Sunday, June 11, 2006

Who Woulda Thunk It?

Powered by Intern Sam

There’s been a lot of talk on this blog and others in the last few weeks about the status of Ron Gardenhire, and whether the Twins’ struggles have him on the hot seat with Terry Ryan. Opinions seem to range widely, all the way from “he’s been useless from the beginning, and the Twins won those three straight division titles in spite of him, not because of him” to “leave the poor guy alone, Terry Ryan screwed him with his offseason moves, and besides, there’s nothing wrong with managing by instinct.”

There are even a few people prowling the discussion boards (and TwinsGeek’s comments section) insisting that, while a few gentle second-guesses might be all right, Twins fans are shaming themselves when they become noticeably frustrated over the judgment of the manager of Our Only Base Ball Team. (This seems like an odd line of reasoning to Intern Sam, mainly because the sports blogosphere is not exactly known for employing a “We’ve Got Spirit, Yes We Do!” style of fan engagement, so it would seem odd for anyone who prefers simply to root, root, root and ignore the gaffes to even be reading the blogs in the first place. But to each his own…)

The fact is, though, it can be awfully hard to judge the performance of a manager from an outsider’s perspective, since most of the moves he makes in a given season aren’t quantifiable in any useful way. To sit down and separate out just how much of the blame for the Twins’ disappointing first half should be laid at Gardenhire’s feet is an impossible task given the statistical tools currently available to us. After all, Gardenhire doesn’t select the players on his team (as has become blindingly obvious in the last week or so,) so can he be held even partially responsible for the struggles of Rondell White, Tony Batista, and Brad Radke? Possibly, but we don’t know for sure.

What infuriates many Twins fans is really Gardenhire’s apparent unwillingness to see the game the way we, the fans, see the game, and to react accordingly. That’s not to say that he’s right and we’re wrong (or vice versa,) just an attempt to put the whole situation in some kind of perspective. (All that having been said, Intern Sam still comes down most frequently on the side of those who say that the Twins could do better than Ron Gardenhire, but he isn’t foaming at the mouth about it.)

Still and all, the Twins remain a club divided against itself, not in the sense of team chemistry, but in the sense that there are now clearly two sets of 2006 Twins: the good ones and the bad ones. We all know who belongs in what set, and it’s up to Terry Ryan to decide, for instance, how much longer Rondell White will be allowed to continue wasting a spot on the roster. But when a season goes south as soon as this one did, with a core of players as talented as the Twins have, there is always a temptation to chalk the whole thing up to bad luck and bad timing. After all, no one dreamed that White, always a steady near-.300 hitter, would have the meltdown of his career the minute he stepped inside the Dome, right? And Carlos Silva? Who would have imagined, after last season, that Silva would be the first guy yanked from the rotation?

Well, as it happens, there are a number of prominent statgeeks who have waded into the nasty business of player performance projection in the last few years, and this weekend, as Intern Sam relaxed in the Upper Club at the Dome, secure in the knowledge that at least Francisco Liriano is bulletproof, he began wondering how much of the current Twin misery was actually foretold by the so-called experts. To answer the question, we’ll turn to arguably the two most well-respected sabermetric publications in the biz: Baseball Prospectus, and The Bill James Handbook, both published annually. The authors of these tomes are not exactly colleagues (and anecdotal evidence suggests that they don’t always play well together,) but that should only help us, in that we’ll be looking at two independently reached conclusions.

Here’s how we’ll do it: below, I’ve listed the six regular players and starting pitchers who constitute the core of the Bad Twins team, along with their 2006 stats through June 11. Listed alongside are the 2006 projections for each player from the 2006 editions of both the Bill James and BP annuals. (One exception: since both annuals were sent to press before anyone knew Tony Batista would be playing in Minnesota in 2006, I’ve used the predictions from the 2005 annuals. This isn’t ideal, of course, but nothing Batista did in Japan in 2005 would have had much of an impact on his numbers, as best I can tell.)

Most of the stats are self-explanatory, but I’ll clarify two less common ones upfront. Bill James uses a stat known as “Runs Created” (RC), which purports to quantify how many of the runs a team scores over the course of a season are attributable directly to that player. I won’t go into the math here, but this is a fairly well respected (if limited) statistic, and it was from the RC stat that James eventually crafted his far more complex (and controversial) Win Shares metric.

On the Baseball Prospectus lines, Value Over Replacement Player (VORP) is included as well. This is a wildly complex stat involving ballpark effects, positional variances, and the effect of the tides on Cristian Guzman’s hair, but all you need to know to get the gist is that VORP is an attempt to measure, by runs per season, how much better or worse a player is than a fictional player whose skills exactly match those of an average replacement player available to an average major league team at minimal cost. Players above zero are better than that average, those below zero are below that average. (There are people who hate this stat, partly because it implies that an average major league ballplayer is worth nothing, but I feel that this is a largely semantic argument in an arena where semantics are meaningless, so I’m using it.)

Why go through all this silliness just to tell us what we already know – namely, that a lot of Twins are underachieving this season? Quite simply, as summer swings into full gear, Intern Sam would like to know once and for all whether anyone (or everyone) should have seen the train wreck coming a mile down the tracks. Let’s have at it…

Rondell White

2006 Stats: .190/.209/.224; 0 HR; 4 BB; 29 K
James: .284/.337/.461; 19 HR; 33 BB; 79 K; 75 RC
BP: .292/.338/.463; 13 HR; 23 BB; 57 K; VORP=15.7


Okay. It’s obvious that no one thought White would be the out-making disaster that he has been. Neither projection system thought he would be as strong as he was in 2005 (when he hit .313 and with an OPS of .836) but to say that Terry Ryan should have known this pick-up wouldn’t work out is unfair. The plan was for White to stay healthy by staying out of the field, and for him to hit something like the average of his career stats, which is what both James and BP projected him to do in ‘06.

It’s worth noting that the BP annual did have White listed as a 40% chance to collapse in 2006 (in BP-speak, a collapse is a 20% or more slump in overall production,) but that’s a marginal measurement at best, and given the Twins’ needs measured against their limited budget, we can’t really blame anyone but the baseball gods for what’s happened to Rondell.

Tony Batista

2006 Stats: .236/.303/.388; 5 HR; 15 BB; 27 K
James (2005 ed.): .245/.284/.451; 30 HR; 32 BB; 90 K; 75 RC
BP: .253/.301/.428; 18 HR; 31 BB; 76 K; VORP=4.5

This is a bit tougher to figure. Both projections are in the ballpark as far as batting average and on-base percentage go, but neither forecasted the power drop. Of course, as I mentioned, these numbers are intended for the 2005 season, which Batista spent in Japan. I wish we had fresher numbers to choose from, since I suspect both models would have docked Batista considerably for having not seen major league caliber pitching in 18 months.

It’s worth noting that while Bill James sees White and Batista as equals as far as creating runs, BP’s VORP ranks Batista as barely better than its imaginary replacement player. Subtract the homers, and he’s surely well below average. So should we have seen this coming? Well. We did, didn’t we? This was the most controversial move Terry Ryan made in the offseason, and in this case, it appears that the fans were right, and the GM goofed.

Juan Castro

2006 Stats: .234/.261/.312; 1 HR; 6 BB; 22 K
James: .241/.277/.355; 7 HR; 19 BB; 64 K; 38 RC
BP: .255/.287/.379; 4 HR; 9 BB; 32 K; VORP=1.0

The experts had it nailed here. Castro is basically your textbook replacement player, and nothing in his career prior to 2006 suggested that he would be even remotely useful as an everyday shortstop. It’s worth noting that both annuals made their predictions assuming that Castro wouldn’t be an everyday shortstop in 2006, because of course, the Twins had…

Jason Bartlett

2006 Stats (AAA): .306/.328/.445; 1 HR; 8 BB; 26 K
James: .273/.343/.386; 7 HR; 47 BB; 67 K; 61 RC
BP: .271/.340/.394; 7 HR; 38 BB; 61 K; VORP=17.3

Now, on the surface, this may not look like a huge deal, numbers-wise. After all, even if the Bartlett predictions were dead on the money, it’s not like the Twins decided to replace Derek Jeter with Cristian Guzman when they kept Castro and demoted Bartlett to Rochester. But the truth of the matter is that Bartlett is 26, and the Twins need to find out if he can be their everyday shortstop of the future before he either gets frustrated and quits, or gets so old that he loses all trade value. Furthermore, there’s just no quantifiable reason for anyone on the Twins to have actually thought that Castro would put up better numbers.

Defense, you say? Baseball Prospectus’s rankings, which take both range and accuracy into account, say that Bartlett is three times the defensive shortstop Castro is. (The fact that he made a few errors last year is countered by his clearly superior range in the field. The laser-like focus on occasional errors to the exclusion of all else is the same kind of backwards attitude that causes Dick and Bert to spend hours extolling the ability of Tony Batista to consistently make good plays on “balls hit to him,” ignoring the obvious fact that he has a total range of about four feet to either side and is therefore a defensive liability.)

The bottom line here is that someone (Gardenhire?) clearly has a problem with Jason Bartlett that has nothing to do with his game, and no one is saying what it is. We’ll probably never know, but just for the sake of history, let’s not forget that the last time a Twins manager ran a middle infielder out of the organization because he just didn’t like the cut of his jib, we got five years of Luis Rivas while Todd Walker batted .291 and thumbed his nose from afar. And if that’s not enough to convince you that Castro over Bartlett is an error of serious proportions, consider this: in Baseball Prospectus’s list of player comparables, Bartlett is listed as being somewhat comparable to Jerry Hairston. Castro is Billy Ripken.

Brad Radke

2006 Stats: 4-7, 6.17 ERA; 22 BB; 38 K
James: 10-10, 3.87 ERA; 28 BB; 108 K
BP: 11-10, 4.17 ERA; 28 BB; 104 K

Look, we all love Bradke. He’s been a workhorse for as long as most of us can remember, and Lord knows, he was good when practically no one else on this team was. But listening to Twins fans talk about Radke in the offseason always makes me feel like I’m back in college in the late 1990s, listening to Cleveland Indians fans explain to me how their team was finally going to get over the hump this season while conveniently ignoring the fact that their starting staff was still anchored by Charles Nagy.

Now, no, the Indians didn’t have Johan Santana, and yes, Radke had a great year in 2004 (despite his 11-8 record,) but the numbers said fairly clearly that Radke wasn’t going to be anything special this year, and even now that he appears to have turned his season around somewhat, he’ll probably finish not too far off from what’s predicted above. And that’s the best-case scenario. Am I saying the Twins should have dumped Radke, or found a trade in the offseason? No. I’m saying that it wasn’t hard to foresee that he is no longer a legitimate #2 starter, and certainly not a useful top-flight starter for a club hoping to contend.

Carlos Silva

2006 Stats: 2-8, 7.73 ERA; 10 BB; 22 K
James: 9-10, 4.11 ERA; 27 BB; 70 K
BP: 10-11, 4.47 ERA; 28 BB; 70 K

Kyle Lohse

2006 Stats (ML only): 2-4, 9.22 ERA; 22 BB; 27 K
James: 8-13, 4.78 ERA; 47 BB; 109 K
BP: 9-10, 4.67 ERA; 42 BB; 87 K

The Baseball Prospectus crew says this far better than I ever could:

“[The Twins haven’t] really adjusted to the fact that their rotation isn’t good enough. While the Sox are making a point of improving their rotation… the Twins are playing make-believe, and continuing to crowd the rotation with the same good-enough guys who are no longer good enough. They’re hoping Brad Radke can be the pitcher he was in 2004 [and] wishcasting that Kyle Lohse might build on his 2003, instead of recognizing that was a good year from a fourth starter… And in doing so, they’ve crowded out their pitchers with real up-side, making only a single rotation slot open to Francisco Liriano, Scott Baker, J.D. Durbin, and perhaps Boof Bonser.”

That’s kind of the 2006 season in a nutshell, isn’t it? The fact that it was written sometime in January just makes it all the more depressing. Because if the question is whether anyone in the Twins’ brain trust should have known that what they had wasn’t going to be good enough to compete in this year’s Central Division, the answer, clearly, is yes. That doesn’t mean there would have been any easy answers regardless, but at least by acknowledging that the chance of contention with this team was slim to none, the front office would have freed itself to spend the year finding out who the real major league players in the system are, and preparing to contend in 2007.

Instead, they’ve clung to mediocre veterans like Castro and allowed White and Batista to continue to flail away despite knowing for a fact that none of those players will figure in the Twins’ long-term future. Meanwhile, Bartlett rots in Rochester and guys like Kubel and Baker get jerked around just when they should be getting a chance to adjust to the majors.

I know this is a buzzkill of a column, and I apologize. The Twins took two out of three from Baltimore this weekend, Joe Mauer has apparently become the God Of On-Base Percentage, Liriano and Santana continue to dazzle, and there are plenty of reasons to keep going out to the Dome this season. This isn’t a Kansas City situation- the Twins have some unbelievable talent on the major league level, and a full-scale rebuilding shouldn’t be necessary to get the club back to playoff caliber. But the current strategy of signing (and keeping) other teams’ cast-offs in the hope that they’ll be magically transformed into the Boston Red Sox isn’t working anymore. And the numbers say that someone should have known better.

13 comments:

John said...

Sam, great column, but I don't think the data supports your conclusion. Let me try and summarize the findings:

White - nobody saw that coming
Batista - as expected
Castro/Bartlett - manager error (though we're tryin to compart AAA stats)
Radke - shouldn't have expected great
Silva - ditto
Lohse - ditto

I think where the conclusion falls down is on the last three. The Twins didn't NEED those three to be gangbusters, as BP suggests. All three should fall squarely under the NO WAY was this expected, because if they were anywhere near their projection, the Twins have at least another 3 wins for each one. They've been that bad.

So by my count, I get

4 - now way we could possibly know
1 - manager error (castro/bartlett)
1 - GM err... no hold it, I'm going with:
1.5 - manager error and .5 GM error. The reason Batista is on this team is because Gardy won't play Cuddy at 3B. Obviously (I don't buy that is why Cuddy is hitting.)

That' still just a lot more bad luck than bad management. It's easy to be frustrated with Gardy's blind spot on the left side of the infield, and we should let him off the hook about it, but mostly this season is about some spectacular collapses, ones that neither we or thers could foresee.

Anonymous said...

Actually, John, I think you're being a bit generous regarding the pitching staff. I think it's quite a reach to suggest that, had Silva and Lohse reached the projections I laid out, that each would have three more wins than they do. In a number of cases, Lohse and Silva haven't even come close to winning, so much of their inflated ERAs result from runs that didn't actually matter in the outcome of the games they were scored in.

I agree that we can't hold anyone responsible for Rondell White, but just because everyone but Ryan and Gardy knew Batista was awful doesn't absolve them of repsonsibility for the signing, especially when they had other options, both in-house and on other clubs. Castro/Bartlett is a similar error.

So by my count, that's two major personnel errors, one that no one could have foreseen, and three mediocre pitchers being counted on to start 60% of the Twins' games, which falls somewhere in between egregious mistake and faultless mishap. Let's assume that Santana wins 20 this year, and let's be generous and say that Liriano could win 15. That's 35 wins, and we'd need another 55-60 to compete for the division. There is just no way that Radke, Silva, and Lohse were ever going to be a good bet to make up that difference, even if the bullpen was solid as a rock, which it isn't.

I'm not saying that the Twins couldn't have competed this year, had absolutely everything gone right. I'm saying that the front office fairly obviously decided to sit back and hope that everything went right, and in so doing, they became largely responsible for what happened when things began to go wrong.

John said...

I'll tell you how I came up with 3 wins apiece. Each one has an ERA that's roughly 3 runs higher than anticipated. I estimated (probabaly incorrectly) that those three pitchers started about half the Twins games, or about 30 games. That's 90 runs which (I think) is about 9 wins. It's not perfect, but it's not a wild guess either.

It also drives home a point that's important. We're used to thinking of a pitcher as "bad" when they have an ERA of 5. These guys haven't been bad, they've been gruesome. They've been so bad that the blow our expectations of badness. They've been so bad that we underestimate just how bad they've been.

So we say "Well, they're bad, but we didn't think they'd be great." No, that's not right. They're unfathomably bad, and they wouldn't have given the 27 Yankees a chance to compete. Batista is a bucket of water out of Lake Superior compared to these guys by any objective metric.

Let's make it objective. BP, at the beginning of the year, even with what they said about the overall staff, picked the Twins to win the AL Central this year. And they knew about Batista, and Bartlett isn't worth more than a couple of wins in their metrics. The difference has been the other four guys.

I think you could even make an argument that the other Batista/Castro/BArltett don't even necessarily belong on this list, certainly not in terms of degree, and maybe not even more than someone like Crain, Baker or even Kubel.

Lastly, I'm not sure how you an justify the piece about "sitting back and hoping". The Twins were all about keeping options around. That's why Lohse and Batista are on the roster. The problem is that they just don't have the depth to fill in all the (mostly unforeseen) implosions. Nobody would.

Again, I'm probably coming off to pro-management on this. I think the Gardy/Bartlett thing is criminal. I've been convinced from Day 1 that the Batista signing was Ryan trying to show Gardy that Cuddyer really is the best option there. And I'm starting to wonder if a change wouldn't make this team a half dozen games better in the standings.

But I still think THE big cause is the four implosions that nobody could foresee. Anything beyond that - including Batista and Bartlett - is small potatos.

John Sharkey, Esq. said...

I disagree quite strongly with some things that went into the construction of this team, but I have a hard time blaming management for the starting pitching. Silva was obviously due for a regression, and Radke is in decline. Lohse was generally viewed as a live arm, and respectable 4th or 5th starter.

If those three guys are simply league-average, or even replacement level, things would look differently. Look at the current VORPs:

Radke: -4
Lohse: -13
Silva: -14

Combined, those three starters are 31 runs worse than replacement level. Let alone league average, they're crappier than guys making $400,000.

When you couple those huge negative numbers with White's -21 VORP, you have nearly all of this team's failings. The VORPs on Castro, Ford, and Batista are -6, -5, and -4, respectively. The Twins have better in-house options than those three, of course. But say we throw Cuddyer at third, Bartlett at short, and Kubel in right. Assume Cuddyer puts up the same +14 (2nd on the team), and Bartlett and Kubel are 2 or 3. That's a nice gain, but not earth-shattering, and we still need a DH.

In retrospect, TR should have rolled the dice with one of the Thomas/Nomar/Thome trio; that move hurts. But White was seen seen as a decent alternative, and the left side of the infield has been about as bad as expected. The best we can hope for is finishing the season with a Santana/Liriano/Baker/Bonsor/Radke rotation, to get that front four settled, then add another of the young pitchers to that #5 spot in the spring.

Either way, though, the problem comes down to TR's problem of being reactive, instead of proactive, in dealing with players that have a potential to decline. He's so risk averse that he's hurting this club's future to compete; that will have to change (starting with how he deals with Silva/Hunter/Stewart/possibly Castillo).

John Sharkey, Esq. said...

Oh, I was going to say:

Part of the error in dealing with the younger talent on this squad has to do with future years, not just this one. It would, of course, be more desirable to win 80 games with the young lineup than 82 games with the old guys (Castro, Batista, et. al.). I think that's been the most harmful thing: this team wasn't going to win anything the way the starting pitching collapsed, but there was (and still is, really, but I see no indication that a Bartlett move is forthcoming) a chance to get the young guys extended time to help with planning for the next few years. That blown opportunity is more harmful that a few wins lost this year.

SBG said...

One thing that is missing from your discussion about pitching is the horrific defense being played behind the pitchers. The Twins have the best FIP (field indepedent pitching) in the American League. They also have the worst, by far, defensive efficiency in all the majors and could have the worst in the last 45 years (which is as far back as BP goes). And why is this defense terrible? Castro, Batista, Stewart, primarily. The decision to put those guys on the field have cost not only the offense tremendously but also the defense (or if you prefer, the pitching).

Anonymous said...

This looks like a good opportunity to ask a question that's been bugging me about Lohse: what accounts for the difference between the outstanding stats he racked up in spring training and his AAA stint and the terrible performance he's had with the Twins during the season? Is this mostly the difference in batter talent and effort, or is there something psychological/emotional going on inside Lohse? From here in Washington, DC, I can't tell, so I'd love to hear a more informed perspective on this.

Anonymous said...

The short answer, anonymous, is that no one really knows why Lohse has only rarely been able to get himself together at the major league level. We've all heard for years that he has "the best stuff on the staff" (although that particular bleat has died down considerably since the emergence of Santana and Liriano,) but that stuff has never quite coalesced into usable pitches.

Lohse is certainly a legit major league pitcher, but many observers far more knowledgable than I have said that he's really better suited to a relief role, where he could come in and blow through three or four batters, then get out before anyone has time to make the adjustments they always seem to make in his starts. Apparently, however, Lohse has made it known that he has no interest in being anything other than a starter.

The sad fact is that the Triple-A level is full of guys who are better than their minor league competition, but just not capable of cutting it in a serious way in the big leagues. It's always a shame, and you've gotta feel for guys like that. It's probably too harsh to tag Lohse as a "Quadruple-A" guy, but he's not far off from that mark.

Anonymous said...

Hear that rustling sound, mnpundit? That's the bone-chilling sound of the Moneyball crowd gathering up Billy Beane bobbleheads to beat you with for the heinous crime of implying that defense is as important as offense and pitching. Run while you still can...

Anonymous said...

One error:

"There are people who hate [VORP], partly because it implies that an average major league ballplayer is worth nothing, but I feel that this is a largely semantic argument in an arena where semantics are meaningless, so I’m using it.)

A replacement level player (the type of Triple-A roster filler or washed up veteran available on waivers to all teams), which VORP uses for its baselines, is far below the average player. Thus, average players do receive positive VORP totals. Below average players, as long as they are above replacement level, also receive positive totals.

VORP does believe that average players have positive value.

Anonymous said...

Another bone to pick:

"The short answer, anonymous, is that no one really knows why Lohse has only rarely been able to get himself together at the major league level. We've all heard for years that he has "the best stuff on the staff" (although that particular bleat has died down considerably since the emergence of Santana and Liriano,) but that stuff has never quite coalesced into usable pitches."

The source of the highlighted section above came from a quote of the unquestonably credible source "an anonymous AL umpire". It was printed when both Santana and Nathan were on the roster.

I'm not quite sure if I know exactly what was meant by "stuff" (an expensive home theater system with a large flat-screen cinema display and surround sound speakers?), but if we're talking the ability to miss bats, this is a guy whose carrer K/9 is below 6.0. For a guy with "great stuff", he sure doesn't make many batters miss.

Anonymous said...

I agree completely about Lohse, Kyle, and what I was trying to point out was that the conventional wisdom about his "stuff" (and I've heard the "best stuff" comment in various form from Twins officials as well as that umpire) is directly at odds with every measurable assessment of pitching success. Sorry if it came out sounding like I meant something else.

As for the VORP stat, I understand that "replacement player" means a far below-average player. This is why I value the stat. But I distinctly recall reading a screed in which some high-profile expert (I think it was actually Bill James, but I'm not sure of that,) decried VORP because of the whole "zero" factor. It struck me as bizarre at the time, James being the god of stats that he is, but anyone who's read a lot of Bill James knows well his tendency to get exercised (and occasionally inaccurate) when his alpha-dog status is challenged. There are complaints about VORP from other camps as well, but they tend to be hard to assess, coming as they usually do from the apparently huge group of sabermetricians who despise the whole BP gang and everything they do...

John Sharkey, Esq. said...

SBG makes a very important point. The entire infield, really, has played poorly, and Stewart is really terrible in left. This is centainly a huge factor.