Callaspo, Trumbo, and the Third Base Profile
The always-excellent John Perrotto recently reviewed some of the Spring Training position battles in the American League. He helpfully includes comments from scouts on each situation. One quote that caught my eye was with respect to the Angels’ third base situation:
This is very interesting. Callapso is a pretty good hitter, but he doesn’t profile as a third baseman. Trumbo has holes in his game, but he does have pop and I think he’ll play a passable third base. It would be hard to take a kid that hit 29 homers last year, send him back to Triple-A, and try to sell that to the fans. But I know Mike Scioscia, and I know Callapso is his kind of player, so I’d be really surprised if he started Trumbo ahead of Callapso.
What struck me was the idea that Callaspo does not “profile” as a third baseman, apparently (judging from the context) because of his offense. The scout is probably thinking of power. Callaspo had a .086 ISO last season (.108 career), while Trumbo had a .223 ISO and 29 home runs while playing first base for the Angels and coming in second in the American League Rookie of the Year voting.
The scout thinks that Mike Scioscia will prefer Callaspo, probably largely based on fielding, as Trumbo has no professional experience playing third. But does Trumbo’s offensive “profile” really gives him any edge over Callaspo?
The first thing that leaps to mind is that Callaspo is likely to be far better in the field than Trumbo. Trumbo is considered to be fairly athletic for a first baseman, and the scout does say that he thinks Trumbo will be “passable.” Still, Trumbo has never played third in the professional baseball prior to this Spring Training (it is going smashingly so far). Callaspo, on the other hand, started out as a shortstop prospect for the Angels prior to stints in Arizona and Kansas City. While he moved to second base where he was, to say the least, not exactly Frank White, when he got a shot at third base in Kansas City in 2009 it was clear both visually and “metrically” that third suited him much better. One is probably on safe ground saying that while Trumbo might have a shot at being passable, Callaspo is a good bet to be substantially better.
But, again, that is not what is at issue here. After all, subjective notions of being passable or having better tools do not tell us how much better Callaspo would be in terms of runs saved in order to compensate for the purported difference in hitting. Fielding metrics claim to give us some idea, but even their most ardent defenders will admit that the error bars on fielding metrics are much larger than those for batting metrics.
So what is the problem with Callaspo’s offense “profile?” Well, the quote does not say specifically, but it is fair to infer that the issue is that Callaspo does not hit for much power. That does not mean that Callaspo has to be Mike Schmidt. But it seems the background of the traditional conception of third base offense is lurking in the background — something like a .250/.350/.450 line.
One way of responding to this would be to say that a player like Brett Gardner does not match the traditional “profile” for offense at left field, and he seems to do the job for the Yankees. However, that might not be persuasive on its own for the hypothetical traditionalist. After all, he or she might have doubts about the relative value of Gardner’s fielding for reasons mentioned above — perhaps that does not outweigh Gardner’s apparently “insufficient” offense.
This is not to dismiss the fielding issue, as it is obviously important. But what about the notion of an “offensive profile?” It would be understandable if one assumed that Trumbo’s 29 home runs made him more valuable at the plate than Callaspo last season. However, we do have ways of establishing the relative value of events using linear weights-based metrics like wOBA. However, the simple heart of my overly-lengthy response is this: Trumbo’s 2011 wOBA was .327 (.254/.291/.477), and Callaspo’s was .330 (.288/.366/.375). Trumbo was worth six batting runs above average, while Callaspo was worth about seven in fewer plate appearances.
Trumbo is younger, of course, so perhaps after making appropriate adjustments, he comes out better. ZiPS does see regression for Callaspo this season, projecting his 2012 wOBA at .314. However, ZiPS sees Trumbo 2012 wOBA as being…. .314. So there is no help there, either.
This is not meant as an attack on some mythically-dogmatic “traditional scouting” conception of positional offense or anything like that. Yes, it would have been quick enough to say that Callaspo is just as good a hitter. What makes this worth going over at greater length from my perspective is that the lure of power over everything else (and Callaspo does pretty much everything better than Trumbo other than hit for power) still tempts even trained observers.
The Angels do have a dilemma when it comes to Trumbo`s fate, as he is part of a three-player DH morass with Bobby Abreu and Kendry Morales. Some difficult decisions have to be made, but moving Trumbo to third base and benching Callaspo is not the right solution given the investment the team has made in winning now. Callaspo’s glove is the main reason he should play at third over Trumbo, but it is not that it “makes up” for an offensive gap between the two. Callaspo is not only better with the glove, he is just as good with the bat.
Mike Scioscia has understandably taken a lot of grief over past miscues such as the Mike Napoli-Jeff Mathis situation. However, if Callaspo is going to get the third base job because he is “Scioscia`s kind of player,“ then Scioscia should be commended for having his kind of player being simply better in this case.
It’s relatively unimportant that Callaspo was marginally better than Trumbo at the dish last season. The crucial piece of information in your argument is how Callaspo and Trumbo’s offensive profiles compare going forward. And just spitting out zips is not sufficient to ground your conclusion that the Angels would be wrong to bench Callaspo and move Trumbo to 3B.
Callaspo’s got 3.5 MLB seasons under his belt and turns 29 this year. He’s pretty much a known (and not all that good) quantity. Trumbo just turned 26 and has 1 year of professional experience. Sure, their median projections for next year might be equal, but the std. dev. on Trumbo’s is going to be a lot larger. His minor league profile alone shows that there’s hope for much more offensively. Plus, Angels scouts and coaches get to work with and watch him year round. If they see him taking a step forward with the bat and being “passable” at 3B, it’s likely he’ll be more valuable than Callaspo this year.
Bottom line: Callaspo is an average-ish 3B with no real upside whos physical peak is ending or over. Trumbo’s floor is a below average 3B, but the upside is there for him to become well above average. If Angels scouts see that upside materializing in the near future, I’d trust their opinions over a zips projection of a player with so few major league data points.
Don’t you undermine your own argument by acknowledging that Trumbo is 26? He’s not a young prospect who still has considerable physical and skills development left. He is relatively inexperienced at the major league level, but his extensive minor league profile doesn’t exactly give reason to believe that he’ll improve at all over his 2011 performance. I also don’t understand why you would suggest that Trumbo’s “upside is there for him to become well above average.” What is that based on?
If anything, I think Klaassen’s analysis rightly underlines the notion that there isn’t much difference in offensive value between Trumbo & Callaspo, it just manifests in different ways. Maybe you could argue that Trumbo’s power could be more useful to the Angels given the abundance of speedy OBP types they already have, but that seems like a stretch. Honestly I’m just surprised that LA hasn’t already traded Trumbo given that there are always teams looking for a cheap, power-hitting first baseman. That would make a lot more sense than trying to shoehorn him in at a position he’s never played.
Let’s not forget that Callaspo is subject to the David Eckstein rule – he does things that sabermetrics don’t necessarily catch (ya know, small ball stuff). At least, I think he does – but then, how can I be sure when there is no statistical evidence to support the notion! Excuse me while I go watch every Angels game from last year so I can properly weigh in.
If Trumbo were 22 or 23 and had his minor and major league resumes, this wouldn’t even be a question – clearly you give him the nod. At 26, he’s certainly got less upside, but he’s (1) in or about to enter his physical peak, and (2) showed better walk rates in the minors. All I’m saying is that there’s some physical and statistical basis to think he can take a step or two forward, and if that’s corroborated by the scouts and coaches who know him best (the Angels’ staff), I’m willing to give them the benefit of the doubt over a zips projection of a guy with 1 year of major league playing time.
As far as the upside of an above average 3B, a lot hinges on how well he can handle the position defensively. But if you give him a BABIP bump (he may turn out to be a low BABIP guy, but it’s probably too early to say) and improved pitch recognition/plate discipline, he could easily turn into a .350′s wOBA hitter. What I’m saying is that if his coaches and scouts see that potential, I’ll defer to them over a zips projection of a guy with so little meaningful input data.
So guesswork, of course a higher standard deviation means possibly higher highs and lower lows. All I’m saying is in cases of guys with so little MLB time, I’ll defer to coaches and scouts over zips.
Statistically (or probabilistically if you prefer), if the medians are the same and one has a greater standard deviation, then that guy has an equal chance to both be better than projected and worse, both by a greater margin. It’s very possible that Trumbo could get better, but it’s also possible he gets worse, and quite a bit worse at that.
If pitchers learn how to prevent Trumbo from hitting home runs, he’s basically useless. So sure, he could get a lot better, or he could get a lot worse. You know what you’re getting from Callaspo, which isn’t good but it comes with decent defense.
If you can minimize the penalty that will result from him underperforming though, then while his future performance is, as far as we’re concerned, random variation, the way it impacts the team is not. Since they’re not committing money to him, only playing time, he will be more valuable in the case he reaches his ceiling than he will be harmful if he falls to the floor. They can bat him low in the order and not start him every game, and if he excels, then they will have his entire future performance to look forward to, whereas if he fails he’ll just harmlessly sit in the minors or get traded.
You speak as if ZiPS is Marcel, using only previous major league data, but I’m 90% certain ZiPS incorporates minor league data into its projections.
At any rate, I know the Oliver system uses minor league data, as they pride themselves on their minor league translations. Oliver projects Callaspo with a .316 wOBA and Trumbo with a .315 wOBA.
I’m not saying your potential and inside information arguments have no merit, but the projections for Trumbo are based on more than one year of data.
Oh I’m sure you’re right about that. I didn’t mean to imply that that was the only data zips was working off of. But minor league numbers just don’t mean much. I mean zips is better off using them than not, but it would have a much better sense of Trumbo if it had 3 years of MLB data than 1 year of MLB data and 5 years of MiLB data. Which is why while it’s reasonable to rely on Callapso’s zips projection over scouting reports and projections, it’s less appropriate to do so for Trumbo.
I tend to believe Trumbo’s wOBA will be much closer to the Bill James .341 and the RotoChamp .339. I think you have to roll with Trumbo and the potential upside over the mediocre at best Callaspo.
What’s funny about that quote is no manager in baseball is better equipped to sell a concept to fans than Scioscia.
comment
Scioscia is better at accidentally sabermetrically evaluating third baseman than catchers, apparently.
I’d thought about this issue quite a bit this winter. Moving Trumbo to third never made agreat dael of sense to me especially after the Angels GM came out and said he wanted to get guys with better OBP’s in the lineup. Callaspo had a better OBP by something like seventy five points. Callaspo actually was one of the better third sackers in the game last year. The Angels are going to have some issues this year with being able to put their best lineup on the fiels day in and day out. Pujols is your first basemen, and Bourjos has to be in your out field. Trout should really be in the outfield as well but what to do with Wells and Hunter. Trumbo, Abreau, Morales, Callaspo, yikes!
First article on the new computer, or last on the old?
Trumbo looks like a replacement-level first baseman to me, who will struggle to be that at 3rd. That said, I can understand the idea of profiles a little bit. There are 6 positions in an AL lineup you’re going to typically have a chance to find plus power, and one of them, catcher, is really bad right now. If you put a slap hitter at 3rd, even an effective one, you’re changing your lineup from 5 power hitters, 2 light hitters, and whatever mess you get out of C and SS into 4/3/2. Maybe that’s better! But maybe it’s not, and it’s unlikely you can make up for the loss in power elsewhere.
But is a lineup with a balance of power and OBP better than a lineup that relies more heavily on one or the other? I suppose you want there to be OBP guys to get one to get driven in by the power hitters, but if you replace that power hitter with another OBP guy, then you’re more likely to extend the inning and get a rally going, even if you’re less likely to get a bunch of runs with one swing of the bat. There could be data on this I don’t know about, but it’s not intuitively obvious to me that a balanced lineup is better than an unbalanced one.
I tried to make it clear that I wasn’t arguing this was necessarily bad, but I do think it makes sense that as one removes power from the lineup, power becomes more valuable to the line up. I would be shocked if 8 OBP-only guys and 1 slugger was a better lineup than 7/2, for instance. It’d be interesting to try and find the ideal.
Thought experiment: a team with 1.000/1.000 OBP/SLG hitters (all singles) will be equivalent to a 1.000/4.000 (all home runs) team: if the rest of the lineup doesn’t make outs, eventually you’ll come around to score either way.
Both teams would be better than a hypothetical .001/.004 team that always hits a home run when it succeeds in getting on base. But that .001/.004 team would be better than a .001/.001 team that only hits singles if it gets on.
Thus, it would seem that the higher a team’s OBP, the less important that team’s SLG. SLG is the only way to score if you know you’re teammates won’t get on. It doesn’t matter at all if you know they won’t make outs.
So my hypothesis is that a SLG-centric player will be a more valuable addition to a team that has a low team OBP, but a slap-hitting high-OBP hitter will be more valuable to a team that already has a high OBP. However, within the normal, narrower range of MLB teams’ OBPs, I suspect that the difference between the two would have a rather small effect, if any, and so I’d just look for the best overall player.
I’ll agree, though, that this is a good area for further study.
Try playing around with this:
http://www.baseballmusings.com/cgi-bin/LineupAnalysis.py
In addition to optimizing the line-ups, it provides an estimate of runs scored per game. I think this is based on Markov chains, but I could be wrong.
It might help flexibility against different types of pitchers. An OBP-heavy lineup, for example, wouldn’t be well-suited to face somebody like Doug Fister or Dallas Braden, because pitchers like that won’t issue many walks, so they remove the advantage of Callaspo types. You’d want a Trumbo type in there to be able to punish a pitch here and there. Conversely, you’d want some OBP-heavy guys to face the Ubaldos of the world.
Probably evens out over the course of a season; might come into sharper relief in playoff matchups.
Newcomer, that was well put together and provides a compelling theory for why it might go the opposite way I expect. I do want to be clear that I’m not some “OBP is overrated” guy; I would expect a .250/.500/.250 guy to be the most valuable hitter in the game. That said, as long as players are more likely to make an out than not, I would think some number of players sacrificing OBP for SLG would be beneficial.
Yikes, that hypothetical player’s wOBA would only be .390, which would be 11th among qualifying hitters in 2011. I didn’t realize Bautista and Cabrera were nearing .450 in OBP!
This is really a no-brainer. If it aint broke don’t fix it. Callaspo was worth about 4 WAR last year if you average B-R and Fangraphs. My best guess? Trumbo ends up getting the lion share of the Plate appearances at DH. The Angels have the same stupid problems in the outfield as they do at the infield corners as they do at DH. The only reaon they are even considering this Trumbo thing is beacause if they play Callaspo, they have Trumbo, Morales, Abreau and Trout fighting for one position. Trading for Wells and allowing Bobby Abreau’s contract to vest were two of the dumbest things they could have done.
This falls into the blind squirrel department. If we have learned anything about Scioscia, it is that he is an irrational defender of his irrational preferences. Process matters as much as outcome.
No in professional sports outcomes easily matter more than process. That’s not an endorsement of the Angels manager and I’m not suggesting their is no correlation between process and outcome because their is. Just win baby!
The hope is that Trumbo will make gradual defensive improvement in March along with better OBA, allowing him to start the majority of games. They can always take him out after the 7th for defensive purposes or use a PR or PH depending on the situation. I don`t think they will be patient with him come April,he has to produce to play.
I prefer OPS to OBP or SLG. And Trumbo (.768) and Callaspo (.740) are about even in that too.