Spreadsheet Baseball: PECOTAing, Part 1

    I received my copy of Baseball Prospectus 2008 a couple days ago, and since I've been up to my usual shenanigans that are brought on by getting "The Book." In effect, it means that every other book I'm reading right now goes on the back burner until I've read all 30 team essays, scanned the player comments, and perused whatever back-of-the-book piece I find the most interesting. You can imagine how well this works what with having a Social Psychology textbook to read. Oh well, psychology classes come and go, but you only get your BP Annual once a year. If you had a friend in college who you wondered how he ever got any work done, I'm channeling him right now. But enough about me and some friend you might not have ever had, and on to the breakdown of what BP sees in store for our beloved boys in blue for 2008.

    First of all, if you're at all unfamiliar with Baseball Prospectus's projection system, PECOTA, you should become close--nay, intimate--with it. The Player Empirical Comparison and Optimization Test Algorithm, created by Nate "The Great" Silver, is considered by the majority of people who know their baseball analysis to be the best projection system out there. This should not be seen as a slight to other projection systems, as Baseball Prospectus is hardly the only fine baseball sabermetrics/analysis site out there. I'm not going to write out the entire details of the projection system here, because it's not my system, but I do recommend the annual and the BP website if you're looking to understand how these things work. And do remember that everything we deal with when it comes to our analysis does deal in probability. That's not to say it's unfair to note when PECOTA or ZiPS or another projection system or a fallible human such as myself is off with their expectations for a player or a team, it's merely to say that--if I haven't hammered this into your heads already--the best way to evaluate baseball players is with a combination of statistical analysis and scouting. In other words, there are caveats when it comes to these projections, starting with the fact that the one listed is the "weighted mean" projection. The word "mean" should suggest to you that it's an approximation of what can be expected from a player, and the projections usually are quite accurate. However, in any data set there are outliers, and so PECOTA does--though these are not seen in the book--project the potential highs and lows of the players as well. If a player goes above and beyond his weighted mean, chances are he's one of those guys who somehow managed to beat his 90th percentile projection. If memory serves, the weighted mean is a little over the 50th percentile projection for a player, taking into account the multiple factors to try and give us the best possible idea of what we're going to see from any given player in the majors in 2008. Additionally, the chances of a "breakout," an "improvement," a "collapse," and "attrition" are all calculated.

    There are detractors of such projection systems that say that since the system never actually hits the nail on the head--and sometimes misses badly--that there's no real point to them. These people might have some valid points--if you dig through any projection system's results, you're bound to find plenty of bad misses--but they're essentially showing why "best" is the enemy of "better." It is easy to see the folly in an argument where the main premise is that if we cannot pin this down exactly, we should not try at all despite the considerable information we can gain. Never try, never win, folks. If you don't ask the pretty young woman to the dance, there's nigh zero chance she'll go with you. If you do, there's a much better chance. If you don't try broccoli, there's zero chance you'll like it. If you, there's a significantly greater chance. We're not about perfect in stats analysis, we're about that "significantly greater chance" that we can be on target with our expectations and learn more about this funky game. We might aspire to perfect, but other than George Washington or Scarlett Johansson, no person has actually attained perfection. The same is true for projection systems.

    Newsflash: Rany Likes the Royals

    Once again we're treated to Rany Jazayerli writing the essay on the Royals section, and once again he does a fine job even if perhaps he's a little higher on the Royals than your average analyst. I'm not writing this piece to reprint every word written, because that would be plagiarism (which is a word is spelled wrong. Why are there two "i"s in that word? I hate english), but I'll share with you all my favorite quote from Jazayerli's work.

    "In the short term, the Royals cannot win on homegrown talent alone, nor do they expect to, judging from an off-season in which the Royals were players for [major free agents]. In the end, Moore landed outfielder Jose Guillen and capitalized on new manager Trey Hillman's experience in the Japanese leagues by bringing set-up man Yasuhiko Yabuta across the pacific. Moore also outbid the Yankees for lefty reliever Ron Mahay and pounced on catcher Miguel Olivo when the Marlins cut him."

    I'm encouraged by the fact that I think Jazayerli and my perception--indeed, a perception some of you all share--of the Royals are basically in sync. Jazayerli, who might know more about the Royals then any other informed outsider, acknowledges that it will be awhile before the Moore-drafted contributors start to bring home the bacon, and that in the mean time Moore had recognized that the team needs outside help to win. While I might have been a little bit more negative in my speculative piece about the Royals being on the upswing, this was a general point of mine. For the current group of young players to win on the Royals while still in their cheap years, the Royals actually are kind of heading into a mini "win now" mode--if that makes sense at all--until the combination of guys like Moustakas arriving and the eventual re-signing of people like Gordon and Butler and Bannister provide the Royals with a base for lasting success. Rany, of course, put this more eloquently and clearly than I did, and it's encouraging to hear someone at BP get behind Moore's machinations of this off-season just because the Guillen contract was one that made me feel as if the Royals were going for a competitive team sooner at the possible expense of later. In reality, if "later" is 2011 or so, it's impossible to condemn that later as being a down period for the team anyway. There's too much time in-between now and then and too much drafting to be done during that time.

    Anyway, the main point here is to say that while Rany is a Royals fan, we should not hold that against him. While he might be high on the Royals future, so am I and you should be too. The problem with a sustained losing, incompetent regime is that when a good front office takes over the rebuilding program that is finally installed still just looks like losing for a while. But seriously folks, this season is the one where we should begin to see the turnaround taking place on the scoreboard and in the standings.

Without further ado, on to the PECOTAs (all lines listed are PECOTA 2008). We'll cover the hitters this week, and the superhuman pitching staff in next week's article.

The Now-Maligned Royals Review Favorites

CF David DeJesus: .270/.350/.400 in 558 PAs, 9.8 VORP

It makes sense, if you think on a bit, that DDJ is someone primed for a bit of a bounceback. His projection takes into account that someone of the damage done to DeJesus's 2007 line had to with a 40-point drop in David's Batting Average on Balls In Play (BABIP). However, it was well-documented on this site in the second half that DDJ hit with all the power of a dying seagull in the second half. A drop in BABIP does not explain away the drop in his Isolated Power. I will once again point out that I think the fact that DeJesus played the entire year actually hurt him; there are some guys who are more effective with a couple days off a week, and it's funny that David was not someone who got those days off when you consider that Joey Gathright was on the roster for most of last year. Anyhow, I believe that the 700+ PA may have tired him out a bit, as he had previously ended up in the mid-500s with his plate appearances for the year. I also like his defense to be more consistently around average if Gathright spells him a bit. In other words, I think the batting average could come back up a bit if he indeed only gets 550 Pas instead of 703. I don't think PECOTA will miss by much though, because I no longer see DeJesus as someone with a power breakout in his future. PECOTA agrees, giving him only a 16 % breakout percentage.

C John Buck: .240/.310/.400, in 315 PAs, 3.4 VORP

What do we know about John Buck at this point? Well, he once again hit incredibly well for two months and was a non-factor for the other four. Look a bit deeper, and you'll see that even if that's true and even if this is all there is, Buck still took significant steps forward last year in his walk rate, power, and defense. Not that he's a "gold glove caliber" guy, but he's solid and his offense at the current level is jes fine sez bug for a catcher. I agree with PECOTA's assessment that he can bring his average back up from the .222 figure it ended at last year. For one, Buck has hit .240 before, and it's not all that hard to hit .240. For another, there's reason to think that the Royals tinkering with Buck's stance may have nerfed a more significant breakout with the bat. Next time that KC has a catcher--whether it be Buck or some other wearer of the tools of ignorance--who is OPSing almost a thousand...they might not want to mess with said player's batting stance. Please. PECOTA is oddly split on our favorite stormin' mormon, with all four percentages--breakout, improve, collapse, and attrition--over 30% percent. Encouraging sign: improvement is by far the highest, at 52%.

IF Esteban German: .280/.360/.390 in 434 PAs, 10.0 VORP

While you hate to wish for Mark Grudzielanek to get out of the way, and by now everyone should know not to bet against Grudz, it would still be an interesting experiment to see what German would actually do with a starting job. While German's average fell off, his modest pop and considerable patience remained in 2007. He's bad defensively, yes, but you could much worse than to have German as a bat off the bence and understudy to injury prone/prone to suckage players like Pena, Grudz and Shealy. I think BP has this projection pretty much nailed.

1B/DH Justin Huber: .250/.320/.420 in 382 PAs, 0.9 VORP

No, I will not shut up about this guy. Until the Royals solve the 1B/DH problem in some satisfying way, there is always going to be the "well, we could give Huber another shot" argument. Huber is 25, has been jerked around his whole minor league career, and there is some indication he could hit reasonably well if given the chance. While Huber's translated EQA of .261 was nothing special, Ross Gload's was .267 last year while he got numerous accolades for his hard nosed play and even "the most productive hitter on the team" by some. Shealy's was .216. You see how it's easy to want this guy to get a shot? He hasn't failed in the majors in any convincing manner--we all know he's been handled badly when up with the big club--and his 60% improvement projection lends to the hope that he could be more than the 740 OPS player that PECOTA projects. Anyhow, the odds are stacked against Huber and he'll never be a star, but this is a team that just re-signed Ross Gload for seven figures and thinks Ryan Shealy is worth a darn. Face it: it's not like the other options blow Huber out of the water.

The Big Fish

RF Jose Guillen: .279/.334/.445 in 497 PAs, 7.8 VORP

As Astro would say, "ruh roh." PECOTA doesn't see much in the of improvement and virtually no breakout potential, so Guillen's going to have to enjoy hitting at the K more than the algorithm says he will to be more than adequate. Actually, I think adequate would be pretty cool compared to Emil Brown, but it's worth noting that A) Guillen is so highly paid that if he posts that PECOTA line the year will definitely be considered a net loss B) this is one of those "win now" players that's supposed to be good for us heading into the end of the decade, and a sub-800 OPS doesn't suggest he'll have much staying power the last two years of his contract and C) PECOTA loves Fukudome, a guy who got about ended up in the same ballpark as Guillen in terms of price range. I'm definitely not condemning Guillen, because I think he will enjoy hitting at the K and that him being a good hitter at Safeco bodes well for his near future prospects. PECOTA might be missing low here, as in my estimation I think you could add fifteen points to those percentages and it would be a perfectly reasonable projection, but I know my opinion is not necessarily the prevailing one.

The Lost Sons

1B/DH Craig Brazell: .260/.300/.440 in 551 PAs, 1.7 VORP

For those who would say "I told you so," I'll retort "I know." It's not like I thought that Craig Brazell had suddenly become a star at age 27 in when he bashed 32 dingers while stilling having a middling .337 OBP, it was more like the idea that Brazell seemed to have earned at least a look what with the absence of any real first basemen on the team. To my way of thinking, there's really not much difference between Ryan Shealy and Craig Brazell except perception caused by the parks that they had played in before their Royals "careers." Shealy was in the Rockies system and got a Colorado Springs/Coors Field boost, Brazell was in the Mets systems and received no such boost. One was regarded as a blocked prospect, the other a journeyman. Now one is going to get an extended look at the first base job and probably has the inside track, while the other is just organizational fodder for whatever team has a hole at first base in AAA. Sometimes life isn't fair.

SS Angel Berroa: .250/.300/.260 in 343 mythical PAs, 0.0 VORP

Lip service. Berroa has next to no chance to ever land a starting job again, and he'll need to display some serious work ethic to get a look at a utility job. He's listed here primarily to make the snarky point that PECOTA projects him to outhit Tony Pena Jr.

LF Emil Brown: .256/.316/.386 in 384 PAs (OAK), -.2 VORP

I like Billy Beane a lot, of course, as he's one of the symbols of sabermetric success and a very smart baseball guy. I see no reason, however, to expect Emil Brown to be worthwhile to carry on a rebuilding A's team for seven figures when your average AAA schlub could do the same for 350K and would thank you daily for the opportunity to provide for his family. Sorry, that's a bit mushy. So's Vlademil's bat.

DH Mike Sweeney: .270/.330/.430 in 290 PAs, 5.4 VORP

PECOTA likes Sweeney more than Shealy, Huber, and Brazell. Unfortunately, that's not much of a compliment. You know the story here. We all love the guy, and he's completely done as a major league baseball player. If Sweeney is so inclined, it's easy to see him coaching in the majors until kingdom come.

C Jason LaRue: .220/.310/.268 in 151 PAs (STL), 2.1 VORP

Jason LaRue had the fifth-lowest batting average of the live ball era with a minimum of 160 plate appearances, "beating" only Jose Oliva's '95 season (.142), Vic Harris in '72 (.140), Chris Bando in '85 (.139), and the incomparable Ray Oyler in 1968 (.135). It figures that the Cardinals, who retread players quite often, are taking a shot at the ultimate retread here. And yes, I suppose the guy can catch and throw. He also is rumored to have all his limbs.

Our Favorite Topics

LF Mark Teahen: .280/.350/.430 in 539 PAs, 9.7 VORP

Well, there it is, and it seems like Teahen's projection is an approximation of what most of think he's capable of. The player comment in the book is quick to note that power in a young player generally isn't a fluke, so it's possible that .430 SLG is a bit low for Mark. Either way, the 52% improvement chance makes it seem as if even PECOTA is counting on Teahen regaining some of his power lost. Either way, we know Teahen has a broad base of skills and plays pretty good outfield defense. 2008 will go a long way towards determining whether he's a long term keeper or better trade bait.

1B Ryan Shealy: .250/.320/.430 in 259 PAs, 0.5 VORP

You have to take PECOTA's reasonably high breakout and improvement rates here with a grain of salt, because since Shealy hit a thundering .221/.286/.308 last year when ambulatory it wouldn't be hard for him to do better. I would love to tell you all that I'm high on Shealy because of his spring training bombs, but the fact of the matter is that a close look at his minor league numbers suggest that he really was never anything to be too enthused about in the first place, and he's no longer young. That projection above looks about right to me, and it's not enough to distinguish Shealy from Justin Huber or Craig Brazell. We all know that Huber's a long shot to produce for the Royals at this point. It says more about Shealy than about Gload that Ross is projected to be a significantly better hitter. Temper your expectations, hot start or no.

"I Spit On Free Passes"

SS Tony Pena Jr: .260/.290/.350 in 402 PAs, -2.5 VORP

PECOTA projects TP to have the same year as last. That's not the worst thing in the world, as while Pena is not at all a long term answer he's a good defense-first stopgap until the Royals can find someone who both hit and field to take his place. Until then, enjoy the defensive wizardry and the first pitch ground outs to second base.

C Miguel Olivo: .241/.273/.388

Solid overall on defense, Olivo's prone to defensive lapses sometimes and swings at everything in the same time zone while at the plate. He's a fine choice for a back-up due to his reasonable pop, but Buck could hit 50 points below Olivo and still be a better player. For the record, Olivo has had a large platoon split in his career (843 to 620) that Trey Hillman might want to consider when he decides how best to utilize his back-up catcher.

Gamers! Gamers! Gamers!

2B Mark Grudzielanek: .280/.310/.380 in 329 PAs, 3.1 VORP

If PECOTA's nailed this one, it's time to parlay Grudz's batting average and defensive reputation into a prospect at the deadline. He's projected to lose 23 points of EQA over last year, and his once-great defense took a slide last year. At 38 years old this season, there's more of a chance that nagging injuries will keep him from getting enough playing time to have the strong second half he did last year. This all said, I've learned not to write Grudzielanek off. He has the "Joe Randa quality" as someone who keeps on producing when you're sure he should have jumped the shark four years ago.

1B/DH Ross Gload: .300/.340/.450 in 363 PAs, 8.3 VORP

He's not good defensively, he doesn't have the power or the patience you'd want in a first basemen, and he's got limited upside. Yep! He's a Royals first base candidate. In Gload's defense, he makes contact and would make for a neat bench player on a contender. In fact, he's a neat bench player here. The problem, of course, is that he might actually be the best first basemen on the team if Butler can't adapt to playing the field, and if that's true starts, it's a "tallest midget at the circus" accomplishment. Even at the somewhat optimistic PECOTA line above he would be one of the least productive 1B/DH players in the league. Gload might contribute here and there, but there's always the possibility he'll fall off a cliff as a guy with little secondary skills who is slow and on the wrong side of 30. Gload will explode!

OF Joey Gathright: .280/.350/.340 in 296 PAs, 1.0 VORP

Gator is exactly what you want in a fifth outfielder, and not much--if anything--more. He's fast, should be able to swipe a base or two, makes good contact, attempts to work the count, and can handle all three outfield spots competently most of the time. The downside is that he's got no power and has virtually no chance of breaking out in that regard now that he's going to be 27 this year. There are a lot of people who consider Gathright a prospect...this is who he is, folks. If he plays good defense and keeps DDJ fresh, his employment by the Royals is warranted. Long term, dozens of other guys have some variation of the "Tom Goodwin skill set" so don't expect Joey to become a fixture in KC.

The Infield Sleeper

Alberto Callaspo: .272/.327/.372 in 392 PAs, 3.1 VORP

Being two years youner than Tony Pena Jr., and having some great OBP numbers in the minors in the past, Buckner-for-Callaspo good pay off as long as Callaspo can swing shortstop (he's done fine playing there in the past) and as long as his terrible line as a utility in Arizona doens't represent his true major league ability (it's very unlikely he's that bad). While Buckner might grow up to be a decent starter, this was not a bad gamble--a gamble, certainly, a but a low risk one--by Dayton Moore.

Buddy Bell's Cleaner Upper of Choice

Shane Costa: .280/.330/.420 in 342 PAs, 4.0 VORP

Personally, I feel bad for Costa, who has creamed AAA pitching but has been able to do next to nothing with his major league opportunities. Still, at a certain point--Costa is 26 this year--you have to wonder whether the guy will ever make the transition to the majors in even an adequate way. He somehow got 109 PAs last year without a single homer, which is downright Gathrightian and certainly not something that suggests that Shane is just an inch away from success. PECOTA likes him thanks to his minor league numbers, and I'll root for him...but I'm not expecting much.

The Young Studs

Billy Butler: .290/.350/.460 in 626 PAs, 20.1 VORP

That line would be downright sweet from the 22-year-old Butler. He's less than a year older than me, and he'd be an average DH with that line and a very good first basemen if he good get to the point where he could not kill the team defensively. Really, the question with Butler is just that: is he a lifelong DH, or can he learn to play 1B well enough to stand around there. Either way, he will hit, and he'll hit well. Butler has a 71 chance of improving according to PECOTA, and a  41% breakout rate. That breakout number is second on the team amongst position players, with the highest breakout percentage going to...

Alex Gordon: .270/.350/.470 in 551 PAs, 19.6 VORP

PECOTA is appropriately cautious after Gordon's rough transition to the majors, but second half stats correlate better than first half stats when projecting a player. That and the minor league track record, no doubt, cement the fact that PECOTA still loves Alex Gordon. With a 67% improvement rate, a 45% chance of breaking out, and improvement across the board in the second half, it's not wishing for too much to imagine that Gordon could beat that projection. Especially the slugging percentage. I think he can hit .500 this year, and that pretty soon we'll all have forgotten the tough April and May that scared us all in 2007.

Spreadsheet Baseball returns next week with PECOTA's harsh take on our pitching staff! Enjoy the ST action to come, and, as always, comments/questions are welcome/encouraged. Also, I fully recommend "the book" this year, as I always do. Happy reading!