One small defense of the Astros: they were under a lot of pressure to DO SOMETHING about the offense. They could have made better signings than this that wouldn't have impressed their fans or the Astros players (Clemens, Pettitte, etc) that they were really trying to turn the offense around. This is a signing that they can really hold up as making a big splash. That said, it was pretty dumb.

Also, it's laughable to say that MGL calls everyone stupid after he has defended the Matthews deal that everyone else was trashing.

Further evidence that the driving force behind these obscene contracts isn't the Yankees, and hasn't been in years.

One thing I've noticed about Lee that seems to speak well of this signing: his power seems to be improving alongside his selectivity, yes, but he's also cutting down on his strikeouts. He doesn't seem to fall into the "old-player skill" trap of sacrificing swings for power, a la the Giambis. In fact, I'm flat-out impressed that he showcased that kind of power (37 HR) while only striking out 65 times. He's not swinging for the fences, nor is he merely hitting better pitches; he seems to be forcing pitchers to throw him better pitches by hanging around deep in the count. Maybe these fools in Houston haven't fallen for a rainmaker after all?

I watched the guy play this year, and I simply do not understand how he stole 19 bases and was caught only twice.

I think he always had the element of surprise on his side.

You should view that as a failure.

Being wrong isn't "failure" unless it's abot WMD. If it's just baseball stat projections, it's a learning experience.

Since we're getting into old threads, talking about Dye and Hidalgo, here's what Dan had to say about these two signings. Sorry Dan :o)

Being wrong isn't "failure" unless it's abot WMD.

Wrong. Or false. Being "wrong" is a failure when it's your job. Being wrong is a failure when you claim, on the basis of your professional credentials and "science" that "this is one of the worst signings of the offseason. Sorry, you failed. You were wrong, no scare quotes necessary. It's one thing to say, "To the best of my ability, this is what I predict player X will do, and here's a range for that." It's another entirely when you say, "This is stupid. This is the worst signing of the offseason."

Some of you guys are incredible. You'll hang Tim Carver or Joe Morgan for a misstatement or overreaching claim when they're talking for 3 hours during a broadcast, but you can't acknowledge that mgl was wrong, simpliciter? He can't acknowledge that he blew that one? That maybe he should be more respectful of the men who make these decisions w/money on the line and get them right (and wrong)?

Well, Pat Burrell at 2yr/$27M is looking like a bargain now.

Lee just might hit 40+ HRs in his new park over the next two seasons, as no park helps RH hitters more, other than perhaps the Cell. His conditioning might become an issue in latter half of the contract; we'll see.

As far as being "wrong" about any particular projection, I would hope that I or the best forecasters in the world are going to be "wrong" (again, whatever that means) some percentage of the time, otherwise there would be something seriously wrong with the universe.

All I can do is take the data and come up with a projection for total player performance in terms of marginal runs above or below an average player at their defensive position. That is not that difficult to do. At least for me and other professional and amatuer forecasters. Everything else (good/bad signing, stupid/smart team, etc.) follows from that tautologically.


I think where most people find your arguments to be strained is how much faith you place on these projections. I'm a big fan of the sabermetric revolution of the past two decades, but let's not view it as a way of pooh-poohing other approaches of evaluating the game - including scouting and observational experience.

You have a model, you project results from your model. Other people arrive at different conclusions from what your model would state. It doesn't make them stupid or wrong. They're simply basing their decisions on factors that they think are important in the evaluative process.

If your method was free of noise or error, then you have a right to say with such conviction at the expense of others. But most of us, at least I anyway, see that current projection systems are full of noise. It's as limited as quantitative models predicting anything more complicated than a helium atom. It's wrong as often as front office decisions. So if you are "wrong," then GMs are often "wrong."

Amen, brother.

stealfirstbase,

not sure what part of the definition of "sarcasm" you don't understand. maybe some background info will help. in 2004, there were several eyebrow-raising signings, starting with the benson deal, which had led to arguments here that most GMs were idiots and that people on this board could do better. my comment in the dye thread really had very little to do with the dye signing per se; it was just a smart-arse remark to foreshadow the general tone of that thread, as I foresaw it. if you'll note, I really had no comments on the dye deal, one way or another. I don't usually pretend to be smart enough to know which deals are terrible, or excellent... unlike the majority of posters here.

re: the article on dye being pursued by the dbacks. that's news to me. I don't recall reading about it in 2004, when dye signed. as you can see from the article, this info was revealed after the 2005 season. and since i don't really care to follow every (or any) article about the white sox, I didn't read about it. I recall the articles about the Shawn Green trade, and about the offer to Burnitz. pardon my ignorance on every little story that pertains to Jermaine Dye. in the end, even if Dye had signed with AZ after 2004, it wouldn't have made one bit of a difference to the Dbacks in 2005, maybe in 2006.

but I'm glad you've got REAL stories to tell about "White Sox steal Dye from AZ => Count the ringzzz". I can tell you that I couldn't give a rat's arse whether Dye signed with AZ or the Dbacks in December of 2004.

I think where most people find your arguments to be strained is how much faith you place on these projections. I'm a big fan of the sabermetric revolution of the past two decades, but let's not view it as a way of pooh-poohing other approaches of evaluating the game - including scouting and observational experience.

You have a model, you project results from your model. Other people arrive at different conclusions from what your model would state. It doesn't make them stupid or wrong. They're simply basing their decisions on factors that they think are important in the evaluative process.

The problem in this case is that I don't think the Astros are using any other model. This has nothing to do with scouts vs. stats. I don't think they think that Lee is a superstar. They believe they don't have any alternatives. That's a failure of imagination. Their problem isn't that they are using a different model. It's that they aren't using any model at all.

This signing is an absolute atrocity. There are no hidden factors that could justify it. There is no, if Soriano is really a good centerfielder, or, if Mathews really achieved a new level of ability, question that someone in those front offices might have insight into. Lee is what he is. An above average corner outfielder. There is no calculation of value that makes this reasonable. Even by the standards already established this offseason this is the worst deal of the offseason.

I agree with Gaelan and Raskolnikov, if that's cool. In general, MGL's stance is not justified by his quantitative analysis, due to the inherent limitations of such knowledge. In this specific case, I'm confident enough to say that 6/100 for Carlos Lee is dumb.

FOr UZR, I don't do anything special with different parks. I simply adjust the numbers using park factors for each part of the outfield (a bunch of zones combined). Anyway it was just an anecdote. I don't know whether a big or small park significantly affects a good or bad fielder one way or another and I have never seen a study that addresses this issue. I also don't think it is intuitive that a good fielder has more value in a large field and that a bad fielder has less bad value (more value) in a small field. Heck it could be the other way around. Manny is about as terrible at home as compared to all other LF in Fenway as he is on the road after comparing him to other left-fielders in those parks (non-Fenway), after adjusting for defensive HFA. That is not evidence of anything of course.

I put "wrong" in quotes (I have never heard of the word "scarequotes") because I have no idea what it means to be "wrong" about a projection, other than it is a projection that is more than a trifle different from a player's actual performance I guess.

A good forecaster does better, using some statistical or "common sense" test, than some standard. How often a good forecaster gets a player "wrong," and what the implications are for getting a forecast "wrong," I haven't a clue.

Manny is about as terrible at home as compared to all other LF in Fenway as he is on the road after comparing him to other left-fielders in those parks (non-Fenway), after adjusting for defensive HFA.

Was the defensive HFA the same for all parks, out of curiousity?

I had understood from somewhere that you had changed how you viewed Manny's Fenway defense at some point in the past year or so; was that incorrect, or did you adjust the zones at Fenway at some point?

I also don't think it is intuitive that a good fielder has more value in a large field and that a bad fielder has less bad value (more value) in a small field. Heck it could be the other way around.

But why wouldn't that be intuitive? If you put Willie Mays (or Dom DiMiggio or whoever) in LF in Fenway, they still couldn't catch otherwise routine lazy fly balls that end up doinking off the wall (or landing in the Monster seats I suppose) but if you put Greg Luzinski in CF in the old Yankee Stadium, that's going to be an issue. I guess it's possible that the reverse is true, or that the effect is so small as to be basically meaningless, but it seems really intuitive to me

Can something be really intuitive? Probably not. Oh well.

Since we're getting into old threads, talking about Dye and Hidalgo, here's what Dan had to say about these two signings.

Ouch!

I would still, however, say the same thing today with the information I had at the time, unlike a lot of other times where I see where my reasoning was based on a false assumption somewhere.

Hidalgo was 29 and coming off of a bad 2004 season, but he was better in 2003 and 2000 than Dye ever was. When Dye was signed, he hadn't had a better-than-average season since 2000. Dye had a good, not great 2005 season and I maintain that nobody thought Dye would have an MVP-calibre season in him in 2006.

The one thing is that Dye had an injury issue; I don't think it's prudent to do a lot of injury-recovery adjustment in forecasts because we have so little good info on specific injuries, specific recovery, and comps for guys. I'd argue the one semi-exception to this is TJ surgery, where the surgery is nearly identical for everyone who has it and the recovery period (or at least the 'until effective' recovery period) is pretty similar across a decent-sized sample.

So when dinging people on Dye forecasts, it's worth considering how much the failure there is one of all forecasters to be able to identify injury-related decline (which I'd suggest Dye had due to his plant leg being wobbly) from noninjury related decline and then, later, when and to what extent recovery from the injury might occur.

Basically, I think if a guy has a truly significant injury, it's not clear that anyone out there has a model which can do any useful forecasting of the situation. I am thus more concerned with forecasting guys who have healthy seasons and where the statistical model may or may not get it right.

I agree with Gaelan and Raskolnikov, if that's cool.

Booorrrriiinnnggg.

A year ago part of the storyline about the White Sox success was that they changed their offensive approach away from one dimensional sluggers. In service to that change Lee was traded for Scott freakin Podsednik. And there was an implied good riddance to Lee.

And now Lee is a 100M man.

Whatta country!

Lee gives the Astros someone scary in the middle of the lineup to back up Berkman. He was the only guy this offseason who could do that. So was it a bad signing? No. Did they pay too much? Yes, but apparently everyone is going to.

Levski, you should really check out that thread. There's a good bit of speculation in there about who's going to win the Russ Ortiz debry. It seems like the two front runners are Williams and whoever was GM of the D-Backs at the time. I, like you, don't really keep up on news of other teams. I think the previous GM of the D-Backs was a guy named Joe Garagliola(sp) who apparently saddled the team with a bunch of bad contracts. But beyond that, I couldn't tell you anything about him.

I didn't realize that you were being sarcastic in your comment due to the fact that--well--I didn't remember what was going on at the BTF two years ago. But I'm not making sport of you. You thought the deal was OK, or had no opinion of it, and that's better than the majority of everyone else here, myself included. Then again, back then I was reading everything at the BTF pretty uncritically. Now I know nobody here knows what they're talking about, but its still fun to speculate. And some deals are so awful--Pierre, Matthews and Soriano--that people here really do know better than GMs.

Also, the models failed to predict Dye's awesomeness the last two years because there was no evidence of that in his numbers. There's no way anyone could have known what was in store just looking at the numbers. And we didn't have anything else to go by at the time. Jermaine Dye, I guess, was a guy who needed to get healthy, or happy, or out of Oakland, or work harder or something. There's no way we could know that he was going to explode.

Why exactly was it so important to land someone "scary" behind Berkman instead of spreading it around to fill the number of holes they have in their lineup and their rotation? In case you forgot, Berkman managed to pick up 136 RBI last season despite no one being particularly scared of Morgan Ensberg or Luke Scott (Though anyone who fears Carlos Lee should have feared both of them at least as much in '06).

Yes, there is a different HFA for (left) fielders at Fenway.

I fully admit that I have a hard time with projections for players who have had a catastrophic injury, like Dye.

For pitchers, projections are especially problematic for those with injury histories. Two current examples are Kip Wells and Randy Wolf. Both were excellent pitchers before their latest injuries. I have no idea where they stand now. Neither one has pitched well since they returned. Benson was another one. He was an excellent young pitcher before his surgery. After it, he was terrible and has since bounced back to mediocrity. Some pitchers are as good or better after surgery than before. I can't predict it one way or another. Maybe someone else can. It is hard enough to project pitching talent without injuries. BTW, we don't really project performance. That is a misnomer and is why we are going to be "wrong" a significant percentage of the time. All we can ever do is estimate talent based on past, sample performance (with all kinds of adjustments for context). That is it. Future performance is necessarily going to be a sample of a player's true talent and thus subject to normal sampling error. Here is a simple mental game. We are the best forecaster in the world. In fact, we are the perfect forecaster. G-d comes down and tells us exactly what a player's true talent is. AND, that true talent does not change from day to day or from PA to PA. So what happens with our perfect forecast? If being "wrong" is missing by one standard deviation in whatever metric we want to use to measure performance (OPS, RC, lwts, etc.), then we are going to be wrong 1/3 of the time. So what? If that is the best you can do, being a perfect forecaster, it seems kind of silly to talk about being "wrong" in a forecast, as if that necessarily means you "did something wrong." `Remember, we forecast talent, not performance!

About small/large outfields, again, I don't know of any research that suggests that there is a significant advantage for good/bad fielders in small/large outfields. My guess is that it does not make that much difference. IOW, if Lee is a bad fielder, playing in a shallow left field like at MM is not going to make much of a difference. Maybe 1 or 2 runs a year (or none at all). Just a guess though.

MWE asserts it as if it were a given. I would like to know where he gets that from. Perhaps he has done some resarch in that arena that I am not aware of.

This isn't too bad of a deal for the astros. Lee is still productive and will probably thrive with that short porch in left.
In his contract, there is only a no-trade clause for the first four years. So after he loses all the range and speed (and hopefully maintains his power), they will trade him to the AL to be a DH.
Yeah, it's big money but it addresses a big need for this team.

Interesting discussion this has turned into, and I'm sorry to backtrack a bit, but what does it mean to "overpay" for a player? In a value-added sense, i.e. "They overpaid for him when a player of equal VORP could have been had for a third of the cost", sure, no problem. And against other teams, as in, "The Rangers overpaid for A-Rod since no other team bid more than $200 million", I can understand. But in the sense that, "they had to overpay in order to get him", I am not so clear, and I have seen that sort of thing said many times. How is that possible? You basically will ALWAYS have to outbid other teams to secure a free agent, and every off-season teams will be competitive over the most talented players. If a top free agent gets $100m/6yrs, and the bidding was close and fair, then that is the player's worth as the market set it and the winner did not overpay.

Except when the player is only a "top" free agent because the bidding teams aren't studying the right metrics and is actually just Craig Wilson with steady playing time.

and the bidding was close and fair, then that is the player's worth as the market set it and the winner did not overpay.

No, that could just mean that two teams were willing to overpay.

By the end of the year you will see salaries all over the place- some players who can reasonably be expected to outproduce Mathews or Lee or Pierre is going to get far less- for no better reason that that the game of musical chairs may run out on them at the wrong time (from their perspective).

You'll see this every year- you have a very small constrained market with non-fungible commodities being sold (individial players)- in such a "market"- market inefficiencies can be very distorting- some players will get far more than their "worth" and some far less

or another way to put it- you are confusing "worth" or "value" with "price" or "cost"

... which some schools of economics do as well...

I would still, however, say the same thing today with the information I had at the time, unlike a lot of other times where I see where my reasoning was based on a false assumption somewhere.

If you put yourself in 2004, it's very hard to see the silver-lining on Jermaine Dye, and you can't fault people for strongly questioning that move.

That said, I think people tend to overemphasize the dollars on these contracts, particularly on teams with reasonable payroll expenditures. The White Sox signed Dye for two years, so the amount of damage he was going to do to their roster was going to be pretty limited, and he had had some pretty decent years in the past. Even if he had been hurt and performed horribly, he wasn't going to take the entire team down, and there was a chance that he would be pretty good. No one of course saw 2006 coming, but those things happen.

stealfirstbase,

If you read thru many of the transaction oracle threads from 2004, you'll see one comment coming thru over and over: "way too much money". It was practically a must. That, and "what the hell is this idiot ______ [insert name of GM] thinking?" Little did everyone on this board know about the wonderful offseason of 2006.

Joe Garagiola Jr. was a certifiable moron when it came to the MAJORITY of the moves AZ made, but you cannot pin the Russ Ortiz signing on him. I have it from a solid source that Joe Jr. was actually concerned about Russ Ortiz, and it was one of the owners, Ken Kendrick, who pushed to sign Ortiz. It was right after Jerry Colangelo was pushed out and Kendrick wanted to leave his stamp on the franchise... by trading Randy Johnson, the remnant of the old regime, and bringing in his own cool kids, such as Russ Ortiz and Troy Glaus. That was certainly a stroke of genius.

In light of the Soriano and Lee signings, do you think Pat Gillick is having second thoughts about his need to dump Abreu's contract for next to nothing?

Hopefully Gillick regretted it when he realized that the trade was just enough to keep the Phils out of the playoffs.

As someone who watched Carlos play regularly and also fortunate enough to speak to him several times the following I believe to be indisputable:

Carlos is a hard-working player.

Carlos is an intelligent player.

Carlos was determined to maximize his "value" in 2006.

Therefore, he was going to hit.

He was going to stay in the lineup.

Which may explain the decrease in K's and the clearly reduced range in the field. Carlos was not going to risk his season a fly ball.

Do I begrudge Lee's approach?

No. Carlos has a limited shelf life and his offense more then compensated for his defense.

I expect him to stay in the lineup and drive in runs. He won't run into any dumb outs on the bases and may make a circus catch every so often since he won't be playing with his back to the wall as he did most of '06.

"Why exactly was it so important to land someone "scary" behind Berkman instead of spreading it around to fill the number of holes they have in their lineup and their rotation? In case you forgot, Berkman managed to pick up 136 RBI last season despite no one being particularly scared of Morgan Ensberg or Luke Scott (Though anyone who fears Carlos Lee should have feared both of them at least as much in '06)."

Well, what holes? The Astros refuse to move Everett, believing (with some reason) that he's the best defensive shortstop in the game. Ausmus is probably going to sort out based on what Clemens and Pettite, who swear by him, do. And since you like Scott and Ensberg, you are obviously not going to move them. So that leaves a hole in the outfield and at second. Well, the Astros have decided that they want Biggio to get 3000 as an Astro, and they've got Burke, so that's that. Leaving the hole in the outfield. The Astros have essentially dug themselves into a hole regarding the offensive output from shortstop, catcher and second. This is their way of compensating, though I'd add for some reason I think it is possible they deal Taveras.

As for the pitching staff, the assumption is that they get Clemens and Pettite, or they get nothing. But Clemens is better than Zito, and Pettite not too far behind, so if the Astros get those two, and fill in their hole in the outfield, I'm not sure how much more is needed to "spread" out. Getting Lidge a lobotomy won't cost that much. But the only way around the bullpen question is to work out a trade, probably, and I don't see many obvious candidates.

Any free agent signing would hopefully make the team doing the signing better. This might be such a case, in fact it probably is. Of course, it probably only makes them better for a couple-three years, not six.

But a real blockbuster would strike fear into the opponents, would make their fans throw in the towel, like when the Yankees sign ARod and everybody goes, "####, I can't compete with that." This deal is not even remotely close to that. Yet they spent blockbuster money.

Still this deal is about fear--the Astros' own, their fear of falling behind. They had to do something, no matter how tenuous. But the fact is, this will only make the 'Stros more fearful in the coming years.

Carlos is a hard-working player.

Carlos is an intelligent player.

I dispute. I remember sitting in the LF seats in the Cell, by the bullpen (great seats, btw) and having the entire section ask Carlos "How many outs are there, Carlos?" El Caballo was well known for his mental lapses. I once saw him settle under a flyball with two outs and a man on third. Carlos caught the ball at a run and reared back to throw the ball to the infield to stop the guy at third from tagging. He never threw, but it was pretty funny to see him try to recover without looking like a dolt.

The funny thing is, back in those days, you could shout to the players on the field, and some days they couldn't pretend like they didn't hear you. There just weren't enough people in the stands. I remember the Mariners backup RFer getting heckled. There was nothing he could do. He couldn't pretend like he didn't here "Ichiro would've HAD that" chanted at him.

So, in conclusion, I'm glad the Sox missed out on Russ Ortiz, and got Jermaine Dye instead in the 2004 offseason. Sorry, AZ, I wish you no ill will, but better you than us.