User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
Page rendered in 0.6929 seconds
81 querie(s) executed
|
| |||||||||
Baseball Primer Newsblog — The Best News Links from the Baseball Newsstand Thursday, January 03, 2008Ump Bump: When Did Peter Gammons Become So Crotchety?
Repoz
Posted: January 03, 2008 at 08:39 PM | 64 comment(s)
Related News: General, Sabermetrics, Boston, Media, Online |
My BookmarksYou must be logged in to view your Bookmarks. Hot TopicsNewsblog: Madden: Omar Minaya's Mets have issues with injuries and inside the clubhouse (5 - 4:32pm, Jul 05) Last: Darren Newsblog: tampabay.com: Tampa Bay Rays minor-league affiliate's Ladies Night promotion causing a stir (22 - 4:26pm, Jul 05) Last: bob gaj Newsblog: Plain Dealer/Pluto: Matt LaPorta is still in the minors because of Grady Sizemore's cranky elbow (10 - 4:22pm, Jul 05) Last: Walt Davis Newsblog: seattlePI.com: Buhner 'still bleeds Mariners blue'
(16 - 4:19pm, Jul 05) Last: SouthSideRyan |
||||||||
|
About Baseball Think Factory | Write for Us | Copyright © 1996-2008 Baseball Think Factory
User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
|
| Page rendered in 0.6929 seconds | |||||||
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
This is just weird, almost every "sabermetric guy" I know love Raines.
The hard data doesn't prove that the "intangibles" don't help because there is no data on these things.
And your point is?? He didn't say that the numbers proved intangible didn't matter, he said there's no evidence in the numbers for the importance of them.
We can quibble over "hard" but it's wrong to say there's "no data" on "intangibles." Gammons gave you some right there -- he thinks Dustin Pedroia and Tim Raines are "winners." Go to Gammons and ask him what other players he thinks makes teams better than they should be. Then go analyse those players' teams vs. their projections and compare to other teams. That doesn't prove anything but it will tell you whether teammates of Pedroia, Raines and other Gammons-identified winners outperform their projections or not.
Want to do it more formally? Develop a list of players, take it to a number of experts and have each of them rate the players' "intangibles" (you'll have to provide some sort of definition), perform some inter-rater reliability tests and see if you can identify a core of "winners/losers". Then perform the analysis.
Basically do what Bradbury did in his analysis of Leo Mazzone. Controlling for other factors, before being on a Mazzone-coached team, pitchers did X; while coached by Mazzone, they did Y; after leaving Mazzone they did Z. Y was MUCH better than X and Z so we can conclude that, given the other factors controlled for, pitchers did better while coached by Mazzone. Of course that tells you nothing about what Mazzone did nor does it definitively identify the effect as a Mazzone effect (e.g. it could be Cox or the presence of Maddux or the combination ... or of course some other factors not controlled in the model) nor did Bradbury (to my memory) investigate whether the effect was limited to particular types of pitchers.
The personal improvement part of the intangibles will be hard to disentangle. That is, Eckstein might be "gritty" ... but it's because of his grittiness that he can perform at the level he does. There is no "non-gritty" Eckstein performance to compare to. But it's certainly feasible to, say, take a bunch of current minor-leaguers, calculate projections for their performance over the next few years, then ask scouts which players they think will over/under-perform those projections due to "makeup" issues ... then, a few years later, see how well the scouts did in their predictions.
I suspect lots of teams now use psych screening of prospects -- maybe before drafting/signing or just after. These teams should be taking data from their scouts' evaluations and data from the psych tests and seeing whether the psych tests really are helping them identify players who are more likely to make it. You can do similar analysis to give you a better read on the "stats vs scouts" debate -- take players' current MLEs, the current scout ratings on the 5 tools (or whatever system the team uses) and see whether or not the scout ratings add any predictive power above the MLEs (and vice versa).
All of that would be hard work. A lot of that would require access to data that we fans don't have access to. Little if any of it will give you a definitive answer -- the criticism that you didn't measure all the things you needed or the right things or you didn't measure them well enough can always be made. But the notion that these ideas can't be analyzed statistically is just wrong.
These sorts of questions might also make for an interesting use of Tango's "wisdom of the mob" type studies.
Stats say Ortiz is great? Stats are good. Stats say Jeter can't field? Anyone who believes those stats is an idiot.
That's what's so weird about the Raines quote. The stats reinforce his argument but he still calls them stupid.
Want to know about winners? (Dustin) Pedroia gave up his scholarship at Arizona State to free up money to sign a much-needed pitcher, so when the Sun Devils reached the College World Series, coaches and players had “DP” on their caps in honor of their leader who never got to Omaha.
This most likely means they should have had "DP's Mom & Dad" on their caps.
The Raines quote is somewhat reminiscent of those who confuse "hit 320 against lefties in day games last year" with "sabermetrics."
And to put a twist on an old favorite: I'm tired of all these statheads who hate Tim Raines and Albert Belle.
And Mike Crudale.
Actually, isn't that the definition of "intangible"? If you can measure it, then it's "tangible" by definition, isn't it?
I was being nitpicky but my point is that he misspoke. He wrote, "None of the hard data proves that these things help in any way shape or form," when really should have said "there is no quantitative data to prove or disprove if the 'intangibles' affect performance." There's a subtle but important difference.
I acually meant to say there is no quantitative data on most of the "intangibles".
Mike Flanagan is a big fan of psych tests. He used Adam Loewen's high score on such a test as a rationale to pony up the money to sign him.
That's interesting idea although I suspect it'll be a highly unreliable, which an inter-rater reliability test would show. For example, if you ask 50 scouts to rate player's X "grittyness" on a scale of 1 to 10, you give them a clear conceptualization of what you mean by grittyness, I suspect you would receive a good bit of variation in the data. For things like grittyness, I prefer more qualitative evaluations rather than trying to a player X down to a number, but of course you can't such data and run them in a regression model.
And let me add that if you did develop a list of players, take it to a number of experts etc., I suspect there will be a small pool of players where the experts shared a high degree of similiarity. For example, the experts would almost uniformly agree that player X and Y are gritty while players A and B are not gritty. It's all those in cases in the middle that would cause a problem. Some scouts would say player X is very gritty, others would say player X isn't gritty at all, you get the drift.
and
The Raines quote is somewhat reminiscent of those who confuse "hit 320 against lefties in day games last year" with "sabermetrics."
I think you are both missing the point of the quote. I recall that article. Gammons was not talking about how Raines made the teams better with his speed or his on base ability. He was talking about intangibles -- how his upbeat personality and attitude positively affected the people around him.
I love the guy, as I loved my dad when he railed against whatever loud music I had playing, but he's fighting to keep his legacy.
Well, yes, but most "intangibles" aren't intangible. Gammons is measuring something to come to his conclusion that Pedroia is a winner. He might not be able to express that measurement any better than to categorize players as winners/losers/normals, but it's still measurement.
I acually meant to say there is no quantitative data on most of the "intangibles".
I would disagree. "Categorical" data (which is what "Pedroia is a winner" is) is easily quantifiable even if it's nothing more precise than a 0/1 measure. Asking whether teams with identified winners on them perform better than expected is no different than asking whether, after controlling for human capital,etc., African-Americans (self-identified using unknown and imprecise personal definitions of "race" or "ethnicity") earn more/less than whites.
Now you might argue that "winner" is so vague, Gammons himself would be so inconsistent in how he defines it and labels players, etc. that it's useless to try to quantify it ... and I wouldn't necessarily disagree but I'd also aruge it's then at least as useless in its un-quantified state.
"It can't be measured" usually means "I don't know why or I can't explain why I think this or maybe even I can't explain what I mean by it, so it's easier if I call it unmeasurable than to blame my own inadequacies." :-)
But "Peter Gammons says Pedroia is a winner, is Peter Gammons correct?" is a quantifiably testable hypothesis ... as long as he defines what he means by "winner" in a measurable way (e.g. his team will outperform expectations). If he's like many of us, he won't really be able to explain what he means by winner (qualitatively or quantitatively) beyond things like "well, y'know, he'll always come out on top." "What does that mean?" "I don't know, he's a good kid, people like him, they'll go out of their way to help him." "What does that have to do with baseball?" "Well, it just does." "Well, if you can't tell me how it impacts baseball, what good does this information do me?" "Stathead punk!"
I don't mean to oversell this. We're never going to be talking about grittiness scores of 72 or career patterns of grittiness or the transition from youthful enthusiasm to veteran presence.
But especially in baseball, where the outcomes (if not the inputs) that everyone is interested in (wins, runs, etc.) are all quantified with little/no measurement error, it is pretty easy to test qualitative hypotheses.
That's interesting idea although I suspect it'll be a highly unreliable, which an inter-rater reliability test would show.
Which would be what I'd expect and is an interesting result in itself. If 'certified baseball experts' can't agree on who is/isn't a winner, gritty, whatever then (1) there's no reason for us to believe them when they claim someone is a winner and (2) they shouldn't be surprised that the rest of us just can't see that X is a winner.
But yes then it becomes about testing winner/loser ratings expert by expert to see if we can find some that actually can identify them with reasonable accuracy. That would likely be a long, pointless process.
And let me add that if you did develop a list of players, take it to a number of experts etc., I suspect there will be a small pool of players where the experts shared a high degree of similiarity. For example, the experts would almost uniformly agree that player X and Y are gritty while players A and B are not gritty. It's all those in cases in the middle that would cause a problem.
But that's not a problem, that's pretty much exactly what multi-rater measures are designed to do. If you got super-high levels of agreement on everyone, then there's no point using multiple raters. It's the same concept, on a smaller scale, as the wisdom of the mob -- there will be a set where everyone is pretty much in agreement and another set where things bounce around more but (you assume and test if you can) that bouncing is essentially random such that the summed score comes out correct on average.
It's really only a problem if that "middle" group is being rated all over your 1-10 scale (assuming a 1-10 scale). But in practice, that middle group will mostly be rated between, say, 4 and 7 -- i.e. folks are generally in agreement that they aren't winners or losers. Or the inter-rater reliability will be quite low. So either your raters are in reasonable agreement with one another (high inter-rater reliability), in which case you have confidence they're all measuring the same thing and your scale is valid, or they aren't (low inter-rater reliability) which means you don't think you're measuring something valid and/or they're measuring it badly.
The main problem, seems to me, would be that all of these writers/fans/etc. will have influenced one another before we get there -- i.e. the ratings aren't independent. Once Gammons calls Pedroia a winner and tells his silly story about "his" great sacrifice of his tuition money (which he probably got from a Red Sox PR hack) then it'll get told by broadcasters and other writers. Basically it wouldn't surprise that once one writer publicly labels a player a winner, pretty much every writer will agree.
For things like grittyness, I prefer more qualitative evaluations rather than trying to a player X down to a number, but of course you can't such data and run them in a regression model.
Of course you can, this is done all the time. For that qualitative evaluation to do you any good, at some point you have to decide you are/are not going to give that player extra credit/demerit for his grittiness. You might even decide (in an arbitrary and vague fashion probably) that you're going to give some players lots of extra credit/demerit, others a little, and others none at all. At some point, whether you express that decision in words or numbers, it's going to be expressed in a way that is going to be pretty easy to quantify and you can test whether the conclusions you came to worked out or not.
I don't want to oversell this. That won't tell you whether grittiness matters, it tells you whether your conclusions based on the qualitative evaluations had predictive power. If they don't, the problem could be your conclusions, the original evaluations, that grittiness doesn't matter or that there are other unmeasured factors counter-acting the grittiness. But it's a start and it's a damn sight better than "I can't possibly test my conclusions so I'm going to stick with them."
Often the issue on testing qualitative data is more around whether you can quantitatively measure the outcome. Are people who say they are "in good health" actually in "good health". Kinda hard to answer. Are people who say they are "in good health" less likely to be admitted to the hospital in the next year (or 5 years or whatever)? Easy (expensive but easy) to answer. In baseball, all of us only care about quantified outcomes.
I think you are both missing the point of the quote.
Not really. I think we are noting that it's very odd to single out sabermetricians for criticism when it comes to Raines. Yes, I got Gammons' meaning. But, if Gammons actually understands the sabermetric view of Raines, his meaning becomes "those damn sabermetricians think Raines was an outstanding player worthy of first ballot induction into the HoF but the stupid bastards don't realise he was also a team leader." Gee, sorry.
For his point about Raines, seems he should direct it to the folks holding Raines' cocaine habit (apparently very short-lived) against him, not sabermetricians.
My point was that the only reason I can think that "sabermetricians" would deserve a shot is if he's confused sabermetricians with the writers and fans who are pointing out his lack of HR and other key counting stats. But those aren't sabermetricians even if their arguments are based on stats.
It's just very odd. Usually character is dragged out to support an argument that a player is better than their numbers. But he singled out a group which is already saying that his numbers put him in the HoF so why is he singling out that group for "overlooking" Raines' leadership?
And, c'mon, we have to think Raines is great -- we're constantly pointing out he was much better than Jim Rice and Rice is clearly an HoFer. :-)
Sure, you can operationalize anything to make it a dummy variable. Whether it's meaingingful valid and reliable is a different story.
But that's not a problem, that's pretty much exactly what multi-rater measures are designed to do. If you got super-high levels of agreement on everyone, then there's no point using multiple raters.
Well, I think you can't tell if here super high agreement until you run a multi-rater measure. A multi-rater measure, I can't remember of the name of the exact statistic off the top of my head, is just suppose to tell you if there high coder reliability.
It's really only a problem if that "middle" group is being rated all over your 1-10 scale (assuming a 1-10 scale).
Well yeah, that's my point.
But in practice, that middle group will mostly be rated between, say, 4 and 7 -- i.e. folks are generally in agreement that they aren't winners or losers.
I think it depends on the subtopic you're looking at.
Of course you can, this is done all the time. For that qualitative evaluation to do you any good, at some point you have to decide you are/are not going to give that player extra credit/demerit for his grittiness. You might even decide (in an arbitrary and vague fashion probably) that you're going to give some players lots of extra credit/demerit, others a little, and others none at all. At some point, whether you express that decision in words or numbers, it's going to be expressed in a way that is going to be pretty easy to quantify and you can test whether the conclusions you came to worked out or not.
I guess, like I said, I'm not sure what we're arguing about here. You can put everything heurestically in terms quantifiable varialbes or there are other non-quantitative methods of analysis.
Often the issue on testing qualitative data is more around whether you can quantitatively measure the outcome.
A lot of qualitative work isn't concerned about measureable outcomes, or as Andrew Abbott puts it, isn't stuck in the world of variables.
Well, yes, but most "intangibles" aren't intangible. Gammons is measuring something to come to his conclusion that Pedroia is a winner.
Categorically quantify the rinnnnnnggg.
Gammons would have told Neyer to get his nose out of a spreadsheet, but he's in no position to tempt fate by referring to somebody else slumped facefirst at their laptop.
Gammons has always been a hack (see this article, which I believe was written back in '01). Of course, you may find his hackiness charming because...
1. you read him when you were a kid;
2. he hangs around Theo Epstein; or
3. he talks about crappy "hipster" music.
Personally, I enjoy him for the same reason I enjoy Sam Waterston -- I just can't tell whether he acts the way he acts because he's a drunk or because he had a stroke. (My hunch: a little from column A, a little from column B.)
But the guy was, is, and probably always will be a hack.
Gammons is like a southern politician who fought civil rights laws all of his life, but then when the voting rights bill passed and it dawned on him that blacks could vote, he suddenly posed as a "civil rights supporter." (A few of these types even tried to claim that they'd "never been against" voting rights.)
But underneath it he was still plenty p.o.'d that his day of calling the shots was over with, and that his unadulterated racist worldview was yesterday's news, and so he still couldn't resist taking potshots at "civil rights extremists," etc. Just like Gammons can't help dropping lines about "guys in garages."
And in both cases you have to admit: It works. Jesse Helms made a few minor adjustments, hired a few blacks for his staff, and had a long career. And Gammons is still hanging in there himself pretty well. The rest of us might as well be Harvey Gantt.
So he takes a few public potshots. I reiterate that folks here need to think big picture. Gammons has been a friend to the cause when this community didn't have many friends. If he needs to yank a few noses or poke a few eyes to keep the peace with the other side that's fine. We're big kids.
So if Neyer decides to have a public hissy I will be most disappointed. It accomplishes nothing and only confirms prejudices.
The first two quotes are pointing at the tendency of some in the sabermetric community to treat the unquantified as irrelevant. He's spot on with that, though in the first quote he allows one to infer that all sabermetricians are living in garages and hating intangibles. The second quote he's not overgeneralizing, but hyperbolic: although some sabermetricians might deem the human element as immaterial to the overall picture, I don't know any who actually "do not believe players are humans". Either way, the first two quotes are fine to me, because I recognize that he's talking about "some" sabermetricians, and he's right.
The third... Whether Rob Neyer is "obsessed" with Rice or not, I don't know, and I'm not even sure Gammons knows. But the only person who should be getting bent out of shape on that comment is Rob Neyer. Gammons isn't denigrating all sabermetricians there, he's singling out one.
Seriously, I think everyone is reading too much into the cited Gammons quotes.
Angel Berroa?
I didn't grow up reading the guy in Boston, I have only read him since around 2000. He has always seemed like nothing but a gossip columnist to me. People change. It seems like the earlier Peter Gammons made much more worthwhile contributions, but I find very little of value from him, and really seldom read much of anything he writes anymore, unless it's linked here as a topic for discussion.
I always thought he was one of the first guys to start covering trade rumors and talking about (slightly) more advanced stats like OPS. Now Robothal has pretty much cornered market rumors, and there are endless stat blogs that talk about far more advanced analysis. His columns aren't really as unique as they once were, but he's still one of few in the mainstream media who even mentions statistical analysis in baseball. He may not be as hip to the newest, latest stuff, but what he is familiar with I think he knows pretty well.
I imagine he's conflicted by his access. I'm sure he's friends with Jim Rice and reads all of Neyer's work, and gets annoyed when he sees Neyer writing about Rice in anything but a positive light. If one of my friends were up for a big honor, and there was someone I perceived as going out of their way to argue against the honor, I might be a little prickly, too. Letting those feeling get in the way doesn't make Gammons a good reporter or analyst, but as you get older I can understand where those feelings might come from.
He does this a lot, and he gets away with it. who knows why, maybe it is because he is 100, maybe it is because everyone envies his monsterous progressive lens glasses, who knows.
The Mussina thing was sloppy. Nothing more came of it. How can Mussina really defend himself in that position anyway?
The Joba/Arod thing was also bad, and he was the only one in the media reporting it, and he reported it several times, without confirmation.
I wonder if Peter Gammons changed, or if "the pictures got smaller." By that, I mean, Peter Gammons may just be doing the same old crap he's always done. But now (as post 29 says), we have so many other places to get analysis and trade rumors. That makes Gammons look a dinosaur, the baseball equivalent of Irv Kupcinet.
which Joba/Arod thing do you mean?
Same here but that's not because of Gammons, but ESPN Insider.
which Joba/Arod thing do you mean?
He reported, both on TV and in print. That Joba looked over at ARod for the OK to throw at Mr. Slumpbuster intentionally.
Nobody else reported it, no one else confirmed this.
huh, I must have missed that.
anyway, I don't think its any surprise that Gammons carries a lot of water for the Red Sox. I think he's now the baseball equivalent of Bob Woodward - a long-tenured respected journalist who values his connections (and sources) above all else, to the point that he will defend in print whoever will talk to him.
I missed this, could you fill me in SJ?
This is contradicted by the fact that the first person in the media, the very first, who condemned the A-rod announcement during Game 4 was Peter Gammons. And anyone who thinks Boras doesn't have Peter on his "Five" or whatever it's called is seriously mistaken. And until quite recently Scott Boras was deemed one of the most powerful people in all of sports.
When Peter Gammons calls for Bill James to be fired from the Red Sox staff I will take the claims of "anti-sabr" seriously.
I find the age-related sniping disappointing. If folks are going to snark at least be creative.
Maybe Gammo values his connections with the Boston front office more than his connections with Boras? Just a WAG.
Gammons has always struck me as someone willing to use sabermetric methods, but frustrated by people who only use sabermetrics without consulting any other ways of looking at it. He thinks sabermetrics brings something to the table, but doesn't like it when stathead deny the validity of anything else. He's made occassional comments like the ones the blog noted for as long as I've been aware of him, and he's always used ideas like OPS as well. I don't see much change.
My two cents, anyway.
Levski beat me to the punch, but I don't see how this proves anything at all. His ties to the Red Sox FO and his Red Sox fandom plainly trump any other concerns. He thought ARod was stepping on Boston's great victory, and blasted ARod and Boras accordingly.
I'm not SJ, but the article linked in #22 says this:
[Gregory] Lynn uncovered another stunning example of Gammons’ thin sourcing, compounded by blow-off writing that may have sold out someone Gammons was shielding. It came in Gammons’ Jan. 18, 2000, column about the then-raging John Rocker racism controversy.
Here’s what Gammons wrote: "What Rocker said was unforgivable. But the fact is, every team has heard similar diatribes those from small towns across America have about New York City. One of the game’s best pitchers, a sophisticated, intellectual graduate of one of the world’s finest universities, has splattered teammates with similar comments and boasts that he will never play without a no-trade clause to the two New York teams."
Even a casual baseball fan, after consulting his handy pocket Gammons-to-English dictionary, could quickly deduce the probable name of the alleged comment-splatterer. There is a process of elimination I could run through, just to prove the point, but I’ll save the word-count. Mike Mussina graduated from Stanford. If he is not the "sophisticated, intellectual graduate" Gammons has in mind, it sure looks like he is.
for the record, I don't think Gammons' connections issue is limited to the Red Sox FO. I think he does it with all sorts of baseball people, its just with the Red Sox that its the most obvious.
He's friends with Jim Rice. They grew up in baseball together. So why don't you sue him for that?
That makes him a nice guy, but not a good journalist.
I'd say it is more Gammons' background--Gammons covered those 1970s-80s Red Sox teams, saw them, wrote about them, remembers them. Rice was a dramatic, important part of that time, place, and team, so it makes sense that Rice would be "huge"--and a HoFer--in Gammons' mind.
GAMMONS:...there was only one thing I could say about the Red Sox. The Red Sox ...
GAMMONS:...and that's all I have to say about that
Probably a good idea if you want to keep that blood pressure down.
"...and boasts that he will never play without a no-trade clause to the two New York teams ... Even a casual baseball fan, after consulting his handy pocket Gammons-to-English dictionary, could quickly deduce the probable name of the alleged comment-splatterer ... Mike Mussina graduated from Stanford."
he may have guessed wrong (Mussina, after all, signed as a free agent with the Yankees in November 2000), which undercuts his argument significantly.
I figured that it was better late than never to thank you for it, Jon. ;-)
And it makes him cranky.
Hey, you kids! Get off my lawn!
The stroke is an unfortunate possibility as well. Phineas Gage turned into a real jerk after he got that spike through his head.
...and lay off Jim Rice, too!
Indeed. I wonder who it would be then.
Mike Remlinger is from upstate New York, went to Dartmouth and played his second major league year for the Mets. I could see an upstate NY guy not liking the big city for good reasons, the fact that people often think of NY as just the city and forget the rest of the state and the city's differences with the rest of its state, and it could be that Remlinger hated the one year he had to spend in NY.
Now obviously hes not the first guy you think of as "best pitchers in the game" but he was one of the best at what he did when Gammons wrote that having gone 10-1, 84 inning, 2.37 ERA year.
He was also Rocker's teammate at the time.
Now I'm not saying it is Remlinger that said that, he was just one example that came off my head as another accomplished pitcher from academic school.
Furthermore, I wouldn't hold it against any big leaguer to get that in their contract. If I was a big leaguer accomplished enough to make contract demands I might put that in my contract..and I might even throw in Boston and maybe LA as well.
The problem wasn't with the no trade, it was the Rocker-esque rants.
You must be Registered and Logged In to post comments.
<< Back to main