clock menu more-arrow no yes mobile

Filed under:

Spreadsheet Baseball: PECOTA's Royal Projections, The Good, The Bad, and the Statisical Anomalies

Happy belated fourth to all. I've been stumbling through allergies the last couple days, as this is the first day in three I haven't woke up and immediately sneezed brain fluid out my nose. There's great imagery for you, eh? Moving swiftly on, you may have noticed we're now through eighty-five games of the season, officially over the halfway point and drawing ever closer to the All-Star extravaganza. In case you were hanging the edge of your seat about this, the Royals are not the worst team in baseball or even the worst team in the AL to date: Texas equally bad and Tampa Bay, mired in a ten-game skid, is 33-50. Even with last night's loss, the Royals are 7-3 in their last ten and a mere two games behind their fourth-place nemesis.

There is, of course, still plenty to complain about. I won't go into any of the specifics yet other than to note that we're still 36-49 and we'd have the worst record at home in baseball if it weren't for Krivsky's Follies. Nonetheless, Gil Meche has ascended to All-Star status, giving the Royals a worthy representative. Gil continues to spit in the face of those who thought he was a huge mistake.

Speaking of huge mistakes, this was a terrible choice for a segway. Anyhow, Since we're about as near to halfway through the season as we're ever going to be again in 2007, today's article looks at PECOTA's projections versus the actual performance of the Royals on the roster. PECOTA, as many of you already know, is the projection system used by Baseball Prospectus. Though it's highly recognized as one of the best in the biz, as with any projection system, there are some things that went very wrong, and others that are impressively on the ball this far. An important thing to remember is that no system is perfect, but some of the less impressive projections remind us that projecting the statistics of non-robot baseball players--by any means--can be a hazardous business.

Position Players

C John Buck: .243/.333/.536, 14.7 VORP

PECOTA Full Year: .253/.310/.416, 5.6 VORP

PECOTA missed low on this one, to be sure, but this wasn't exactly entirely unexpected. The whole "Age 27" thing might be overrated by some extreme sabermetricians (NOTE: this is not a new kind of snowboarding), but it is an age where a lot of players experience career years as they arrive in their physical prime. Clearly Buck has experienced a power spike of large proportions, and that's the most glaring issue between where he is now and where PECOTA expects him to be at the end of the year. There are reasons to believe, however, that Buck will finish well above his projection now: he's slugged 14 homers in 213 PAs, which is sweet, he's walked 22 times to last year's 26 and a projected 28 for this year, and he's shown the ability to hit for a considerably higher average. All these factor combine to convince me that PECOTA's weighted mean projection is off, and Buck has officially broken out. Perhaps PECOTA is exonerated, however, by the 54% projection that Buck will improve and 31% chance it projected that he would out-and-out breakout this year.

C Jason LaRue: .194/.266/.357, -2.1 VORP

PECOTA: .233/.314/.369, -.7 VORP

In his brief heyday, Lala survived by getting enough singles to fall in to keep his batting average in the mid-.200s and swatting a reasonable number of longballs to offset the fact that he always had a bit of a slow, strikeout-prone swing. Now that his batting average and what patience he had is dried up, we're left with a catcher hitting below the Mendoza line whose only visible skill at the plate is occasional power. That's okay for a back-up, given what people expect/don't expect out of back-up backstops, but it's not worth the money he's being paid.

PECOTA didn't have the extent of LaRue's collapse nailed, but give the projection system props for predicting he would fall below replacement level. PECOTA hardly ever projects anyone to hit this badly, so it's really just a case where you have to look at the mere 9% chance of LaRue breaking out from his 2006 663 OPS, and say that's basically as close as the algorithim can get to saying "this guy stinks." He's still a good enough defensive catcher and third basemen that we might be able to flip him for something reasonably alive.

1B Ross Gload: .259/.296/.398, -.0.8 VORP

PECOTA: .305/.353/.456, 9.4 VORP

Projection notwithstanding, everyone and his brother knew that Gload would be exposed if used as an everyday player. This isn't MVP 2005, folks, where the "batting average" rating means anyone who hit over .300, even in limited playing time, can be a successful starter for you. This is MLB, where bench players who play too much and strain their quads hit fifty points lower than they were projected to hit had they remained in their previous role. That PECOTA line is for 277 PAs, as the PECOTA expected Gload to be used as the role player he is. I mean, I guess he's better than Ryan Shealy. That doesn't mean he's a good hitter at all.

Gload has taken six walks in 115 plate appearances, which is just ugly. This is the problem with relying on a slow corner-type who is reliant on batting average to be good. Lose the BA, and you lose the value. Well done, Buddy.

UTIL Esteban German: .276/.365/.383, 6.0 VORP

PECOTA: .287/.361/.386, 10.4 VORP

Even when his batting average drops, German continues to provide good value via his plate discipline and occasional pop. While it was kind of a given that he wasn't going to repeat last year's stellar performance, PECOTA pretty much had him down to the letter. It expected him to get about 400 PAs which explains the slight difference between German's current VORP and half of the projected VORP, and those OBP and SLG numbers are so close that it's pretty cool. As I've been saying all year long, the Royals would get more value from 2B if German was the regular. He's also going to continue to be cheap.

2B Mark Grudzielanek: .267/.312/.417, 3.9 VORP

PECOTA: .285/.327/.383, 9.3 VORP

Again, PECOTA is dangerously close. Extrapolate Grudz's current performance over a whole season and the VORP comes out to right around the one-win mark that PECOTA projected. That said, it's fair to say that we're seeing the signs of Grudzielanek's inevitable decline; his batting average has dropped and his OBP is lower than usual, and no one's as impressed with your defense if you can't stop getting hurt. Grudzielanek is a Royal who probably won't be wearing blue for that much longer. If he's still here after the trade deadline, it'll come as a surprise to me.

Tony Pena Jr.: .275/.293/.356, -1.5 VORP

PECOTA: .247/.283/.341, -6.1 VORP

Pretty much the only thing that PECOTA missed was that Pena is good enough of a contact hitter to hit in the .270s or .280s when he's going well. Rightly, he was pegged as having virtually no plate discipline without any impressive power to compensate. Now yes, he's good defensively, and he's not "the problem" but it's damn hard to hit .275 and have a sub-.300 OBP, folks. He's 26 this season with no good minor league track record, so he's not "the solution" either.

Alex Gordon: .231/.323/.361, -.1 VORP

PECOTA: .282/.363/.509, 35.7 VORP

If there's anything more encouraging this season than Gil Meche moving up out of the "mostly harmless" category of starting pitchers, it's that Gordon has recently flashed the ability that everyone was pretty sure he would display from day one. Keep this in perspective: PECOTA's going to miss low when the year's over, but so is virtually everyone else. It's not often that someone is the consensus number one prospect in baseball and then worries the hell out of everyone for so long.

Gordon's slow start means his batting average is probably "ruined" for the year, as it'd be surprising to see him get up to that .282 projected by PECOTA, but his June progress hints that he's come around finally. A .265/.350/.450 finish would satisfy me after the struggles in the first two months, and I think Gordon's very capable of reaching that line.

LF Emil Brown: .225/.288/.315, -7.0 VORP

PECOTA: .279/.346/.438

I'm not getting any particular glee out of this, but it's my duty to tell you that Brown's bat has done died, folks. For all the talk about Brownie being a good second half player, he's 2-for-29 with four walks in his last ten games and he's plummetting back towards the Mendoza line. Hopefully some desperate team will see Brown's track record and think he's not done, but I'm not a desperate team and I say that a 603 OPS through June is plenty evidence that the guy's not worth giving any more playing time on a team that's trying to rebuild.

PECOTA didn't really see this coming, forecasting another modestly productive year from Emil. Nonetheless, the thing that a lot of people forget about players who break on the scene late is that they won't necessarily stay late. Call it Lew Ford Syndrome.

CF David DeJesus: .288/.369/.424, 19.7 VORP

PECOTA: .290/.364/.446

I need to stop flip-flopping here, though David is yanking my chain all over the place. He started off hitting like Pujols, then hit like Neifi Perez for a month, and is now hitting like...himself. Resist the temptation to see DeJesus as a trading chip, as he's leading the team in offensive VORP and even if he stays what he is now...he's still good, and signed to a cheap contract. PECOTA would like everyone to know that it has David down pat, and that it thinks that DeJesus am what he am: his breakout percentage was only 6%. Myself, I remain optimistic for a couple extra dingers in the second half. I'm sorry I doubted you, David.

RF-1B-3B-Bat Boy Mark Teahen: .290/.368/.416, 13.9 VORP

PECOTA: .283/.357/.474, 25.5 VORP

Aside from not being a Russian Czar or a Circus Clown, Mark Teahen is not a first basemen. That Buddy Bell seriously thinks it's a good idea to take a guy who was perfectly good at third base and then right field (READ: he's a good defender at position harder than first) and put him at the least-demanding fielding position possible is a joke that's not particularly funny. Teahen was fine in right, and his bat is plus-average there as opposed to thoroughly mediocre for a first basemen.

Speaking of his bat, Teahen's power numbers have been down this season but I'm here to tell you Teahen fans to take heart: the lad is 25, has already taken only five less walks than he did all year, and is on pace to equal his weighted mean projection even if his slugging percentage doesn't come up. While .416 isn't a good slugging percentage for a corner guy, Teahen is still a developing player and I am optimistic that his power will come back up as he adjusts. Teahen came out of the blue last year to rake pitching in the second half, and it's clear that pitchers have adjusted a little to him. If can be a little more aggressive when he gets his cookies (this take n' raker is rake a little too little) then I don't see why he can't match that .474 number that PECOTA projected. In any case, I'm encouraged by the broad skillset Teahen's shown on offense. Fun fact: he's now 25-for-30 in stolen bases for his career.

LF-CF Joey "Gator" Gathright: .317/.406/.367, 3.8 VORP

PECOTA: .273/.340/.350, 3.1 VORP

He's not a starter, and it really is that simple. Yeah, he's better than the declining version of Emil Brown, but so is Shane Costa. Here is the problem with players like Gathright who have no power:

1. They MUST have a good OBP to be successful

2. OBP is largely a product of walk rate, which is largely a product of the batter being a reasonable threat to smack the ball if the pitcher grooves one

3. Players like Gathright never smack the ball

4. Pitchers then have no incentive to walk them

5. The walk rate comes down, and without power...

6. The only thing the player can do is hit for a high average

7. And players who rely on batting average have all their value totally collapse when they're hitting below .280 or so.

And when you can't slug over .400, you shouldn't really even be considered for a corner spot because those are positions where you need some sock. That .400 is in probably the best 73 PAs of Gathright's career, and they're in pretty much direct opposition to his line of .256/.327/.309. He's a fine role player, but not the kind of guy you want starting.

OF Shane Costa: .185/.194/.231, -7.2 VORP

PECOTA: .288/.331/.447

PECOTA must have been high when it projected this line for Costa. Shane probably isn't as bad as his numbers thus far, but he definitely isn't good enough to give the PAs to him to find out. He's back-up at best, and he is probably the worst hitter to hit clean-up so far in 2007. It's pencilling guys like this into the line-up that makes me worry about Bell's sanity.

Hitter Billy Ray Butler: .274/.289/.425, 0.0 VORP

PECOTA: .295/.347/.454, 14.4 VORP

His plate discipline hasn't translated to the majors yet, and he's hitting .274 instead of .295 which accounts for the difference in SLG as the IsoP is right on the dot. Other than that, Billy Ray is looking a lot like PECOTA projected. The plate discipline thing isn't really surprising as a lot of young players take less walks in the majors than they did at AAA, but it is a concern as Butler has walked once in 70-odd PAs. Nonetheless, his ability is considerable (his six-RBI game was cool), and I'd bet his free-swinging will decrease a little once he gets his at-bats consistently for a while. It seems like, for some reason, Bell would rather play LaRue at third in a chicken suit and DH Teahen before he gives Butler a week's starts in a row. All this said, Butler could easily get to the projection above by the end of the year.


SP Gil Meche: 3.26 ERA, 116 IP, 84 K, 32 BB, 11 HRA, .25.9 VORP, ALL-STAR

PECOTA: 5.37 ERA, 171 IP, 125 K, 72 BB, 22 HRA

Meche extends a curveball-twirlin' middle finger to all those who doubted his ability to be a front rotation starter, including yours truly. Contrary to what everyone expected moving out of Safeco, Meche's home run rate has gone down to .85 per nine innings, a marked improvement from last year. Perhaps equally important is his walk rate decreasing from over four per nine to two and a half per nine. This ensures that when Meche does give up a tater (less likely these days), there should be less baserunners on average. The durability he's displaying is impressive as well, especially from a guy who once had labrum surgery, and Meche looks poised to reach the 200 innings plateau for the first time in his career.

What does all this mean in relation to his projection? Well, it just goes to show that you can't just say Mediocre Pitcher + Hitter's Park = Bad Pitcher. You have to consider that some pitcher's breakout later than others, and individual skillsets sometimes translate to different parks and situations in ways that aren't necessarily linear. Meche has clearly established a new level of performance in Kansas City and, while it's too early to call him a definite success, he's done very well. I do have to wonder if his improvement has something to do wiht being another year removed from surgery: there isn't much information right now on what to expect years down the road from a torn labrum patient because not many of them hang around.

SP Brian Bannister: 3.71 ERA, 80 IP, 43 K, 24 BB, 4 HRA, 13.8 VORP

PECOTA: 5.77 ERA, 107 IP, 60 K, 45 BB, 15 HRA, -0.5 VORP

Another pretty significant miss by PECOTA on the pitching side of the ledger. Bannister is probably due to regress over 4.00 ERA, but he's clearly better than expected. He has kind of an inconsistent record when it comes to allowing home runs, and a big part of his success thus far is the fact that he's only allowed four so far and currently has a home run rate almost half of Meche's. His low strikeout rate suggests that perhaps he's been a little lucky so far this year, as does his .275 BABIP number (it's usually around .300 for most pitchers), but Bannister seems to have established himself as a legitimate major league pitcher. Given his shaky trial with the Mets last year, that's pretty cool. Why did PECOTA screw up here? Well, Bannister's low minor league walk rates have come with him to the majors this time, and that home run rate is pretty low given his history. Apparently those factors threw PECOTA for a loop, though Bannister had a reasonably optmisitic 38% improvement projection.

SP Jorge de la Rosa: 5.27 ERA, 100.2 IP, 58 K, 34 BB, 13 HRA, 3.1 VORP

PECOTA: 5.78 ERA, 97 IP, 72 K, 56 BB, 12 HRA, -0.7

That's more like it. While PECOTA is pessismistic here, the fact is that Rosy has a veyr average K-rate, a nothing-special walk rate, and gives up too many home runs and hits. Last year he wasn't this hittable, but he walked waaaaaay too many batters. This year, he can throw more strikes but he's given up 124 hits in 100 IPs. That might be bad luck, and it might be that de la Rosa doesn't have the stuff to avoid both walks and hits to the point where you want him in the rotation. Whatever the case, the home run rate has to come down one way or another or he's living on borrowed time. Nonetheless, he's shown enough flashes that there's plenty reason to keep him around to see if he's worth a rotation spot for next year.

SP Odalis Perez: 5.56 ERA, 90.2 IP, 44 K, 33 BB, 10 HRA, 2.5 VORP

PECOTA: 4.97 ERA, 132 IP, 77 K, 34 BB, 17 HRA, 11.0 VORP

He's history as an effective starter, end of story. This is now two year in a row where Odalis has given up an above average BABIP and three where he's given up more runs than DIPS would predict, to the point where you just have to think that his stuff doesn't fool enough hitters to get by anymore. His home run rate and walk rate are both similar to de la Rosa's, his K-rate is worse, and he has none of the upside that Rosy has. He'll take the ball and struggle through five or six innings for you, but not much more. If the Royals or another team want to try and salvage anything from him, a move to the pen is in order.

PECOTA didn't seem to think he was done, but recognized he was in decline. Obviously, the projection system didn't recognize the BABIP as a running problem, but it does suggest that he is going to lose his rotation slot some time this year. With an ERA that fills one corner of clubhouse by itself, that seems likely to come true.

SP John Thomson

Not enough data here to really bother with PECOTA, so I'll just say that I really hope the Royals can flip him at the deadline. Otherwise there's not much point to acquiring a back-of-the-rotation starter to run on every fifth day for a last place team. In fact, there's no point when...

SP/RP Zack Greinke: 4.52 ERA, 67.2 IP, 55 K, 18 BB, 8 HRA, 9.4 VORP

PECOTA: 5.44 ERA, 121 IP, 77 K, 35 BB, 20 HRA, 4.7 VORP consider that this guy is in the bullpen. Not that Greinke hasn't done well in the bullpen, but I'd like an explanation for his short leash (moved to the pen after seven starts) when de la Rosa and Odalis Perez remain in the rotation with ERAs above five and John Thomson is signed off the scrapheap. PECOTA obviously through Greinke would have trouble this season, but when you consider the tough times Greinke went through in '04-'05, PECOTA really didn't have much to go on. Zack still possesses a k-rate above seven, a servicable walk rate, and gives up the home run per nine that you'd expect from a guy who doesn't throw that hard and who isn't really a ground ball pitcher. Thing is, the upside of using Greinke in relief is so much lower than what he could be if turned out to be a mid-rotation starter. Free Zack!

RP David Riske: 2.29 ERA, 39.1 IP, 31 K, 16 BB, 4 HRA, 14.9 VORP

PECOTA: 4.78 ERA, 38 K, 20 BB, 8 HRA, 6.0 VORP

PECOTA and I were pretty damn sure that Riske was regressing to the land of the fungible relievers, but David has had other ideas recently. He last gave up a run on June 15th, and his ERA has come down to the point where some contender should be willing to part with a reasonable commodity in exchange for him at the deadline. He stills walks his fair share of hitters and hasn't exactly been homer proof, but his K-rate that was missing the last two years is back over seven. That's enough that it makes him a desirable reliever for a contender. Props to you, David.

Capital "C" Octavio Dotel: 3.94 ERA, 16.0 IP, 19 K, 6 BB, 2 HRA, 3.4 VORP

PECOTA: 4.85 ERA, 41 K, 20 BB, 40 IP, 7 HRA, 4.0 VORP

Maybe he's lost a few MPH off his fastball now that he's 33, but he's the same pitcher he's always been; he can smoke anyone at the plate, but his aggressiveness within the strike zone leaves him prone to giving up a bomb or two more than you'd want out of a relief ace. Still, he's shown that he's back from his surgery, and he's got that name brand value that GMs love when the trade deadline rolls around. Dotel could haul in a pretty good prospect for KC without ever actually contributing anything meaningful during his Royals tenure. PECOTA was understandably pessismistic of a guy who looked like he was back a little too early from TJ surgery last year.

RP Joel Peralta: 3.54 ERA, 53.1 IP, 46 K, 12 BB, 5 HRA, 11.8

PECOTA: 4.69 ERA, 59 IP, 46 K, 18 BB, 9 HRA, 7.5

Peralta is on the verge of turning in a second solid year in a row in Royal Blue, despite a shaky couple weeks early on. Interestingly, he's walking a few more batters than last year but also allowing fewer home runs. Couple this with a nice K-rate, and it portends well for Peralta's future. Downsides are that he's already in his thirties, and that if he continues to pitch this many innings he'll be totally burnt out for next year. It's pretty much a toss-up which way the Royals want to go with this guy in terms of keeping or trading.

Closer Joakim S-O-R-I-A: 2.27 ERA, 35.2 IP, 39 K, 14 BB, 0 HRA, 13.2 VORP

PECOTA: 4.84 ERA, 54 IP, 41 K, 22 BB, 8 HRA, 5.4 VORP

If you can pitch inning after inning without allowing any home runs and couple that with a great strikeout rate, you're badass. And that's what Soria is: a live arm who Ks more than a batter per inning without allowing a home run yet, a young guy with plenty of upside to possibly start in the future, and a gamer who refuses to let those peski batters get on base. Give PECOTA a break, as Soria was in A-ball last year: it's the Padres who get left with the dunce cap for letting this guy into the Rule 5 Draft. If Buddy Bell shreds Soria's arm, he must be euthanized, not fired.

I apologize to Jimmy Gobble's three fans, but I must head off to work for the next four hours. The end result seems to confirm what we already know, that PECOTA can be deadly accurate and there is usually a good reason if it's off. Gil Meche...well...there's always some things you can't explain with just the stats, and that's why scouting and statistics have to be paired together to be most effective.

Spreadsheet Baseball returns next week, hopefully on the correct day. For this week, comments are, as always, welcome/encouraged. Happy reading.