THE BOOK--Playing The Percentages In Baseball

I don’t know, it seems like we should add a few more teams to that list if your only criterion is that one significant player on the team was juicing in the past.  We have the Yankees with Giambi, Pettitte, Rodriguez, and Clemens.  The Red Sox had Manny.  The Giants had Bonds.  The Orioles had Tejada and Palmeiro.  For that matter, the A’s at one point had Giambi and Tejada.  Going a little farther back, you’ve got Cubs/Sosa.  And I’m not sure if Gary Matthews, Jr. was good enough for you to consider him significant, but he was implicated in using HGH while with the Rangers.  Eric Gagne was in the Mitchell Report, which would give you a significant player on the Dodgers.  That’s another 4-8 teams where you might suspect more widespread usage based on reports of high-profile players having used PEDs if we’re going by your theory.  Do those teams fit together with the Padres and Indians?

I don’t think this really gets us anywhere, though.  At this point it seems like usage was widespread enough that it’s not going to be worthwhile to try to guess which teams had the most users or the most high-impact users.  Not without a lot more evidence, anyway.

Also, in the case of the Indians, last year their Pythag record was roughly 86-76, which doesn’t seem to suggest that the individual contributions were particularly lacking.  And this year they are 5 games below their Pythag record already, which is still pretty bad, but suggests that the individual performances haven’t been as bad as their raw record would indicate.

Maybe I am misunderstanding mgl’s point but:
The fact that the Padres and Indians are doing significantly worse than their pythag record (and similar more sophisticated estimates) project has nothing to do with the fact that Giles, Hafner or whoever may have juiced in the past.
Pythag etc. take into account the current performance of these players and project the win-loss record of the teams based on performances by their players this season.

vj- I’m pretty sure that MGL knows about the basics of Pythag… he’s pretty good at the whole “baseball analysis” thing .

If I am understanding him correctly, he is saying that the Padres and the Indians have vastly underperformed their preseasons projections.  Projections are made up of the past performance of players, so if a player did steroids in the past, his projection would be inflated.  However, if he were to go off steroids, he would likely under perform his inflated projection. 

If a team had a lot of players who were coming off of steroids, they would likely under perform for that same reason.  MGL is saying that because Giles and Hafner *probably* used steroids in the past, some of their teammates likely used them as well.

I just realized that you were probably talking to ubelmann, so feel free to disregard my condescension.

The Indians are hitting the ball just fine.  They just don’t have any pitching after Cliff Lee.

Giles is old.  I suspect he was a roider, but probably went off about the time he went to San Diego, and MLB started their testing policy.  That theory coincides with him dropping from a 35 homer man to a 15 homer man.  If he had been beating the steroid tests from 2005-08, why stop now?  I think his being 38 is a more likely culprit.

Also, the ballpark hurts his HR total, but doesn’t explain the magnitude of the drop.  In Pittsburgh he hit more homers every year on the road than he does now in total.

Yes, I am not talking about Pythag record versus actual record.  As #3 says, I am talking about record as implied by each player’s pre-season projection pro-rated by their playing time.  I wrote two articles in the THT Annual about this.  SD was off the charts last year and probably so this year as well.  My principal point was that the manager should be held accountable for this enormous anomaly.  I threw in the steroids thing as a bonus.

#1, that is true.  And if any of those teams had the same profile (vastly under-performing) then they would be on the list too.

As I said, it is a weak theory.  I don’t even know that the players overall on SD and CLE have in 08 and 09 vastly underperformed their projections other than Giles and Hafner.  But feel free to support or rebut.

As an Indians fan, I agree with Rally...the hitting is fine.  The pitching has been horrible, and it’s tough to chalk that up to steroids, as Carmona has been unable to throw strikes for the last two years, Perez has been unable to throw strikes, a couple of the back end of the rotation guys (who were not great by any means, but did win the jobs over the guys who replaced them) are injured,… I don’t really see any performance-related declines that are easily chalked up to steroids.

The Indians went into this season counting on Carl Pavano as their #3 starter.  I thought they would get away with it, but they didn’t, and I think it sums it up in a nutshell.

Also, Pronk hasn’t been half-bad this year...he’s hitting 289/389/556, which is better than he did in ‘07.  Obviously it’s nowhere near his 05-06 production, and he hasn’t played everyday because his shoulder is still bothering him.  But he is 32 now and he always had old player skills, so I’m far from convinced that natural aging is not the culprit rather than PED use.

More on Indians hitters exceeding/missing their projections (I didn’t formally look at any projections, just going off gut):

Martinez: power sapped, likely by elbow injury in 08.  Back to expected level in 09 although pitiful the last few weeks

Garko: he is what he is

Cabrera: struggled last year and got sent down, but his projection probably wasn’t that high to begin with.  Has been productive since being recalled last year

Peralta: power outage in first 2 months this year, but coming back to his expected level (and did in 08 about what he did in 07)

Francisco: below projection, but he should have never profiled as much more than a 4th outfielder anyway, but for some reason the Indians don’t feel it’s important to have a left fielder who can hit

Sizemore: met expectations last year, way down this year--likely culprit is elbow injury, not drugs

Choo: if anything, has hit better than could have been expected the last 2 years

The one team I really think got a boost from PED’s was the A’s during there magical run that made Beane look like genius.  The Giambi’s and other completely over preformed and are shell’s of themsleves once they had to quit.  Now the team is trying to use the same model, but the under lying bust in isn’t there.

I think the topic is still causing people’s emotions to run high.  I would love for a group of people to sit down and look at the effects and keep emotions out of it.

I don’t think Giles is the face of this theory for SD.  I look at his hitting from 2004-08, and it seems pretty standard (ups-and-downs).  His 2008 season should have disspelled the notion that he’s found a lower level of performance vis-a-vis this theory, especially at the age of 37.

As for Hafner, he was injured in 2006 and 2008 wasn’t he? 

I don’t have necessarily a problem with a conspiracy-nut theory, as MGL is calling it.  I don’t think putting some poster boys on it helps though, especially since you can find reasonable reasons not to put them on it.  This basically would derail the theory before it has a chance to get started.

In any case, my conspiracy-nut theory is that pitchers are the ones most affected by PED.  I think it’s easier to say what adding 3mph on a pitch would do, than saying what someone who can bench-press an extra 50lbs would do.

Other people have pretty much already said this, but based on my quickie analysis comparing the Tribe’s actual performance to the PECOTA projections for all the players, the offense is somewhat outplaying its projections while the pitching is underperforming greatly. I calculate that the pitchers have given up about 65 more earned runs than they were projected to. The biggest underperformers (accounting for 60 of those runs):

Carmona -19
R.Perez -16
Huff -14
A.Reyes -6
Wood -6

Huff’s a rookie and Reyes and Wood are fairly new to the franchise, so it’s hard to imagine that they were caught up in a team-wide steroid party years ago.

Carmona and Perez both debuted in 2006. I have no idea whether they’ve used PEDs, but based on differences in veteran status, ethnicity, and position, they don’t seem like the most likely candidates to be hanging out with Hafner in the gym.

Don’t forget. It’s not just steroids that are banned now. It’s stimulants as well, and they’ve been around for a lot longer than steroids, so it’s harder to tell how much effect the ban on them will have. I suspect that the fact that they are no longer freely available will have the biggest effect on older players, since they are more likely to have days on which they are just plain fatigued and don’t feel like they have gas in the tank.

"The Indians went into this season counting on Carl Pavano as their #3 starter.”

Scary thing is that as bad as he’s been, he’s been their 2nd best starter.

As I said, it was a “wild-ass” theory, if you can even call it a theory, in the first place.

In any case, other than the idea that you should not fire a manager unless you are on the “inside” and you know what goes on in the clubhouse, etc., if you are ever going to fire any manager in baseball for his team under-performing legitimate projections, don’t Wedge and Black, especially the latter, have to be at the top of the list?

I’m interested in the pitchers-more-affected-by-PEDs theory, and I’m wondering if the poster boy is Barry Zito… [Pure speculation.  No accusation.] I haven’t really looked into him too much; was his fastball a faster pitch with Oakland?  Of course, pitchers can lose fastball velocity naturally.

Seems a bit of cherrypicking going on in suggesting the Indians and the Padres. I’m not sure if I can name a team that did not have at least one minor league or major league player connected to steroids at one point or another. Also, just because a player is a significant i.e. star player does not mean people will take steroids just because he “suggests” they do. If the player is unpopular or if there is a language barrier, it might not happen. Can you imagine, for example, Barry Bonds trying to persuade Jeff Kent to take steroids? Or even one of their fourth outfielders? Maybe someone might smoke pot with Manny, but would they let Manny inject them in the butt with a drug?

I think you simply have to be on the “inside” to make these judgements. If the owner and GM don’t think that these managers are significantly responsible for their teams’ apparent underperformance, and are still appropriate managers for the present team, and choose not to fire them out of knee-jerkedness---well,I support that approach.

To say that they should be fired because their teams have underperformed their preseason projections for a couple years seems to go against what mgl has said in the past about manager evaluation, and the difficulty of ascribing responsibility to them, at least without a very large sample size of games managed.

I think that Bill James many years ago made the most insightful comment about managers comings and goings, something like “I believe that most of the time managers are fired and hired for very good reasons.”

And Bill James should have, and probably would have, if he had written it again, included the term “retained” in his quote

re: Zito

He was never a hard thrower.  His average fastball was around 87 with the A’s (all on Fangraphs), and slipped to 85.8 his last year there.  (It is insane to give that much money to a guy throwing 85 with only so-so control.) He dropped even further, to 84 with the Giants, but is back up to 86.4 this year.

ALL pitchers (on the average) experience drops in velocity starting very early in their MLB careers.  Many pitchers because of the stress on the arm and body, minor and major injuries, tinkering with mechanics, etc., experience considerable drops in velocity (and occasional increases, like Cliff Lee last year and this year).  I don’t think looking at drops in fastball velocity alone as evidence of possible PED use will bear much fruit.  I suppose it could be part of an investigation though.

“I believe that most of the time managers are fired and hired for very good reasons.”

I believe that is one of many things that James says that he pulls out if his a**.  He often gives more deference to teams and team management than they deserve, for whatever reasons, probably because he either wanted to work for them or now works for them of course.  I strongly disagree with that statement, but it is just an opinion on my part and of course it depends on the definition of “most of the time” and “very good reasons.”

“...seems to go against what mgl has said in the past about manager evaluation...”

Not really.  Yes, I have said there is virtually no way of knowing the influence that a manager has on his team’s w/l record and if you could know, it would likely be very small, but…

I also said, in this thread and in other posts, that when a team vastly (REALLY vastly) underperforms (or overperforms, like Mike Scoscia’s Angles over the last few years) for quite some time, I am in favor of canning the manager anyway.  The Padres and to a lesser extent, the Indians, have underperformed to an historical extent. If you read my articles in the 08 and 09 THT Annuals, you will see how/why.

dave/17- I think that using expected vs. actual performance as a rough measure for managerial prowess is okay. 

Think of it like ERA.  We know that a lot of ERA isn’t under a player’s control, but we do know that most good pitchers will have good ERA’s and most bad pitchers will have bad ERA’s.  Furthermore, the bigger the sample size, the more pitchers’ ERA’s will regress to their skill. 

If all we have to go by to judge managers by is expected vs. actual performance, we know that most good managers will help their teams over perform and vice versa.  We may miss some guys who are simply getting lucky or unlucky, but in a 2 or 3 year sample size, like MGL is referring to, it is very likely that the two managers in question are just bad.

Let’s say there is a player who is greatly over-performing his projection.  I’ll call him Ben Zobrist, although the comments that follow are not specific to him.  If we start trying to figure out why he’s over-performing his projection, how many different possible causes are we going to go through before someone suggests it’s because his manager is incompetent?

The first thought of course would be random chance.  Then someone might suggest that he has benefitted from regular playing time.  Someone might suggest that his projection was too low to begin with because he was hitting a lot of line drives and our projection system only considered that in terms of the traditional stat results.  A wild-eyed blogger (but never, ever a responsible professional journalist) might suggest he’s on steroids.  Someone might suggest that he’s learned how to take advantage of his park.  Someone might suggest that he’s vastly improved his plate discipline.  Someone might suggest that he stopped trying to pull everything and learned to take the ball the other way.

I think that a whole bunch of things would be floated before someone says “Joe Maddon is a genius”.  Certainly some of the potential factors COULD be as a result of instruction, but players are certainly capable of making adjustments on their own.  Perhaps it was the batting coach that spurred the change in approach.  Or the guest instructor at spring training.  Or maybe it’s his old college coach who he works out with the winter.  Or…

Furthermore, in the case of a long-time manager like Wedge (he’s been there since 2003), the projections for long-time team members (in the Indians’ case you would have in particular Hafner, Peralta, Lee, Sizemore, Martinez) are based on the performances those players posted under Wedge.  So if Hafner played great in 2006, and poorly in 2009, we conclude that Wedge is doing something wrong.  But if we’re accepting the premise that the manager can materially impact a player’s performance vis-a-vis expectations, maybe what actually happened was that Wedge coaxed an unsustainable over-performance out of Hafner in 2006.

I have no problem with under/over performance being considered as AN indicator of managerial performance.  But I think that you can easily get carried away with it too.

On Wedge in particular, I would give you 1-3 odds that he’s not managing the team in 2010.  I think he’s an ordinary manager, and since managers are hired to be fired, as the cliche goes, it’s probably time for him to go.  But I applaud Mark Shapiro for not doing the knee-jerk thing and firing him during the season, and I don’t believe that the Indians would play better the rest of the season if Joel Skinner, Jeff Datz, or Torey Luvullo was the manager. 

I have never understood what bringing in some retread for 2 months of caretaking is supposed to accomplish, unless the current manager is so awful that he is actively harming the future of the team.

I ditto Patriot.

MGL/19, I don’t think B James pulled that comment out of his a$$ at all. I think managers are fired not strictly because their team is underperforming, but because of that PLUS how it is affecting the manager’s day to day performance with the team. If the brass senses that the manager is “losing control” of the team (whatever that exactly means), it becomes appropriate to fire him. I don’t think managers are ever fired for their tactical decisions, etc., although we as analytical observers like to focus on that stuff. It’s almost always a function of the way the manager is perceived to be leading the team (although sometimes even if they think he is still managing well, they might still make a change just for the sake of change, or to satisfy the fans, etc.)

If Bill James comments were true, then we would see managers get fired all the time when the team is not doing particularly poorly.  The fact that managers almost always get eventually fired when a team is doing dismally and almost never get fired when a team is doing fantastic suggests (strongly) that James statement is pure BS.  I don’t see any other way to look at it.

Patriot, I am NOT talking about the manager potentially being responsible for the underlying performance of the players.  I am talking about the win/loss record of the team, which incorporates a lot of things that the manager has some control over.

I may not have been clear about that before.  If we take each player’s projection (hitting, pitching, defense, baserunning - everything) and prorate it for the amount of playing time for the season, we can come up with a w/l record that the team “should” have had.  That is what I did in those THT articles.  If we compare that number to the team’s actual w/l record we actually have SOME idea as to the manager’s (and everyone else who has something to do with team “leadership") input on the team.  How much, I don’t know.  It depends on the sample size of games of course.

I am contending that after a year or two, if that gap, positive or negative, is extraordinarily large, we have to start talking about “management.” In the case of the Padres, that gap, this year and last year, is unbelievably, extraordinarily large. It is likely that SOMEONE is doing something wrong.  You HAVE to fire Black even if you think he is a wonderful manager (other than the team’s w/l record).  ANY other business would let him go.  It would be like, “The manager of our store on Elm Street seems like a great guy and a great manager.  But somehow, we aren’t selling anything at all in that store, while all the other stores are doing fine.  We don’t know why.” Well, guess what?  You have to replace that store manager. You have to.

Re Dave Smyth’s #23, “I don’t think managers are ever fired for their tactical decisions, etc,” you might want to grab a copy of Seth Mnookin’s Feeding the Monster; check the index under “Henry, John, and Little.”

It is rumored that Little was fired for the Pedro debacle, but my guess is that if any manager is fired for a perceived tactical decision or decisions, there are probably other things that their bosses did not like about them.  While it is generally true that managers do not get fired for specific poor tactical decisions (one reason is that management does not KNOW a poor one from a good one), I have little doubt that when a team is considering retaining or firing a manager, that those kinds of things, among many, come up in their meetings and discussions.  But, as with most things like this, unless you have worked in a front office and have been privy to these kinds of discussions, who knows?

Re MGL’s #26, a book report with some comments:

Mnookin’s end matter ("A Note on Sources and Methodology") explicitly claims that he was in fact privy to first-hand reports of those discussions, saying that his is/was “...the first book in which a writer was given unfettered access to every level of the club.”

Mnookin also has John Henry saying of Grady Little “My feeling was that we were here to win a championship, and I thought that, sooner or later, when it came down to crunch time, he was going to hurt us,” and “There was a total lack of preparation.”

To me, since Henry’s remarks have never (as far as I know) been disavowed, this suggests that maybe the answer to “Who knows?” is “We do.” Little’s bosses may well have fired him on account of serial tactical blunders.

In a pinch, we could call this incorrigible dumnicity “other things that their bosses did not like about them;” then again, maybe we could say “tactical blunder after tactical blunder = strategic incomprehension.”

I’d prefer to keep it in my pants better than BPro’s book [Mind Game], but I do think it’s arguable that in this case management in fact did “KNOW a poor thing from a good one.”

One of the interesting things with Boston and Little is, if the Boston brass, James, Epstein, et al., knew that Little was making tactical error after tactical error, why didn’t they straighten him out well before the Pedro debacle or before he was fired?  If I were operating a team, there is no way I would stand for my manager making more than an occasional tactical error.  First of all, I would be telling him in advance what I wanted him to do or not to do.  And then, I would be constantly evaluating his decisions and discussing them with him. If he is the type of manager that is not receptive to either of those things - does not want the FO interfering with his job - I would not have hired him in the first place.  Does that not go on for teams like Boston?  I know it is often thought of as anathema for GM’s and other FO personnel to interfere with managerial decisions.  I have always thought that that mindset was ridiculous.  The GM is the manager’s boss and he has every right to “interfere” with the manager’s decision- making processes.  And to let an employee do things that are clearly wrong and that you know costs your team wins and money is irresponsible - in fact negligent.  So again, if Boston was so smart in that they knew that Little was making lots of tactical errors and that is why they fired him, why did they allow him to do those things in the first place, and why did they hire him in the first place.  It is not that I am doubting Mnookin’s account (although you have to be pretty naive to automatically believe anything you read in a book). I am simply asking reasonable questions which go along with his account.

Re MGL’s #28: Excellent points throughout.

My best guess is that the FO was (overly?) cowed by the local press; maybe Little put the nail in his own P.R. coffin by leaving Pedro in the playoff game, making it safe to fire the manager.

There also seems to have been considerable emphasis on Expanding The Market (aka Branding), pressure exerted downward from CEO to GM, perhaps resulting in excessive looking over the shoulder when making personnel decisions.

No matter how we slice it, any beatification of the organization (e.g. Mnookin’s book, the BP book) is a bit premature, just as you suggest.