Brawndo the Thirst Mutilat

You've answered your own question when you said "Dillon played quite a bit in Nashville." The better projection systems like ZiPS use MLEs for minor league stats and then treat them as substitutes for MLB stats where those are lacking -- so what's been really driving Dillon's projections are his 983 PA in AA and AAA between 2004 and 2007, during which his OPS was over 1000 every year ( MLE OPSes in the mid 800s). Properly adjusted for difficulty of league and park factor, AA and AAA stats are virtually identical to actual major league stats in their predictive power for future MLB performance. We have plenty of information to rely on, we just have to recognize that we can quantify the difference between the upper minor leagues and the major leagues. I understand the concerns about Dillon's defensive limitations, but I think that the org is underestimating Dillon as a hitter, and I think it is likely that the manner of thinking in which they are underestimating him would go something like "well, we can't bring him back as a PH, we tried to give him that job in '08 and he stank." Baseball men believe in the existence of "professional pinch hitters" who are able to "handle the pressures and difficulties of hitting without getting regular playing time" and, as a consequence, they are often wont to give the last spot on the bench to "Lenny Harris."

ZiPS 2008 projection for Dillon was 280 / 338 / 449, so his crappy actual 2008 (aside from a nice walk rate) actually did a lot of damage to how highly ZiPS thinks of him. IIRC, that 280 / 338 / 449 line was the lowest of any of the notable projection systems, so ZiPS wasn't that high on him to begin with...now it's even less high on him...and it STILL thinks he's close to an average major league hitter. As for how they can project a gain from his 2008 numbers; he only had 90 MLB PA and 206 AAA PA in 2008. That's not anywhere near enough to tell us that his hitting skills have meaningfully declined, so what you'll see from projection systems is that the mass of data from 2004 through 2007 that says that he's an excellent hitter is balanced against the more recent info (which is more heavily weighted pound for pound but which is a lot smaller sample). It's not so much that the system "projects improvement" as it is that it is built to be sensibly skeptical about sample error in small samples of data -- something it often seems baseball fans and even scouts and management types are constitutionally incapable of doing. It's hard for me to see any sort of rational basis for doing a projection any other way.

Even if there was some sort of 40 man crunch involved, this was dumb dumb dumb. The Brewers just dumped a guy whose ZiPS projection, even after a really rough year, is still a totally decent and quite-useful-as-a-bench-player 255 / 334 / 401 for a guy whose 2008 ZiPS projection (243 / 288 / 361) accurately forecast his 2008 numbers, which worked out to an MLE of 242 / 295 / 341. If 40 man space was going to become an issue at some point this offseason, just wait and cut somebody when it does. It's never a good idea to dump useful spare parts for completely useless ones.

Here's a description of the methodology used to derive those projections: These are based on 5 years of data from 2004-2008, weighted 1, .8, .6, .4, .2. The numbers are a combination of Stats ZR, Revised Zone rating, and Totalzone (which is not available for 2008). Instead of regressing these to a league average, the regression is to a run value derived from the Fan's scouting report. http://lanaheimangelfan.blogspot.com/ So it certainly isn't that Chone Smith has developed any revolutionary way of dealing with infield shifting. That's as unadjusted-for here as it is everywhere else. Given the lack of a clear-cut winner in the advanced defensive metric fray, he's simply making a pretty sensible move and tossing all the freely-available zone ratings based on the two different sets of stringer data available (STATS and BIS) in a pot, letting it simmer for 5 years, and throwing in a dash of Tango's Fan Scouting Reports for flavor. These numbers always look the "most correct" to me too -- what this tells us is one of two things. Maybe both of two things. 1) One year's worth of defensive statistics is simply too small a sample size for quantifiable results to match up with actual defensive talent. This is most certainly true -- the usual rule of thumb is that 2 years of defensive data is approximately equal to one year of offensive data. So one year plus-minus or defensive runs saved totals are about as likely to match up with a given player's actual ability as, say, post AS-Break OPS. 2) What our "eyes see" isn't nearly as objective as we tend to like to believe it is. Not only are our great big brains heavily involved in the actual "seeing", they do a tremendous amount of filtering and interpretation, a huge chunk of which is defined by preconceptions about the reality we believe we are seeing. As a result of this, our "eyes" may persist in telling us that Weeks still stinks, even if he isn't stinking anymore, because for most of his career, he really has stunk. At some point, "Weeks stinks" just becomes a sort of tenet of faith that is stubbornly stuck in our heads. So when somebody estimates his true talent level at "stinks" by using a weighted mixture of all the available info over a 5 year period, we are pleased with the results and believe them to be accurate. I really do like the methodology here (guess I'd rather he was using Dewan's plus-minus and UZR in the pot, but those aren't freely available, so what are you gonna do?) and I believe that for the bulk of the player population this kind of approach is going to give you the best estimate of true talent you're likely to get. But, assuming for the sake of argument that Weeks has actually improved at 2B from his pretty disasterous early career numbers, this approach will still punish him for older data that isn't quite as relevant to his current performance level as it tends to be for the bulk of the player population.

He certainly wasn't in better in '08. Gross was much better. According to numbers here and here, even if you include Gross's time in Milwaukee he was worth something like 15 - 20 more runs than Hart was on offense and defense. Hart's non-SB baserunning (steals and CS are included in the offensive values given) might eat into that difference by a couple of runs, but there's really no way that Gross wasn't at least a full win better than Hart in '08. That's not to say Gross will continue to be better in '09. Hart was much better in '07. But it wouldn't surprise me much if the Rays would view subbing Hart for Gross as a lateral move. Dumping Gross for nothing (I feel pretty comfortable saying that Butler is nothing at this point) was a really terrible move.

I certainly wouldn't dispute that Pujols is better at making contact than Howard. It's just that when you say "contact hitter", Albert Pujols isn't exactly the first guy that pops into my head. And for the most part, the kind of guy that I do think of (Ichiro, for example) is a lot worse at getting on base than people tend to think.

I think power vs. contact is a false dichotomy, at least as it relates to OBP. I see a lot of posts where it seems like people posit that there are two types of hitters: 1) Low AVG, high K sluggers with low OBP (We have too many!) 2) High AVG, low K slappers who hit lots of line drive singles and have high OBP (We don't have enough!) In actuality, the yearly OBP leader board is routinely populated mostly by sluggers that drew a crapload of walks, and didn't bad-luck into a .230 AVG that year. The Adam Dunns, Jim Thomes, and Travis Hafners of the world typically tend to cream the Mark Lorettas and Ryan Theriots in OBP. Our problem isn't a lack of focus on contact hitters in the draft, it's that our sluggers don't control the strike zone well enough. That has traditionally been viewed as an instructional failure -- but I think that's mostly wrong. I think that strike zone judgment ought to be treated as the 6th tool...and maybe the most important one. So I guess maybe I think it does still come back to Jack Z...but not because he's focused on power hitters in the draft. Drafting power hitters is the best way to find OBP, but they have to be selective power hitters instead of clueless hackers.

Pull for Haudricourt to be wrong. That would be doubly satisfying.

As far as I know, nobody has figured out a good way of doing this so far, since the two companies that have stringers at every game (STATS and BIS) don't track anything other than where a hit goes (the field is divided into a bunch of different zones) and what kind of hit it is (GB, LD, FB, "fliner"). All the defensive metrics are attempts to compare a given player's rate of turning that ball into an out to the league average rate at which that ball is turned into an out. The two things you listed and the effects of defensive shifting are the most notable holes in the current state of defensive research (of which I am aware, at least). They are problems, to be sure, but I think that you can still get a pretty good idea whether a guy is helping or hurting his team with his glovework, and usually at least some approximate notion of about how drastically, as long as you are careful to make sure that the various advanced metrics (whatever methodology is used to turn plays made and missed into runs and whichever set of source data is used) are all telling you more or less the same thing. If, after at least a couple of seasons, you're still not seeing any sort of consensus (as with Hardy for some reason), I don't think any of the people doing this sort of research are using a system sufficiently better than the others that you can just believe them over everybody else. Which isn't to say that some of these aren't better and some aren't worse. I think that Dewan's , MGL's UZR, and Dial's system are all very good. Non-zone systems like BP's FRAA are much less reliable. In effect, you are assuming that the guy is average at all the stuff that is in a blind spot. For most players, I don't know that that's a crazy assumption. Even for guys like Weeks and Fielder, who seem really atrocious at the respective skills you listed, I'm sceptical that the results are impacted by more than 5 or 10 runs. As a for instance, Weeks actually turned 2 more double plays than the average 2B in the number of innings he played (84 vs. 82 expected). Given that the Brewers put less runners on than the average team (so if anything he should've had less opportunities), I find it difficult to believe that Weeks could've choked away more than 10 or so more DPs than the average 2B would have and still met his DP quota for the year. I saw him blow a bunch of them, just like everybody. Maybe every 2B blows a bunch of them. Maybe his arm strength and release time are such that when he doesn't totally duff it in a hugely noticeable and irritating fashion, he's actually converting on a few extra DPs that an average 2B's throw would've been a split second late. I don't really know. I do think it's important not to be too dismissive because of the blind spots. Batting Average, for example, has huge blind spots that we've come to recognize are better measured by OBP (how often do you walk?) and SLG (what kind of hits were they?) respectively. But for the century that everybody was using AVG, they were still better off with AVG than they would have been just using their eyes and trying to look at how pretty some dude's swing was, or maybe trying to remember how often it seemed like he got hits.

Agreed. Hardy was 5.9 above average according to this guy, who has a freely available project similar to Chris Dial's O+D. When that happens (the metrics can't agree), it's probably best to fall back on what your eyes tell you. Mine tell me he's a little above average, and judging from Tango's fan scouting reports, most of you agree with me.

It doesn't say too much for your defense if your team is consistently slotting Mark Reynolds (he of the 34 errors, and runs saved ratings in the -10 to -15 range according to the advanced metrics I've seen) one spot further along the defensive spectrum than you.

Before he developed the bizarre metabolic issues, Baldelli: 1) tore his ACL prior to the '05 season 2) injured his elbow while rehabbing from that, eventually requiring Tommy John surgery to fix the damage, which caused him to miss most of '06 3) pulled a hammy in ST of '07, which lingered throughout the season, was aggravated during a minor league rehab stint, and ultimately caused him to appear in only 35 games. That last one seems kind of weird, right? Well, those kind of muscle problems (and extreme fatigue from even light workouts) are what caused him to undergo extensive testing prior to this season. Doctors never figured out exactly what was wrong with him. It might be a lottery ticket worth purchasing for like1 year and maybe a couple million bucks, tops. Anything more than that seems reckless to me. And I think you'd have to be crazy to go into a season expecting him to hold down a starting spot in the OF all year. That kind of move makes a lot more sense for the Mariners or the Nationals or somebody that doesn't have the remotest chance to contend in the first place. Part of my problem is that he was never anywhere near as good a player as everybody seemed to think even during his two healthy seasons (now 4 years removed). The Joe DiMaggio comparisons, for instance, were patently absurd.

My favorite successful former manager hasn't been mentioned anywhere in this thread. Aside from a bizarre proclivity for intentionally walking Barry Bonds, he always showed that he had an excellent handle on what a manager should and shouldn't do, and how statistics should and shouldn't impact that. His team won its division 4 of the 5 years he managed it -- his only real failing was a total lack of playoff success. Doesn't really affect my judgment of him much -- count me amongst those that attribute playoff success or failure almost entirely to the whims of fortune. He's probably not a serious candidate, as by all accounts he seems happy to go on being the Astros' version of Bill Schroeder and writing a book from time to time. But if Larry Dierker wanted to manage again, he'd be my first choice, unless Earl Weaver somehow felt up to the task at 78. Among those who have been mentioned, Davey Johnson would make me happy. It doesn't seem like anybody would really like Macha very much, but at least he wouldn't bunt every stupid inning. So I wouldn't freak out if they picked him. Everybody else gives me the willies. Especially Brenly.

That's spot on. Either possible outcome will produce an equally insipid "emotions and intangibles" narrative. Ever thus from the mainstream baseball media.

I'll echo doubts that Giles would be available, but if he were, he'd be a huge addition. He was about half a win more valuable on offense than Braun this year -- that's how good a hitter he is, though nobody seems to realize it. Advanced metrics still like his defense a lot, too, but even if you discount those he was almost 4 wins better than Corey Hart on offense alone. If you believe that Giles really was 15 runs better than an average RF, he was over 5 wins better than Corey Hart overall. That's a cataclysmically large gap.

I started this topic under the premise replacing a corner OF'er would be easier to do than replacing an All Star caliber ss or 1B. So just trading Hart for another RF'er would sort of disprove my premise. For whatever reason, unless the Padres or Halos inexplicably decline options on Brian Giles or Vlad Guerrero, the FA market doesn't have any decent regular RF this offseason. As for Abreu, I'm inclined to agree with Eric Seidman's assessment over at fangraphs. He's a guy all the advanced metrics agree on -- his defense has become a complete disaster over the past couple of seasons. He might be as bad as 25 or 30 runs below an average RF, which is basically as bad as what we saw from Braun over at 3B. His hitting is still solid, but it's slipping, and he's not exactly the kind of guy you expect to age well. Even if he continues to hit at his current levels, his defense will be altogether too atrocious to justify giving him the kind of contract it will take to sign him (I'm guessing at least 15 mil per, probably at least 3 years). Because MLB GMs still haven't totally gotten the hang of evaluating park factors and defense, Brian Giles would probably come cheaper than Abreu. He's a similar hitter (was actually slightly better in 08), and is a much much better fielder. He'd certainly be worth paying what he would make on the open market -- he may even be worth paying what Abreu would make, though not for 3 more years at his age -- but I highly doubt that he'll be available, since the Padres are smarter than your average bears and they have to know what an insanely good value Giles's option is (just 9 mil, and since his buyout is 3 mil, it's really only 6 mil savings if they cut him). Baldelli would be a total flyer -- he's spent so much time on the DL the last 4 years that he's had around just one full season's worth of PA over that entire duration. He's a total hacker like Hart, and even if he has somehow fully recovered from his injuries, he was only an average hitter before, so that's your optimistic outlook. I can't imagine a team hoping to contend bringing him in as anything more than a reserve, unless he goes nuts in the WS or something and somebody is dumb enough to believe that means he's a big game clutch performer.

Unless you think that Gamel should be next year's RF (which I'll grant isn't totally crazy), trading Hart to fill 3B or SP would simply open another hole that needs filling, so a challenge trade for another RF makes as much sense as a trade for either of those other 2 commodities. At bottom, I brought Hermida up because I think that he will be a more valuable player over the next 3 years than Hart will.

I found the '09 Marcels here. They have Crawford at 296 / 345 / 455 for a wOBA of .346, so yes, he also projects as an equally valuable hitter to Hart. But whereas Hart and Hermida are both entering arbitration for the first time, Crawford is already signed to a arb-buyout extension that has options for '09 and '10 (8.25 mil and 10 -11.5 mil depending on escalators). So you'd have Crawford for one fewer year, and he'd cost more than either Hart or Hermida during the time you'd have him. And while he's probably the best defensive LF in the business, his arm won't play in RF and he hasn't played a ton of CF, so you'd probably have to be willing to move Braun again. Plus, Crawford walks as seldom as Hart does, and his perceived value has always seemed way too high to me for the player he actually is because he steals so darn many bases. I'd be willing to think about Hart for Crawford plus (depending on what the plus was), or Hart for Crawford straight-up if the Rays would kick in some dough (maybe 6 mil?) to offset the price difference, but I doubt the Rays would be too interested in either of those trades. Hart for Hermida plus seems much more plausible from the other team's perspective.

It would be a dreadful waste to convert Martin to any other position full-time, since he can catch and can hit, which is a pretty rare combination. Considering what Martin would likely cost in talent, I don't really see any way the Brewers make a play for him unless the Dodgers would take Kendall as part of the package of players for Martin. You probably don't want to end up paying Kendall 4.6 mil to be a seldom-used backup and probably a royal pain in the butt to boot.

Apparently the Marlins are looking to deal Jeremy Hermida: http://www.marlinsbaseball.com/hermida-sought-after/ Hart for Hermida might make for an interesting challenge trade. Both play RF and have virtually identical service time according to Cot's (.005 years less for Hermida). A recent '09 Marcel projection for both found them to be almost perfectly equal in expected hitting value: Hart -- 277 / 331 / 477 for a .346 wOBA Hermida -- 273 / 355 / 447 for a .347 wOBA As far as fielding goes, most metrics had both as a little above average in RF last year. They were exactly identical in RZR (.890) and made a similar number of OOZ plays per inning (Hermida would have made 59 OOZ plays in Hart's 1376 innings to Hart's 62 OOZ plays). Previous year's RZRs are very similar as well. Tango's fan scouting reports show that we like Hart a lot more than Marlins fans like Hermida, so I suppose you have to give a nod to Hart based on subjective consensus. Hart's obviously the better baserunner. Hermida is two years younger and was much more highly valued as a prospect than Hart, and I'm sure most scouts would still agree has a significantly higher ceiling as a hitter. He also bats LH, and while LH-RH balance in the lineup isn't as big a deal to me as it seems to be to a lot of folks around here, I guess that still needs to be noted as a plus. In sum, I think Hart has been the slightly better player to this point, but they should be considered about equal going forward, and Hermida still has some chance to break out as a genuine superstar-caliber player, which I don't think I'd say about Hart. I think it would be a pretty fair straight-up trade (you'd be hard-pressed to find two more astonishingly equal players in nearly every important respect), and the market may still value things like RBI and SB enough that we could actually get the Marlins to throw a little something extra in, since most MSM-types and casual fans probably think Hart is a substantially better player.

Swisher's BABIP this year was .251. His LD% was a career high 20.9%, so his BABIP would have been expected to be around .330. He'll likely have a huge rebound season next year. If the White Sox really view him as a guy worth dumping at those prices (not sure they do), sign me up.

Weeks is also perenially among the best hitters in baseball at getting hit by a ton of pitches. Really, the only perspective from which it makes sense to say he isn't the sort of hitter who "tries to find any way to get on base" is if you absolutely require your leadoff hitter to be the sort of guy who tries to bunt for base hits all the time. There just aren't that many Brett Butler types in the big leagues today. And most guys that approximate that type really sorta stink.

In order to correct for this problem, you calculate a park factor by comparing a team's performance in that park vs. in all other parks. If you look here, you'll find a very long, and mathematically rather daunting, explanation of how B-R goes about doing that. To give an example of how detail-oriented their approach is, a batter's park factor is even adjusted for the fact that he doesn't get to (or have to -- if they're good instead of bad) face his own team's pitchers.

I actually think that of the freely-available on the net park-adjusted stats, Sean Forman's 3 year park factors used for OPS+ are about the best that you'll find. Some other places still use one year park factors, and while that might make more sense for a place like Wrigley, where the weather is the predominant reason it plays as a pitcher's park or a hitter's park, for most parks the more data you use, the more accurate your park factor. The only real problem with OPS+, in my view, is that it treats OBP and SLG as equally important, when in fact we know that a marginal point of OBP is around 1.8 times as valuable as a point of SLG, leading to the underrating of low SLG guys like Ichiro, as kramnoj mentioned. For quick and dirty purposes, I don't think there's anything wrong with it at all.

OPS+ ain't perfect, but you have to do something about park effects when comparing two players who have played in as drastically different environments as Hudson and Ellis have. Ellis's BABIP this year was a terrifically unlucky .249; given his 20.1% LD% you'd expect his BABIP to be around .320. Ellis and Hudson have comparable discipline (virtually identical BB%). Hudson has had better batting averages (leading to higher OBP), Ellis has had more power. Batting averages being less stable than discipline and power, I'm really not sure that I'd say that Ellis is harder to project reliably just because his last few years have had more variance. I'm willing to give the slight nod to Hudson offensively anyway, because he has had better recent results. I just think that you have to give a much more definitive nod to Ellis defensively. Hudson is good (didn't call him average, even though advanced metrics tend to), but Ellis is outstanding -- one of the true elite defensive players in the game today. It's not shocking to me that somebody would prefer Hudson, but it was pretty surprising to me that somebody as stat-savvy as you wouldn't think enough of Ellis to list him as a FA 2B worth a starting job. For my part, I think that Hudson vs. a cheaper Ellis is a complete no-brainer in favor of Ellis. I'm not really in favor of getting rid of Weeks, but Ellis is the one guy I think offers enough value at a price I'm guessing is reasonable enough that I'd be really happy if they signed him.