target audience is anyone how might read this board.

Alou/63: The 7 "magic" points is straight out of his chapter, and yes, you would have to have a grade for each of those things. But, you already have a mental idea of each of those things anyway. All MGL's model needs is to hear those observations from you (with whatever uncertainty level you have for each of those 7 magic points). You provide the inputs, and MGL's engine will analyze your own data.

This is no different than any financial engine telling you how to allocate your money for the next 50 years. You set the values, you decide the parameters, and the engine will do the hard work for you. You could do it yourself, of course. The *only* thing a computer program buys you is time. That's its value.

As for the other claim that MGL needs to support: most people thought that he said too much, but now you are saying he didn't say enough!

Can I address something not really related to MGL's persona?

I have a problem with this:

Q: Should Cecil Cooper have had Mark Loretta attempt a bunt in that situation?

A: We'll never really know for certain.

That's the start of village/44's summary of MGL's post. And the problem I have is that "I don't know" is not an acceptable answer for Cecil Cooper. He faced a binary decision: bunt/not bunt. The question, even for analysis with hindsight, should be whether he made the right decision given the information available to him at that particular moment. A manager can't send the "I don't know" sign to his third base coach. He has to decide. Even if the probabilities and expected values of potential outcomes favor one of the two options by only a tiny margin, then that should be his choice, right?

If I understand what MGL is saying, the correct answer is unknowable due to limitations of the data. And I don't see how that's helpful in the context of 1) making baseball decisions or 2) evaluating the decision-makers.

I could have this totally wrong, and maybe it's the 6 am wakeup call combined with the frustration of a 3-hour layover in ATL due to my connecting flight being delayed by 90 minutes talking, but it seems to me that in nearly every binary decision in baseball we can speak of the superior option. It may not be overwhelmingly superior, but "I don't know" and "we can't tell" aren't good answer for the person who has to make the yes/no decision.

He has really put in some hours pouring over data.

Be specific. Maple syrup? Catsup? Tabasco?

37. kevin Posted: April 04, 2008 at 07:47 AM (#2730188)
...


Insight is the beginning of wisdom. . . .


My advice to is 1) put the arrogance BS in cold storage

One hardly knows where to begin . . . .

He does not have a duty, and if he did, he isn't and shouldn't take that duty to reach an audience completely unfamiliar with analysis.

Then why bother posting a 1500 word screed in the first place? If you are going to take the time to dump all that information and analysis on an internet blog, the least you can do is try to make it clear. Of course, we have disagreed before about the importance of clear communication -- then again, as lawyers, maybe we have a vested interest in a world of ambiguous communications over which to argue endlessly.

Great point, Keith. And it helps to demonstrate why the 'traditonalists' often scoff at the saber types. If all that analysis by MGL produced a "we don't know", then why not rely on the instincts and experienc of the manager? Now THAT is time-saving value.

Frida/69, Keith/67: "I don't know" basically means that the manager can go with his gut, without being second-guessed.

For example, this chart I had in 2002 for when to walk Barry Bonds gave clear signals as to when to walk him, when to not, and when the manager could go with this gut.

There's some things that I don't know, and I'll defer to the manager to fill in those holes for those close calls.

When it comes to the bunt, there's alot of close calls. MGL has 20 rules of thumb that you have to remember to make a good decision.

It is true that, in practice, the bunt call from a manager is typically a binary decision, with runners on. This is an inefficient industry standard.

What makes sense, in my view, is three options. Yes, no, and "it's up to the batter". The original notion of the sacrifice bunt was that the batter would first try to reach base, and as a poor second choice if that failed would at least advance the runners. Leaving the choice whether to attempt a bunt up to the batter in most cases would return the sacrifice bunt to its original and sensible purpose, rather than the modern version with the telegraphed bunt ordered by the manager. With a veteran hitter, down by two, and runners on first and second in the sixth, it seems to me that door #3 is the obvious choice.

Great point, Keith. And it helps to demonstrate why the 'traditonalists' often scoff at the saber types. If all that analysis by MGL produced a "we don't know", then why not rely on the instincts and experienc of the manager? Now THAT is time-saving value.

I think it points out when instinct is useful. It is useful when there is no objective way of arriving at a single best answer. When the best the data can say is, "inconclusive" we have to go with out gut. You can argue Cooper was wrong, whatever his decision, but you can't prove he was wrong. It's unlike valuing OBP, which can be conclusively shown to be important. A manager can't just decide he doesn't value the ability to get on-base (rather, he can, but we can prove he's wrong) but he can decide to bunt, even if most of us think it's the wrong decision. Over time, if his gut is wrong most of the time, his boss can use his gut and fire him.

Frida/69: To be clear, while in most specific instances it would be a "we don't know", so let the manager go with his experience, the overall strategy (game theory) still counts. If all the infielders play Ortiz deep and to the right side, then bunting to the left side would be a good thing. The average Ortiz PA will add less than +0.10 runs per PA if he swings away. But, if he can get on base, he gets +0.27 runs (and an out would be a similar -0.27 runs). If he can bunt his way on and be safe more than half the time, he should bunt. He should keep bunting until the 3B gets back to his spot. Then he can start hitting away.

As long as the opponents take a reasonable approach to their fielding alignment (play somewhat deeper for the great hitters, somewhat shallow for the bad ones), you can set it up so that the choice of whether to bunt or not (with no one on base) becomes a "I don't know" play. But, if the fielding alignment is skewed, then you will get a clear "bunt, I can't believe they're giving you this much room" or "do not bunt, they're right on top of you".

Btw, Hawk/64 has a fantastic piece on analyzing the bunts, along with his followup. If you are going to spend 10 minutes reading something, ignore most of this thread and all of my posts, and read his posts.

I agree with 71 almost entirely-the hitter almost certainly has the best assessment of the ``true'' probabilities of each outcome from a bunt in this situation, and a ``veteran'' (i.e. not stupid) player ought to be able to know enough of the correct bunting rules of thumb to do better than a manager using MGLs model (or anyone else's).

PreBeane/75 ... except Adam Laroche.

If you are going to spend 10 minutes reading something, ignore most of this thread and all of my posts, and read his posts.

Now you tell me!

Finally, if MGL were my graduate student, I'd bring him into my office, sit him down, close the door and tell him right from the get-go "Look, you can't keep going around telling everyone how great your work is and how crappy everyone else's is. That's BS. For 2 reasons: it's not up to you to judge how good your work is. That's up to others. And if you believe all your stuff is so great, it prevents you from maintaining a proper critical distance from it. My advice to is 1) put the arrogance BS in cold storage, 2) learn how to write a sentence without putting everyone to sleep, and 3) develop some interpersonal skills. It's a small planet and we all have to live on it."

Really funny.

Kevin, I think your old graduate professor wants to have a little sit-down with you.

That it was not appreciated doesn't mean that one gets the green light for rudeness and insult.

On the Internet, every post carries an implicit green light for rudeness and insult - a fact that's as true at BTF as it is anyplace else I've been.

When it comes to the bunt, there's alot of close calls. MGL has 20 rules of thumb that you have to remember to make a good decision.

So, even if Antonio Alfonseca becomes a manager, there's no way of keeping track of that many...

That is the problem. You are trying to define a target audience for him.

From MGL: "I honestly thought that I was doing a valuables service to this guy and his blog community."

I'm not trying to define anything; I'm trying to work within the confines of the definition MGL provided us.

I doubt he gives a flying #### about that audience.

Care to reconsider?

Right now and in my previous post, I am talking to you. Anyone that can follow the conversation is able to join it, but I don't take a duty to explain it to anyone beyond you. That is the nature of conversation, and this blogging is an outreach of conversation.

Agreed. My point was that MGL, in his post, took no such duty beyond his initial post, and it's debatable that he even considered the audience initially.

Given the ridicule he got back I can understand why one might not want to take the high road there, but a simple response of, "What part didn't you understand?" would've been a good start.

That is making a pretty broad read on someone's objectives.

I was only saying how it seemed to me, not how it is; and admitted I could be wrong. I'd be pleased to find out I was wrong.

I posted to Justice on his blog:

Richard, I think the only real issue some of us have is when you said: "They've got so much stuff in their brains they can't get a coherent thought out."

Do you continue to stand behind publicly writing this statement, or do you agree that it was entirely misplaced at the time you said it?

He responded:

As for the statement, yes, I do stand by it.

At his request, he wants to come here. (That post has yet to post on his blog.)

I crafted something pretty long winded (ironically) and decided I should probably make it a blog post, so my response is here.

Actually, when I invited you, you said "Shhh... I was hired by the Sox, but I haven't closed the deal yet".

If I understand what MGL is saying, the correct answer is unknowable due to limitations of the data.

Well, here's where the fun part is. Go back to MGL's post and find the part where he actually said, or even suggested, the data is too limited, or that the answer is unknowable.

.

.

.

It's not there.

Earlier in this thread, Walt Davis suggested the data's too thin. In post 44, I suggested it was part of a more concise answer MGL could give. I inferred it from MGL's post simply because, after asserting over and over that data, not opinion, provides the answer, MGL does not provide the data-driven answer. Were there a data-driven answer, and were MGL really trying to do a valuable service to Justice and his blog patrons, I assume that somewhere in his post I'd find it. Is that so wrong to assume?

MGL's post lacks the data-driven answer, and lacks an indication of whether there is enough data for Cecil Cooper to find the correct answer for his given dilemma. But it does say that opinion - the only thing left absent the data-driven answer - is irrelevant. This is coherent?

MGL spends a lot of time telling us who doesn't know the answer; but he doesn't give the answer. He pretty much reduces the field of people who could possibly know the answer to just himself - and then doesn't deliver. It's one thing to say, "I've studied this a lot more than most people," but it's completely another thing to tell an audience, "There's no way possible that you know more than I do." It's even worse not to give a relevant answer after that.

I don't believe for a minute that MGL intended his post to read that way. But to a new audience it reads as arrogant, self-serving, and rude. "Incoherent" is a compliment.

As someone who's interacted with MGL via email, I'll agree with Walt upthread and say that I don't find him to be an ####### or overly arrogant or overbearing at all.

I actually find him to have one of the sharpest minds in terms of analyzing baseball, whether it's decisions on the field or in the front office, on the Internet.

I think what hurts him is precisely that he is good at breaking down a situation in terms of things he knows based on research, things that he doesn't yet know that could be known/approximated, and things that cannot be known. This is in fact an exhaustive listing of all the various things that can go into a decision.

So when he speaks definitively about things, he's not speaking definitively about the things he knows, but his framework for analysis, which I generally find to be both rigorous and more or less correct, exactly because he doesn't just include the things that he knows but the things that he doesn't and does his best to scope the things he doesn't know.

Honestly, I don't think there's a more useful way to structure or think about complex decision-making situations, and he's better at it than most of the sabermetrics work I see out there.

I think where people run into trouble is when they take his posts personally, which is absolutely not the spirit in which they are intended. I agree that it can be hard not to, but if people could manage to take that step, I think they would get a lot more out of his analysis.

Ditto 86.

Reading mgl's post, I don't think he comes off as abrasive. Sure, he comes off as arrogant, but I think we can all agree that the vast majority of people who spend as much time on a subject as he has on stats will be arrogant about that subject.

I'd rather read an arrogant post that goes into the stats than a post that is nothing but clichés and tautologies.

Is that what I said? My memory is really bad (DAMN YOU SEROQUEL!!!) so I was just hoping I wasn't rude about it. My attitude when I was with the Red Sox was that anything I said could be implied to be something coming straight from the Red Sox front office, so I decided not to say anything ever.

Besides you guys probably did much better without me. I say "probably" because you STILL HAVEN'T SENT ME THE BOOK YOU PROMISED, TOM! :)

If you are going to take the time to dump all that information and analysis on an internet blog, the least you can do is try to make it clear. Of course, we have disagreed before about the importance of clear communication -- then again, as lawyers, maybe we have a vested interest in a world of ambiguous communications over which to argue endlessly.


It is pretty clear to most people. It may not be clear to others, but for many of them it would take large amounts of effort to make it clear.

For instance, in this thread, even in the later posts, people still think that the objections to DIPS is because it was not understood by the recipient.

Honestly, I don't think there's a more useful way to structure or think about complex decision-making situations, and he's better at it than most of the sabermetrics work I see out there.


I agree. I can't see how people would favor:

I've done research on one year's worth of data, and I can say X is true;

over

I've looked at 20 years worth of data; we feel confident about X; but cannot yet prove Y.

Also, just as an aside, I don't think its arrogant to name the work, "The Book". In fact, I think time is better spent by Tango and MGL doing the research than coming up with the name.

I got a kick out of the posters who wriote in to take "Richard" to task fro being rude to MGL:

Mitchel is a very respected baseball researcher, perhaps you should apologize to him.

I, for one, appreciated Mitchel's detailed post/primer on bunting. It was in-depth and informative, while still being relatively easy to understand.

Richard, what is wrong with you? A guy makes a great post on bunting (the discussion at hand), and you proceed to insult him. How about a 'thanks for the effort' rather than a put-down?

Seriously, I think the thread derailed when MGL posted this:

Wow. If you are serious with that statement, you are a real (deleted)..... I'd rather be incoherent.

Thanks for the appreciation for the time I took to write a thoughtful primer on the value of the sac bunt attempt. Clearly you know little about it yourself, despite "checking with stat people" or whatever it is you said you did.

MGL DID NOT have to respond that way, he could have done any number of things,
He could actually have applied or tried to apply some of his analysis points to the specific scenario that was being discussed before he came along
He could have made a little joke/backhanded compliment at Richard's expense.

No, what MGL does, is throw up a few points, assert that everyone else's opinion is immaterial, throw in a few references to his book, and makes assertions like:

1) NO ONE, but someone who has (competently) studied or read about sac bunt attempts and all of its statistical intracacies, can possibly know or figure out when a sac bunt attempt is correct or not ... so everyone can take their "opinions" and well, you know...

and

As it turns out, there are enough singles and ROE's (and walks) when a bunt it attempted to justify a bunt in certain situations... Unfortunately, there are no "opinions" which will help you figure out the right answer.

and later my favorite from MGL:

Do you think that calling someone a "profanity" is any worse than saying someone is "incoherent?" Do you have something in particular against profanity? My long post clearly was not incoherent. If you thought that it was, perhaps something is wrong with your reading comprehension skills.

1: Yes calling someone a profanity is worse than saying someone is incoherent- that's one of the very reasons profanities exist and are used.

2: No individual sentences in your post were incoherent, taken as a whole, in the context of the thread, it was very nearly incoherent. (a flaw of many of my longer posts as well) One of your fans effectively paraphrased a main point of your long post in about 3 lines.

I've done research on one year's worth of data, and I can say X is true;

I keep seeing this from you. Where did you get this idea about DIPS? Who told you this?

It's not true, it's never been true from day freakin' one and I've let it pass by without comment long enough. I did the research originally from 1993-1999, and soon thereafter (very soon, long before the Prospectus article) expanded it from 1901-1999. I repeated it for pitchers who changed organizations from one year to the next. I did a matched pair comparison. I looked at non-pitchers pitching. I looked at career rates and compared it to what could be expected from chance. I looked at the difference in HBIP rates by count. I didn't initially look at knuckleballers and trick delivery guys because I didn't have the data. I didn't look at fly ball rates because I didn't have the data. When I got the data for some of that stuff, I re-did a bunch of things and published it here on DIPS 2.0 (long before Tippett's piece). I had always suspected knuckleballers (and other trick delivery guys) might be exceptions and said so on several occasions early on. I always suspected flyballers might allow fewer hits, but didn't have the data and also assumed (correctly I might add) that what flyballers gained by doing so was likely lost in giving up more extra base hits.

And you wonder why I get angry. My work spanned hundreds of hours and several years and then you come here and make false assertions and do it in a way that suggests you're 100% positive you are right.

If I said something in the past personal to you that was rude or offensive, I apologize. I had been known to do so in the past and I apologize. But on factual matters, please get things correct or at least ask for a clarification. I'm honestly trying to make things different this time around, but I'm not going to say things I don't believe in an attempt to be "humble" and I'm not going to say things are true when they are not just so that I don't appear "arrogant." Ultimately if that means I'm arrogant, I gotta live with that but it isn't going to change.

I don't post anymore on BTF, but few rules should be intractable, so I'll briefly (less than 1500 words) chime in:

BTW, if someone wants to make $50, I'll pay them to research all writings I have ever made on the internet, and compile a list of every attempt at humor I have ever made. If the list is less than 10, I'll donate $1,000 to Kevin's favorite charity and if it is 10 or more, he will donate $1,000. If Kevin turns down the offer, which I'm pretty sure he will, I will ask him to publicly retract his statement that "it is obvious that MGL has no sense of humor whatsoever" AND to write, "I am a buffoon" 100 times. Fair enough?

If I understand what MGL is saying, the correct answer is unknowable due to limitations of the data. And I don't see how that's helpful in the context of 1) making baseball decisions or 2) evaluating the decision-makers.

Not at all what I am saying, or at least thinking/espousing. I am a HUGE proponent of taking research on strategic questions and then boiling then down to a set of simple rules or "rules of thumb" if you will, that anyone can use. That is the whole essence of "The Book" and why we we intersperse in every chapter, "The Book says" boxes.

For example, for sac bunting:

If your pitcher is not an anemic hitter, you should bunt him mostly and occasionally hit him. If he is an above average hitter for a pitcher, you should bunt him about half the time with 0 outs and about 20% of the time with 1 out. If the batter is slow or a poor bunter, you should rarely if ever bunt him and if he is fast and a good bunter, you should bunt him about half the time.

All of these "rules of thumb" can get tweaked by other relevant variables. How much a manager wants to tweak them is up to him, as long as the tweaking is sensible (I hate to say it, but in a sabermetric sense) of course. Lots of these other relevant variables are the so-called "things the manager knows but the analyst may not." Such as whether the third baseman is good at fielding bunts, or perhaps was just hit in the hand by a pitch (of course, there is no reason why the analyst can't know that).

What I try and do is educate the manager (fan) about what is relevant and what is not, give some basic rules of thumb, and then let the manager (fan) make a cogent decision, based on what he knows and what is relevant, and the "weights" of those relevant variables. Things traditionally used by managers/fans to justify bunting or not (and other strategies, like the IBB), such as, "We want to put an early run on the board," or, "We have not been scoring runs lately," (or, "We don't want to let so-and-do neat us") by and large are founded on superstition and ignorance (or lack of understanding), and are therefore not generally particularly fruitful in terms of giving the deciding team the best chance to win the game.

I don't feel like looking up the relevant data for the Loretta situation, but if the team was down by 2 runs or less, there were runners on first and second, and the defense was not especially expecting a bunt (another good rule of thumb is that when the defense is especially anticipating a bunt, a bunt is almost never correct, except perhaps by the speediest of players who are great bunters, like a Pierre, Ichiro or Taveras), AND Loretta has decent speed and is a decent bunter, he SHOULD be bunting some percentage of time, more than 10 and less than 90. Does that answer the question? When you are dealing with game theory, as you MUST when it comes to the sac bunt attempt, there is almost NEVER a you should or should not bunt answer. The answer is almost ALWAYS, the percentage of time that you should bunt is X.

Interestingly, that means that if you are only watching one instance of a certain situation, you can NOT say whether the manager did the right or wrong thing, even if God came down and told you the exact answer as to how often the bunt should have occurred in that exact situation, UNLESS you asked the manager what his strategy was (how often he would bunt if that situation came up 100 times), UNLESS the correct answer from God was 100% bunt or 100% not bunt, which is not usually the case in conventional sacrifice situations.

An example I often give of that "connundrum", which is interesting, I think, is this:

Let's say that you are a new pitching coach for a team, and in your first game, the count is 0-2 on the batter and your pitcher throws a fastball, not right down the middle, but he catches a lot of plate.

After the game, you walk up to the pitcher and start to scold him about that pitch. Before you have a chance to finish, your pitcher, who is a Brian Bannister-type, only smarter, and an expert on game theory (he is a world-class poker player to boot!), says, "Skip, I only throw a fastball over the plate in that situation 1% of the time. You just happened to see that 1% pitch! I have to do that so batters won't completely eliminate a fastball over the plate on an 0-2 count!"

The pitcher would be right (more or less - let's not debate that), of course. In order to criticize a particular decision when game theory is involved such that the correct strategy is to do one thing X percentage of the time and another thing Y percentage of the time (with possibly more than 2 alternatives), you have to either watch a bunch of those same situations, or ask the decision-maker what his intention was!

1008 words! Phew!

It is pretty clear to most people.

Maybe, maybe not. You like to assume.

It may not be clear to others, but for many of them it would take large amounts of effort to make it clear.

Maybe, maybe not. You like to assume.

For instance, in this thread, even in the later posts, people still think that the objections to DIPS is because it was not understood by the recipient.

How many people? And who is to say they are wrong? A lot of people have misquoted it from the get-go, and/or refused to even consider later permutations.

Before you have a chance to finish, your pitcher, who is a Brian Bannister-type, only smarter, and an expert on game theory (he is a world-class poker player to boot!), says, "Skip, I only throw a fastball over the plate in that situation 1% of the time. You just happened to see that 1% pitch! I have to do that so batters won't completely eliminate a fastball over the plate on an 0-2 count!"

If he's so smart, why is he calling the pitching coach "Skip"?

After the game, you walk up to the pitcher and start to scold him about that pitch. Before you have a chance to finish, your pitcher, who is a Brian Bannister-type, only smarter, and an expert on game theory (he is a world-class poker player to boot!), says, "Skip, I only throw a fastball over the plate in that situation 1% of the time. You just happened to see that 1% pitch! I have to do that so batters won't completely eliminate a fastball over the plate on an 0-2 count!"

If he was "a Brian Bannister-type, only smarter" he wouldn't say that (he might think it), he'd say, "damn, the ball got away from me, missed over the plate"...

If Kevin turns down the offer, which I'm pretty sure he will, I will ask him to publicly retract his statement that "it is obvious that MGL has no sense of humor whatsoever" AND to write, "I am a buffoon" 100 times. Fair enough?

There's more than a few posters here who would gladly write, "Kevin is a buffoon" 100 times for free...

"How arrogant do you have to be to call your book, 'The Book.'"

Why is that arrogant? I call my car "the car", and my house "the house" ... maybe he's just not that good with titles.

"If he's so smart, why is he calling the pitching coach 'Skip'?"

Pitching coaches have names too, y'know.

I'm honestly trying to make things different this time around, but I'm not going to say things I don't believe in an attempt to be "humble" and I'm not going to say things are true when they are not just so that I don't appear "arrogant."

Then maybe you can enlighten me.

People are still going around saying: Voros says that pitchers have no control over balls in play.

I think this is dead wrong. I think what you said is that the difference in that skill among most bona fide major league pitchers is so minimal, that other skills (K, BB, HR) are more important.

Which is it? Or is it something else? Or is it so complicated, you are going to have to ask MGL for a 1500 word post to explain it?

There's more than a few posters here who would gladly write, "Kevin is a buffoon" 100 times for free...

Actually, if you add them all up, and accept various synonyms for "buffoon", we are way past 100 by now.

For the week, I mean.

Voros' article on DIPS

1008 words! Phew!
one!

All of these "rules of thumb" can get tweaked by other relevant variables. How much a manager wants to tweak them is up to him, as long as the tweaking is sensible

While what you say is logical, the question is how practical is it? You are providing a template with certain factors, whose weights are undetermined and vary by situation, then creating a concrete instance of the solution might involve more effort than the value of the result. As Emeigh said upthread, most managers have lot of these "variables" wired into their psyche as instincts.
If you are providing a method to the fan/nouveau manager, the packaging is as important as the content ( As a slaving Ph.D student, I should know ). Sometimes making things too esoteric for the sake of complete ( or as close to complete you think you can get ) correctness can turn people off your research. And from what I gathered out of this fracas, thats what happened to you.
You were right in your own sense, but your packaging of your views turned off said fans, as they were looking for a simpler explanation from "people in the know". And as things happen on the internet, such misunderstandings get overblown. Justice was at fault, but in a sense, you are culpable too by virtue of using a machine gun to kill a housefly.