"Michael Young is the best player in the A.L. West" -Rob Neyer, March 2007

Call me silly, but I don't want this guy any closer to a Hall ballot than any of the Underwood Typewriter set.

outside of vlad, who are you thinking is better? don't forget the positional adjustment.

That was an off-the-cuff answer, and Guerrero might have been a better one. But over the last three seasons, Young's totaled 80 Win Shares, Guerrero 78. Now, you might believe Win Shares is bunk, as analysis. But it's pretty hard to argue, all things considered, that Young's not in the same neighborhood as Guerrero.

And if not one of those two guys, then who? Considering Eric Chavez's decline, nobody on the Athletics qualifies for this discussion. And the Mariners? Pass. Which leaves other Angels and Rangers. And with the possible exception of Teixeira -- who wasn't real good in 2006 -- I'm not sure who we're missing. But Halofan, don't worry about that Hall ballot; I'll never have one! -r

Isn't Young a pretty terrible SS according to the Fielding Bible and UZR?

A quick and dirty method, Poor Man's SWLTs, has Teixiera, Swisher, OCab, Matthews, Vlad, Bradley, and Johjima (maybe others) ahead of him in 06. That's just position players. Not sure how they project for the future, though.

Yes, but look around the league. I'm sure more than a few teams would trade mediocre range (though Young had a considerable uptick in his fielding rating last year) and arm strength for his offense. Every team in the AL short of NY, maybe BAL and DET (how far has Tejada declined?) would see Young as a major upgrade.

This is not about a single comment, though, because Neyer's been wrong about players before. But I'd wager a considerable amount that his comments are far more often spot on or near the mark, and he is a good read.

Can you imagine if every valuation statement and prediction you ever made were available for viewing? I shudder at how often I would have been proven the amateur analyst that I am...

Teixiera, Swisher, OCab, Matthews, Vlad, Bradley, and Johjima

Well, part of the connotation of "best player" is that he's proven himself at a certain level for a while. That requirement reduces the above list to Teix, the OC, and Vlad. To take the OC over Michael Young, you have to have great faith that whatever defensive advantage the OC has outweighs the extra 30-50 runs that Young creates every year. That takes a major leap just to bring them even.

And Milton Bradley? He's played 171 games the past two seasons combined. Young plays 160 every year. Win Shares may give too much credit for just showing up, but shouldn't you give some?

"how far has Tejada declined?"

Not that far. But your point is right on, I think Young has a pretty solid hold on 4th best SS in the AL (probably in the top 7 or so (Furcal, Reyes and maybe Ramirez or Rollins) in majors. That's a pretty good player.

Going forward, I'd take Guerrero, Teixiera, and maybe Swisher over him, but that's about it.

And looking back over the past few years, one might rate Ichiro even with Young, especially considering baserunning. Both are solid players when they don't win batting titles, and excellent ones when they do. Ichiro is three years older, of course ...

Bob,

I agree with you; that's why I said I'm not sure how they project. I'd put a few guys ahead of Young but not a whole lot of them. (I was as surprised as you that Bradley scored so well.)

It is dangerous to speak of player's value or talent without fully incorporating their estimated defensive value, to the best of your (or someone else's) ability, regardless of what you think of the various defensive metrics. And saying something like, "Well Young is not the best defensive SS in the game so I'll dock him a few ticks" is just not going to cut it. If you ignore or mischaracteriae defense, you end up greatly overvaluing players like Jeter, Young, Griffey, and Manny, and undervaluing players like Adam Everett, who gets blasted all the time as a "black hole" in the HOU lineup, along with Ausums, when he is in fact almost 3 wins above replacement, including defense and other peripherals.

Michael Young, according to most of the PBP metrics, is almost unequivocably one of the worst fielding SS in baseball. I say "almost unequivocably" because it is based on a large sample of opportunities. Having a "problem" with these PBP metrics, in terms of them accurately describing a player's fielding value is another story. For know, that is all we have to go on, and when I estimate a player's fielding talent, there is quite a bit of regression involved in the sample numbers to make up for much of the "innacuracies" in the methodology and the problematic nature of quantifiying fielding in general.

Anyway, Young is NOWHERE near the best player in the AL West, including positional adjustments of course. Not even close. I have him projected at slightly more than 2.5 wins above replacement in 07. Here is a list of AL West players who are projected to be better, according to my Slwys projections, again, including positional adjustments:

Bobby Crosby
Jason Kendall
Ichiro (especially in CF)
E. Chavez
Tejeira
Ellis
Vlad
Bradley
Beltre

That is in reverse order, BTW, with Beltre being the best of the bunch, at almost 4 wins above replacement.

Doesn't Bradley have to play more than 101 games to be rated that high? Something he has only done once in his career.

He's already missing games this year.

I have Adrian Beltre on my bench in our Lounge Fantasy league. And Juan Pierre,too. On my bench.

Here is a list of AL West players who are projected to be better, according to my Slwys projections, again, including positional adjustments:

Bobby Crosby
Jason Kendall
Ichiro (especially in CF)
E. Chavez
Tejeira
Ellis
Vlad
Bradley
Beltre

That is in reverse order, BTW, with Beltre being the best of the bunch, at almost 4 wins above replacement.

How many games do you project Bradley playing? He's 29 years old and has qualified for the batting title once in his career and didn't come close any other time. Unless the guy hits like Ruth and fields like Ozzie, I have a hard time believing he is the second best player in the division.

Rob Neyer: Still an Outsider?

Which one? Ponyboy or Sodapop?

Adam Everett, who gets blasted all the time as a "black hole" in the HOU lineup, along with Ausums, when he is in fact almost 3 wins above replacement, including defense and other peripherals.
....
Young is NOWHERE near the best player in the AL West, including positional adjustments of course. Not even close. I have him projected at slightly more than 2.5 wins above replacement in 07.

So Everett projects as better than Young overall? How far apart do you have them on offense? My offensive projections (based on Marcels) have Young at +15 and Everett at -24 (not adjusted for position). I find it pretty hard to believe that defense and other peripherals could make up that difference, much less put Everett ahead, especially considering the regression to the mean in defensive projections.

I have bent over backwards over the years to try not to call someone stupid-although I may have, I've written a ton of words over the years. I have called a certain move stupid but I try not to call a person stupid.

I like Rob Neyer. I genuinely do. I have all his books. But this selective memory at its finest.

To wit, this:

"How stupid is Buddy Bell? I’m serious about this. I honestly think the Royals hired a stupid man to manage the team."

I'll grant him this though: He's doing much better about put-downs in recent years. And being a Royals fan entitles anyone to temporary fits of rage.

I've been reading Rob since the start of the late, unlamented Rob Neyer Message Board.

Roy Hobbs is absolutely right, especially in the early days. I recall that Rob wrote a column a few years back, where he essentially admitted that he had been intemperate in his comments, and agreed to change the tone.

But when he started out -- when he probably had far fewer readers except on those days when he pointed out that Jeter could not field -- he would go after guys.

Of course, in the current FireJoeMorgan/Don Imus/Baseball Prospectus atmospohere, his comments and tone on average were pretty tame.

So Everett projects as better than Young overall? How far apart do you have them on offense? My offensive projections (based on Marcels) have Young at +15 and Everett at -24 (not adjusted for position). I find it pretty hard to believe that defense and other peripherals could make up that difference, much less put Everett ahead, especially considering the regression to the mean in defensive projections.

The Fielding Bible had Everett at +76 plays over the last 3 years and Young -73. I'm not sure how to convert those plays to runs, but the difference is 50 plays per year.

"Bobby Crosby"

Really? He hasn't played a full season since 2004 (which incidentally, is his only full season). The only time he was a good hitter was in 2005. He blew hard last year. I guess he rates pretty well defensively?

The Fielding Bible had Everett at +76 plays over the last 3 years and Young -73. I'm not sure how to convert those plays to runs, but the difference is 50 plays per year.

Well, AROM's defensive projections, which are the only readily available such projections I'm aware of and I believe were endorsed by MGL, have Everett at +15 (the best of anyone in baseball) and Young at -3. From what MGL says, it sounds like he has Young quite a bit worse, though.

Bobby Crosby
Jason Kendall
Ichiro (especially in CF)
E. Chavez
Tejeira
Ellis
Vlad
Bradley
Beltre

Wow. MGL is one of the few people that I pretty much believe, no matter what, but this list was surprising to say the least. I read it over three times just to make sure I understood it. And the Everett comment also made me cough up my dinner.

Young is NOWHERE near the best player in the AL West, including positional adjustments of course. Not even close. I have him projected at slightly more than 2.5 wins above replacement in 07. Here is a list of AL West players who are projected to be better

And that's fine, but again, when you talk about who's currently the best player, the usual connotation is "guy who has solidly established himself for a few years now," not "younger guy who projects to do things in '07 that he really hasn't achieved in the past with any consistency," e.g. Crosby, Ellis, Bradley, Beltre ...

MGL is one of the few people that I pretty much believe, no matter what

I'm sure you know this, but you shouldn't do this. I've been reading mgl's posts as long as I've been here (over 5 years), and while I respect him, I often disagree with him, even after reading his explanation on his position. I'm sure he would agree.

Dialed In: Gold Glove Winners

Dial has Young as an above average SS defensively.

Wow. MGL is one of the few people that I pretty much believe, no matter what, but this list was surprising to say the least. I read it over three times just to make sure I understood it. And the Everett comment also made me cough up my dinner.

You missed his post of 56 position players better than Derrek Lee(Spring Training '06)

Here's that list: Post #35: 56 position players better than Derrek Lee

Roy Hobbs is absolutely right

Really? I find this hard to believe. I started reading Neyer about when he started writing for ESPN (though I didn't participate in the message board. Too much nastiness every time I peeked in). The thing that always struck me about his writing style was the classic midwestern "oh, gosh!" tone. As in "Gosh, I don't pretend to know how to manage a major-league baseball team, but geez it would be nice if Tony Muser could develop some starting pitching ..." There was often a certain acidity in the sense of the analysis and there were some regretable columns (the one where he slagged scouts based on the language of their reports stands out), but the actual language of the columns was notably polite, and irony was not much in evidence. Indeed, as internet discourse became increasingly vituperative (e.g. Prospectus), Neyer's apparent sense of decorum seemed almost quaint.

Did he call someone stupid? I'm sure he did. As he notes, he's written a ton of words. "Stupid" will come up now and again. Not remembering is not an instance of selective memory but rather of the altogether common failure of memory.

Dial, Schmial. I was talking about "rate" and not playing time. Personally, I don't project playing time and I have not read or heard of a study that analyzes any particular methodology for doing so. IOW, if a guy has averaged 100 games over the last 4 years, I have no idea (and no opinion one way or another) whether he is projected to play 100 games or 140 (or 90). IOW, I have no idea, and I don't know anyone who does, how predictive number of games played in tha past is. I am talking about games lost to injury as the assumption is that a very good player is going to play full-time if he is healthy. Anyway, the above numbers and hierarchy are "per some number of games (150 usually), so anyone can make any adjustments they want based on projected playing time. As I said, I have no projections on playing time and no "opinions" one way or another.

As far as "established players" versus "young ones who have not achieved any consistency, yada, yada, yada," I have no idea what that means and all my and everyone else's projections are OBVIOUSLY based on EXACTLY what they have in fact achieved in the past, regressed towards some mean, and asjusted for age. A projection is a projection. It is the mean (or median I suppose) of the distribution of possible outcomes for player, regardless of how much playing time he has had in the past. The uncertainty on that projection depends mostly on that historical playing time, but a projection is a projection. It is what it is. If someone wants to say that a player with a +10 runs per 150 games projection is "better" than another player with a +12 per 150 becuase the second player has a larger uncertainty on that projection, that is fine by me. It is semantics. I was only answering the question of who had the best 2007 projection, per unit time (who would I want at bat in a random game situation, adjusted for position and defense), of any AL player, position adjusted, according to Slwts. Wheher that is the "best" player, I don't know or care. Depends on what you mean by "best." Could be "best looking" I suppose.

My offensive and defensive projections for A Everett, per 150 games, are -25 and +23. The +23 for a projection is almost ridiculously good, but that is around what all the numbers say, even after regressing appropriately to the mean. For Young, it is -14 and +11.

How could you (Hobbs) say "selective memory at its best??" Rob says that he MAY HAVE CALLED SOMEONE STUPID

Dial, Schmial. I was talking about "rate" and not playing time. Personally, I don't project playing time and I have not read or heard of a study that analyzes any particular methodology for doing so. IOW, if a guy has averaged 100 games over the last 4 years, I have no idea (and no opinion one way or another) whether he is projected to play 100 games or 140 (or 90). IOW, I have no idea, and I don't know anyone who does, how predictive number of games played in tha past is. I am talking about games lost to injury as the assumption is that a very good player is going to play full-time if he is healthy. Anyway, the above numbers and hierarchy are "per some number of games (150 usually), so anyone can make any adjustments they want based on projected playing time. As I said, I have no projections on playing time and no "opinions" one way or another.

As far as "established players" versus "young ones who have not achieved any consistency, yada, yada, yada," I have no idea what that means and all my and everyone else's projections are OBVIOUSLY based on EXACTLY what they have in fact achieved in the past, regressed towards some mean, and asjusted for age. A projection is a projection. It is the mean (or median I suppose) of the distribution of possible outcomes for player, regardless of how much playing time he has had in the past. The uncertainty on that projection depends mostly on that historical playing time, but a projection is a projection. It is what it is. If someone wants to say that a player with a +10 runs per 150 games projection is "better" than another player with a +12 per 150 becuase the second player has a larger uncertainty on that projection, that is fine by me. It is semantics. I was only answering the question of who had the best 2007 projection, per unit time (who would I want at bat in a random game situation, adjusted for position and defense), of any AL player, position adjusted, according to Slwts. Wheher that is the "best" player, I don't know or care. Depends on what you mean by "best." Could be "best looking" I suppose.

My offensive and defensive projections for A Everett, per 150 games, are -25 and +23. The +23 for a projection is almost ridiculously good, but that is around what all the numbers say, even after regressing appropriately to the mean. For Young, it is -14 and +11.

How could you (Hobbs) say "selective memory at its best??" Rob says that he MAY HAVE CALLED SOMEONE STUPID

Dial, Schmial!

I was just about to comment on Hobbs post! How can you say, "...selective memory at its best" when he (Neyer) just said that "HE MAY HAVE CALLED SOMEONE STUPID IN THE PAST" but that he bends over backwards trying not to do so??

So he admits that he may have called someone stupid in the past and you (Hobbs) point out that he called someone stupid in the past! That sounds like a good memory, but maybe not a great one. And we have no idea whether he in fact does try and not call someone stupid. We know that he has and we know that he does not do it a lot. Is that "bending over backward not to?" I have no idea, but we certainly know that he admits to possibly saying that. Where is the "selective memory?" Isn't this somewhat the opposite of selective memory!

As far as Everett and Young, I have Everett projected at -25 in offense and +23 in defense for 07, per 150 games, and Young at -13 in defense and +11 in offense, per 150 games.

BTW, I don't do playing time projections. My numbers and my analysis were based on a "per unit time" projection. If someone wants to include projected playing time and change the numbers, that's fine with me. I was answering a specific question and perhaps I should have made that clear - namely, who has the best overall projection in the AL west, per unit time, according to my Slwts projections. If that is different that "who is the best player" that's fine with me also. If "best" means a projection rate prorated by playing time, as I said, that is fine with me.

As far as "established players" versus young players who have not yada, yada, yada," I have no idea what that means, but a projection is a projection. It is based on what a player actually has done in the past (majors and minors in the case of most projection systems), regressed toward some mean and age adjusted. Does that mean that all projections have the same uncertainty (or certainty) level? No. Generally, the more sample historical data, the more certainty on the projection. But, regardless of the amount of data and the certainty of the projection, a projection is still a projection. It is the mean of the distribution of likely events in the future, based on a number of parameters. If someone wants to make the argument that a player with a +10 projection with a large certainty (lots of "established" data) is "better" than a player with a +15 projection but a large uncertainty factor, that is also fine with me. Personally, I'll take the +15 player most of the time, but I wouldn't crucify you if you took the +10 player (wanting to sacrifice some expectancy for uncertainty, assuming that both projections were reasonably accurate means in the first place).

that should be "...sacrifice some expectancy for CERTAINTY..." and sorry for the double post as I thought the first one was lost in cyberspace and had to write it all over again, changing some things around obviously...

Whoa, you rewrote that whole thing?

Wow. MGL is one of the few people that I pretty much believe, no matter what

This is something that I would imagine that a lot of people on this site would agree with, and I don't understand why. If a regular sportswriter wrote the kind of stuff that MGL often says, we would all be saying he's the worst sportswriter in America. The system MGL uses makes predictions that are really, really strange, to put it kindly. I don't think its because MGL is so much smarter than the rest of us - not just people on the site, but virtually every other writer in America. I think its because the system, for whatever reason, simply doesn't work.

I was only answering the question of who had the best 2007 projection, per unit time (who would I want at bat in a random game situation, adjusted for position and defense), of any AL player, position adjusted, according to Slwts

Well, the question of who the tenth-best AL West 2007 projection, per unit time, position adjusted, according to Slwts, is probably definitively answered, then. IOW you did not mean to say that Michael Young "is NOWHERE near the best player in the AL West," you just meant to say that he has nowhere near the best 2007 projection, per unit time, position adjusted, according to Slwts. So when Rob Neyer or any of us is asked that question, we will know how to respond :)

The system MGL uses makes predictions that are really, really strange

Though, to be fair, there were a hell of a lot of position players who had a better 2006 than Derrek Lee :)

You missed his post of 56 position players better than Derrek Lee(Spring Training '06)

Here's that list: Post #35: 56 position players better than Derrek Lee

Well, that's actually not a bad list, all things considered. Obviously, there's a huge number of problems with projecting simply performance per 150 games and not taking health concerns into account, but it's not like that last is full of comical names. There's a handful of guys (Bobby Crosby, notably, who apparently continues to project as some sort of great player despite his inability to either combine hitting and staying healthy since 2004) who are lousy choices, but as my fantasy team would attest, Lee himself was no great shakes in 2006 thanks to fluke injury, so there's some value in raw value analysis as well

nobody on the Athletics qualifies for this discussion.

Piazza! [/fanboy]

You missed his post of 56 position players better than Derrek Lee

I'd guess more than 56 position players were better than Lee last year.

That is in reverse order, BTW, with Beltre being the best of the bunch, at almost 4 wins above replacement.

Beltre? Adrian Beltre? That's some serious third base there.

Not that I have to justify Slwts projections - their strength is in including all of a player's peripherals in his total value. Other than that, most projection systems, including my own, are basically a modified Marcel. As Tango likes to say, and he is right, any idiot that has a basic familiarity with a Marcel type system (which is EXACTLY what works) can do a reasonable projection. FWIW, in a test between my batting projections and Chone and Pecota, Pecota and mine are just about a dead heat (OPS RMSE, etc.) and both mine and Pecota outperformed Chone. To be fair, there are many reasonable ways to compare the "accuracy" of systems and one is not necessarily better than the other. That being said, if any of the f****** geniuses on this thread want to put their projections up with my "really, really strange ones" I'll happily put up any amount of money they want and we'll let some third party be the arbiter of accuracy at the end of the season!

Remember one of MGL's most important rules! Anyone can proffer any opinion they want. Offer to bet money against them and they often realize that maybe they're not so smart after all. Of course, they usually just mumble something like, "Well, if I had the money.." as they are walking away with their head hanging down...

I might as well throw my AL west list out here:

Projected runs over replacement, counting offense, defense, position, baserunning, and playing time:

Vlad +45
Beltre +42
Teixiera +38
Chavez +37
Ichiro +32
Swisher +31
Johjima +28
The OC +25

Young is at +21, same ballpark as Kendall, Jose Lopez, and Howie Kendrick, and includes a -9 defense rating.

Of course, they usually just mumble something like, "Well, if I had the money.." as they are walking away with their head hanging down...

I always thought not having the money was a pretty good reason not to gamble, but I guess if the degree of confidence in a baseball projection is at stake, that's truly the time to throw caution to the wind.

@ AROM

Would you mind breaking down the components of Beltre and Ichiro? I'm genuinely curious where Beltre's 10 run aggregate advantage over Ichiro comes from.

I have Beltre at +8 in defense and +11 in offense. There is a +1 positional adjustment for 3B. Ichiro is +5 on defense including arm, another +2 or so in baserunning, and +13 in offense. But a corner outfield adjustment is -8.

Thanks mgl. Okay, that makes sense. I was thinking that the projections were basing Ichiro's contributions as a CF.

He's probably worth more in CF. I don't really know. My defensive ratings for him in RF have been less than his "rep" (and speed) suggests and some of the other defensive metrics.

Where is Joe Mauer?

In the AL Central.

if any of the f****** geniuses on this thread want to put their projections up with my "really, really strange ones" I'll happily put up any amount of money they want and we'll let some third party be the arbiter of accuracy at the end of the season

And you will win; but the original discussion was not about projection systems, but about who the best player is, which admittedly may be too fuzzy a question for your taste. I think that the assorted geniuses may have some legitimate questions about the value of defense to a baseball team. Basically, I think that More Indecisive back in #5 was right: almost any club in the AL, except the Yankees, Tigers, and Orioles, would happily trade their starting shortstop even-up for Michael Young. The Astros would do it. But by your system of evaluation, many of these clubs would be idiots.

The Young/Everett comparison is instructive. Danny in #19 suggests that Everett typically makes 50 more plays a year than Young. (Presumably at shortstop these would all turn into singles.) A quick glance at B-Ref indicates that Young typically creates 65 more runs a year than Everett. That first-glance discrepancy is so enormous -- 65 runs versus 50 singles -- that people have a hard time accepting that Everett is more valuable than Young (and since Everett just turned 30 and is only a few months younger than Young, people have a hard time accepting that Everett projects to get much better compared to Young in the near future, either). Now I can readily see that the 65 runs in Young's favor incorporate bunches of illusions that your analysis may well whittle away to nothing. But prima facie this takes a great leap of faith ... "strange" is not an overstatement. Indeed if you turn out to be both "strange" and correct, that's a compliment: you are a visionary.

That being said, if any of the f****** geniuses on this thread want to put their projections up with my "really, really strange ones" I'll happily put up any amount of money they want and we'll let some third party be the arbiter of accuracy at the end of the season!

My monkeys say they happily accept your challenge. However, they are not paid with money, per se, so if you are willing to accept, say, bananas or raisins in the unlikely event that you out-project" them, you have a bet.

"Dial schmial" may be the best thing I've seen posted on this site, ever.

What just struck me in a moment of irony is that suddenly MGL's prescience is being debated in large part because of an assertion about an AL West shorstop (Crosby) whom he sees as having higher value than many of the readers here do. Much the same, as beginning with post #1, as Neyer's 'quality' as a writer has hinged upon ONE statement--I throw out the Buddy Bell comment, because a) I agreed then and now, and b) Rob's a KC fan and likely sickened by how right he was--rather than the body of his work.

When one has read him consistently, as I have for at least the past few years, it becomes pretty clear that:

1. Rob really, really digs baseball
2. he is not afraid to take on baseball icons (e.g. Joe Morgan, Derek Jeter)
3. his commentary covers a variety of teams, so much so that it would be hard to pin him as a KC homer if he weren't so open about it
4. he is wrong sometimes in valuations and predictions
5. he is willing to admit (4) - I have read it, seen him admit it in chats, and had a nice email discussion with him

I, for one, would love to have more writers on the sports with a similar set of tools.

So in sum, we're skewering someone for being marginally wrong on one statement (which at least has given us a healthy debate from what he calls an 'off-the-cuff' comment). I'd much rather discuss the idiocy of some of the announcers that provide average fans their understanding of baseball acumen, or debate for 50+ posts the valuations of GMs past-and-present, which have real implications (i.e. SIGNING Young to a massive extenion, or Gary Matthews at all, or Mike Piazza, or what Mark Teixeira's non-Texas value--see home/road splits--might be, or Bill Bavasi vs. a monkey as GM...the list goes on and on).

or Bill Bavasi vs. a monkey as GM...

Tread lightly, sir.

Bob, Batting everyday in the top 3 spots means Young creates more runs, but uses a lot more outs as well. For their careers, and it seems fair to compare careers as they are the same age and came up at the same time, Young creates 2.09 more per 27 outs. Park and league factors ("AIR" on b-ref) are close to a wash. If you compare them at 400 outs, thats 31 runs, 39 at 500 outs, 35 seems like a decent compromise.

I've got Beltre at +11 offense, +9 defense, and +5 position. Ichiro is at +10 offense and +5 position - I have as a CF but don't have a rating for him there yet, not enough data.

I guess the two things that really jump out at me are the Beltre and Everett comments. It's almost the same problem for both, I just can't see their fielding making up for their problems with the bat. I had to double check, AB's OBP was 303 and 328 the last 2 years, both below "average". His OPS+ were 90 and 108. I thought at the time, and I still do, his signing was historically bad.

I may be beyond convincing that AE is a positive for a team.

I didn't really have too much of a problem with mgl's list of 56 players more valuable than Lee. Names like Kent and Beltre and a couple of others jumped out at me, but otherwise I thought it was pretty reasonable.

Swisher has been playing center field this season and not too badly as far as I can tell. He's much more valuable there than first base. But he needs a haircut.

But he needs a haircut.

He does, but he's growing it out for charity (when it gets longer, he's cutting it off and donating the hair to make wigs for cancer patients) so he gets a pass.

IT WAS FOR CHARITY!!!!!!

AB's OBP was 303 and 328 the last 2 years, both below "average". His OPS+ were 90 and 108. I thought at the time, and I still do, his signing was historically bad.


It depends on if you decide to completely throw out Beltre's 2004 season. I don't, and I know MGL doesn't either.

In my case, his 2004 season is weighted, combined with the other season, and regressed, but its such a monster season that any projection I do for Beltre is going to look better than any season he's actually put up, except for 2004. If you make the decision that 2004 was not Beltre, but perhaps the alien Josh Exley inhabiting his body, then you can throw that season out and get a weaker projection for him.

In any case, its not historically bad. Last Beltre was about average on offense, played everyday, and was great defensively at a key position. I'll trade you Gary Matthews Jr for him anyday, and Matthews wasn't the worst signing of last offseason.

I guess I thought that there was a chance Gary Matthews was legit, though small. I never thought Beltre's 2004 was anything other than a one-year, FA-binge type of thing.

1. The day someone like Neyer or Sheehan gets a vote, the players will revolt.
2. If MGL is so smart, why isn't he in Vegas making millions or working for a team? I know he DID work for the Cardinals in their failed experiement and thigs like his D-Lee tirarde are probably why he's not there.

1. The day someone like Neyer or Sheehan gets a vote, the players will revolt.

I don't think the players would care either way. However, it's much more likely that a Neyer or Sheehan would see the actual 'value' in a player of subtle contribution than many of the members within the "traditional" media. Too many of them are like Joe Morgan, any 'true insight' would be placed under arrest in a deep, dark dungeon with the key thrown into an abyss.

And so it goes...

--------
trevise

2. If MGL is so smart, why isn't he in Vegas making millions or working for a team? I know he DID work for the Cardinals in their failed experiement and thigs like his D-Lee tirarde are probably why he's not there.

1. How in the world would you know what I do and how I make my living?

2. Neither I nor the Cardinals has any idea what the "D-Lee tirade" is, although I'm sure I must have said something about D-Lee's value/ability that either is different from the conventional wisdom and/or different from what he actually ended up doing. And there was no "experiment" that I am aware of in the Cardinals organization that either failed or succeeded, unless by "failed" you mean one NL pennant and one WS championship. The Cardinals have at least one analyst working with them full time now who is doing exactly the same work as I was doing and more.

It is amazing how a moron/troll can pull misinformation out of their ass and then think nothing of spreading it, like an infant spreads his feces on the wall and in the crib. You (Jessex) should work for a corrupt politician.

mgl Posted: April 15, 2007 at 07:59 PM (#2335834)

... a corrupt politician.

dude i do believe that is what is called oxymoron

or rather the definition of politician - "a moron/troll can pull misinformation out of their ass and then think nothing of spreading it, like an infant spreads his feces on the wall and in the crib"

dude i do believe that is what is called oxymoron

Actually, I think it's a redundancy, unless you're looking to pay pols a compliment.

Picky picky.

Dial has Young as an above average SS defensively.

I appreciate that someone remembers. A +4 isn't really oustanding, and I comment after the list:

"Michael Young's defense really jumped from last season. That's odd, but it could be he worked on it."

Or it could just be a lucky season. And he's a 4 there, which is pretty close to average. I believe BIS had him around average in 2006 as well.

Dial, Schmial.

What did you have for Young last season? 0?

Young was -3 runs according to BIS data last season, but he's still projected at a pretty terrible -14.

RB in NYC (Now with Job!) Posted: April 15, 2007 at 08:43 PM (#2335863)

dude i do believe that is what is called oxymoron

Actually, I think it's a redundancy, unless you're looking to pay pols a compliment.

- grinning
i guess you are the other guy editing my stuff at THT

Yes, I had M Young as -4 per 150 last season, but so what? A projection is a projection. It could be that his true talent all of sudden changed for the better last year, but it is not likely. I sort of take that back. Actually, our estimate of his true talent DID change after last year. It got better. As I have said a billion times before (maybe more), our estimate of talent and hence our porjection is based on some kind of weighted average of a player's career numbers. Period. Yes, it could be that we are wrong with respect to Young or anyone else. What is the point of quoting how someone did last year, last month, or two years ago, when we are talking about our estimate of true talent and about future projections? A projection is a projection. A projection is a projection. A projection is a projection. A projection is a projection...

Yes, I had M Young as -4 per 150 last season, but so what?

Well, you poo-pooed the fact that I had him rated as about averaqge last year, as did you. So what is - why the "Dial Schmial" part of the comment. My commentary after the rating was that it was surprising.

I wasn't commenting on your projection, and I noted last year that it was an odd jump - so I'd just appreciate a lack of disparaging remarks, when you agree (essentially) with what I rated him at last season.

The Beltre thing struck me as a little off, too. I would guess that there isn't a single GM in baseball who would trade Michael Young for Adrian Beltre at this point, yet MGL has Beltre not just better but lots better.

There are really two possibilities here: MGL knows more about player evaluation than everyone in baseball, or MGL's rankings are wack.

- grinning
i guess you are the other guy editing my stuff at THT

Don't get a 4th editor, or the finished product might likely end up the same as before you submitted it to the 1st editor. ;-)

Well, Boots, I think that GMs may still undervalue defense at some positions. I think the Beltre rating is a bit high, and Young plays SS, not 3B.

I suspect MGL (and others) still project Beltre too high based on that one goofy season.

I suspect MGL (and others) still project Beltre too high based on that one goofy season.

Have you, or anyone else, looked at projections for players who have one totally outlying season? Someone like Derrek Lee has been consistent at a certain level, around .275/.370/.500, except for that one crazy season where he hit .335/.418/.662. I wonder if seasons like that shouldn't be strongly discounted in making projections.

Have you, or anyone else, looked at projections for players who have one totally outlying season? Someone like Derrek Lee has been consistent at a certain level, around .275/.370/.500, except for that one crazy season where he hit .335/.418/.662. I wonder if seasons like that shouldn't be strongly discounted in making projections.

Yes, I think I wrote an article on it...

guess you are the other guy editing my stuff at THT

Trust me, I'm the last person you want editing things. I'm the guy who struggles with keeping straight that although one can say Derek Jeter has led the Yankees led the Yankees to six pennants, you can never say that Derek Jeter has lead the Yankees to six pennants, even if it is pronounced the same way, at least insofar as what they make bullets out of is concerned. And let's not even get into my issues with run-on sentences.

There are really two possibilities here: MGL knows more about player evaluation than everyone in baseball, or MGL's rankings are wack.

***

I'd go with the former. The THT projections have Beltre at +3.9 wins above replacement, and Young at +2.5. Beltre is projected for a .799 OPS, which is pretty much exactly what he did last season (.793).

What amazes me is how bad that means the Mariners are. Most people think that one of the reasons they're expected to finish under .500 again is that Beltre has underperformed expectations. But Beltre is one of the biggest reasons they're as good as they are!

"What amazes me is how bad that means the Mariners are."

It's hard to imagine the impact of having Jose Vidro as the starting DH can have on a team.

50. More Indecisive than Lonnie Smith on 2nd...
I, for one, would love to have more writers on the sports with a similar set of tools.


This is probably what spoiled me on sportswriters, Rob was the one of the first sports writers that I followed and after reading his stuff, almost everyone else paled in comparison, for pretty much the reasons you gave. I mean can you imagine Mariotti admitting he was wrong? I started to read prospectus when it was free, and thought it was refreshing how it wouldn't hold back, but then you noticed that they would stick to their assumptions even when their own facts contradicted them. Rob was always an entertaining read and is almost reason enough to go to ESPN insider, but considering how ESPN.com has handled their message boards, their news programs, their website, their coverage there really is no reason for me to support them in any way.

I have Beltre projected (in his own park) at .785 in OPS. The M's actually don't have a bad lineup at all. In fact, it is well-above average. Sexson still projects well. Ichiro projects very well for a center fielder. Johjima projects as one of the better hitting catchers in baseball. Ibanez and Guillen are OK and Lopez and Betancout are both above average at their positions. As most of you know, DH is not a great hitting position as it should be, for various reasons. Vidro is OK, as he projects as a league-average hitter, a little below that of a DH.

So do you think they could actually win the west?

Their pitching isn't that bad, and nobody else seems to be any good either. The Angels' offense isn't good, and their pitching injuries are mounting. The A's offense is terrible, and now Rich Harden's down. The Rangers look like the best bet to me at this point, but they're not exactly good either. That's just eyeballing things, of course. My attempts at actual projections last year were so bad that I didn't do any this time.

2. If MGL is so smart, why isn't he in Vegas making millions or working for a team? I know he DID work for the Cardinals in their failed experiement and thigs like his D-Lee tirarde are probably why he's not there.

He has worked for one more team than most of us and how do you know he isn't making millions in Vegas??

This site is becoming about as much fun as a trip to the dunny, when you realise you don't have enough paper half way through.

I did my team projections before the season started using BP's playing time estimates. I had them dead last with 75 wins and a .075 chance to win the division. I don't know that has changed much. What changes projections, of course, is major injury or trades/acquisitions, mostly the former. It is a myth that Oakland has a very poor offense. It is around average. If Harden stays down, then yes, that will be a big hit (3-4 wins or so). The Angels have a horrible offense, but excellent pitching. Texas' offense is pertty bad, especially with Sammy out there. Other than Millwood, no one in Texas rotation is better than average, although McCarthy and Tejeda are not too bad. Seattle has of course one of the better pitchers in baseball in King. I probably undervalued him going into the season. Other than him, Batista is not bad, but that is about it for starting pitching and Putz of course is a good closer. Can they win the division? Of course, ANYONE can win their division. No team has a zero chance. I would say maybe 10% chance if everyone has equal health. If the M's stay healthy (at the least the players that matter) and one or more of the other teams do not, then that number goes up. Now that I just looked at the standings, they give SEA even more of a chance since they are around 1 game ahead of their "schedule," Tex is 1 game behind and OAK is .5 game behind (every team that is around a .500 team should be around .500 right now). So maybe 15% for SEA?

Thanks, mgl. To use the parlance of the younger generation, you rock!

Of course, ANYONE can win their division.

Especially with only 3 other teams to beat... it's really not fair that the NL Central has 6 teams and the AL West has only 4 teams. That's a very significant difference in terms of competition... not to give the Pirates and Cubs any other excuses for their suckitude, but I think this has not been stressed enough when evaluating such teams' futility.

Beltre is projected for a .799 OPS, which is pretty much exactly what he did last season (.793).

I have Beltre projected (in his own park) at .785 in OPS.

Yes, I'd have him a lot lower than that. That's weighting his big season too highly. I'd put him closer to .750. But he is 28, and it is the AL, so maybe. I think .720 is a perfectly acceptable target for him in a year or two.

I've got an .811 OPS for Beltre. I could get a .750 OPS if I did a marcel using 2006, 2005, skip 2004 completely, and use 2003. I can understand discounting the big year, but not ignoring it completely. Would you think more highly of Beltre had he merely put up an .800 his last year with the Dodgers? That would give him a marcel of about .765-.770.

Yes, I think I wrote an article on it...

Do you have a link?

Would you think more highly of Beltre had he merely put up an .800 his last year with the Dodgers? That would give him a marcel of about .765-.770.

I suppose I am probably not giving him enough credit for age-related improvement (as I hedged at the end there...), But okay, .765 at age 28 seems good for him.

But by-and-large, wrt Beltre's development, that 2004 should be discounted completely. I think *Beltre* has shown that (not me).

Sure- with other links

The MAriners do have a pretty good lineup. It is my hope that they continue to offset every Felix Hernandez start with a start by Jeff Weaver, the Anti-Felix.

Thanks Chris. I'm convinced Beltre is not a future Hall of Fame 3rd baseman, but the question of how to treat out-of-place years like Beltre's remains.

The projection systems have him right about where he was last year, if we are right he's got as good a chance to hit .340/.520 as he does to hit like he did in 2005, or 2001-03 again. We'll see.

It doesn't exactly answer the question, but Tango had a little study on banner years.

if we are right he's got as good a chance to hit .340/.520

Well, it is the AL, so that should inflate his stats some.

From that study:

Seeing that the first year didn't have any predictive value for the first group, let's redo our study, this time looking at a 4 year cycle instead of 5 years. (This will also allow us to increase our sample size.) There were 61 players in the first group and 61 in the second. Here are the results, along with the group's performance in year 4:


Year 1 1.32 1.02
Year 2 1.02 1.00
Year 3 1.02 1.34
Year 4 1.09 1.13

The year 4 results are consistent with Marcel.

In projecting a player, there is absolutely NO REASON to ever ignore or discount an anomolously good season unless that season involved cheating or you have reason to believe the player's true talent subsequently changed (for the worse) other than by the normal effects of aging. This is regardless of how anomolous that season is. For example, .730, .710, .940, .700, .714.

We can ALWAYS definitely answer a question like that (as long as we are not assuming other facts about our player in question) by looking at all players in history who have had similar patterns of performance. IOW, look at all players in history who have had an anomolously good season (banner year) and see how they did in the subsequent season or seasons. You will find that you can accurately predict any of those subsequent seasons by using some kind of Marcel. If you ignore or "discount" that anomolous season, your projection will NOT be accurate - it will be too low. So why would we ignore or discount Beltre's banner year? Can we assume he was on steroids? Even then, can we assume that the steroids added some gigantic amount of performance. Even if I was reliably told that he was on steroids for that year, I would not be comfortable adjusting his performance by more than 5 or 10% and then redoing his projection.

Stupid question: why do people think the Angels offense will be terrible this year? They've been around league average for several years now, and the starters they got rid of or benched this year (1B, 2B, CF) were, by OPS+, their worst regulars last year.

there is absolutely NO REASON to ever ignore or discount an anomolously good season

Hey, that philosophy worked out great for me.

Since every year is a sample of a player's true talent, it is simply his ecact true talent for that year plus or minus random sample error. That sample error can be anything from zero to infinity, normally distributed. In a large group of players, we expect to see any possible "pattern" you can think of by chance alone. Obviously the "weird" patterns (e.g., one great year surrounded by a bunch of much lesser ones) will occur less frequently.

Bottom line is that unless we know something (other than by surmising because of the stats - which is a gigantic mistake) about some change in a player's true talent, any pattern we see in a player's historical stats means nothing from the standpoint of projecting his future performance. Our best projection is always a Marcel adjusted for age. Period.

That being said, there are some caveats/exceptions to that. Research HAS shown us that certain patterns combined with age can tell us something unique about a player's development (and thus a modified Marcel is the best predictor of future performance), but it is a little tricky and one has to be careful with these kinds of things...

Our best projection is always a Marcel adjusted for age. Period.

Except that a projection is just a projection. I'm adjusting for table texture.

I take back what I said about the Angels offense being terrible. It is below average, but not terrible. Dag Nabbit, be careful about just doing a "crude" assessment. If you want to look at them over the last few years as compared to now, look at PARK ADJUSTED team numbers, like team OTS, and then do your substitutions and compare that to the average AL team OTS. Or simply take their current lineup and apply Pecota's offensive projections for each player. Trust me, Pecota's (or Chone, or ZIPS, or Shandler's, or Tango's, or THT's) is as good a projection as you are going to get from ANYONE.