There’s No Such Thing as a Position Battle

Two men, one position. The spring training meme is at least as old as spring training itself, and yet is it a really real thing? If we believe in the merits of large sample sizes – and we do here – why would we forget that approach when the teams head south for spring? Perhaps this idea of the positional battle is overblown.

For the most part, the contenders for any starting role have left a trail of data behind them. This data isn’t always made of numbers, but it should be substantial by the time that player is eligible for a starting role on a major league team. His organization has scouted him at least once, and probably twice – for the draft and/or for a possible trade. On the way up through the minor leagues, his coaches have formed opinions, his scouting director has collated and informed those opinions, and his general manager has, at the least, read the file. He might have even sought out the coaches for additional conversation about the player. Add these subjective opinions to the data on hand, and you have yourself a dossier, sir.

With some possible leeway given for offseason development, why should that collection of data be outweighed by that player’s performance in 50 spring plate appearances? We want the front office to drag the net as wide as possible and not to weigh some handful of hits or muffed grounders equally against that players’ entire history. It goes against our nature to believe such a thing is the right approach, and yet we participate in the breathless coverage of the Spring Position Battle, if to a lesser extent. And, in defense of such coverage, there’s plenty of team officials that will discuss the pending position decision – this is not something that is made up.

But if you asked a team official in an honest moment, they might admit that most of the roster is set by the time the new year begins. And that they probably have strong opinions in place about their preferred or predicted winner of said position battle. That seems to render the position battle into some sort of Kabuki theater, played out for appearances’ sake. The media needs something to report, fantasy baseball prognosticators need something to debate, and serious fans need something to ponder as they drive to work.

There are times when it makes sense to open up a position. Perhaps a pitcher has radically changed his mix – by adding a cutter or focusing on a lost changeup. Sometimes they have completely revamped their idea at the plate, like Jose Bautista did last spring, and it becomes immediately obvious that, despite the small sample, something has changed. In New York, there is a new guy in town via the Rule 5 Draft (Brad Emaus) and a player trying a new position (Daniel Murphy). That seems like an okay time to open up the position and get some eyes on the battle.

In the end, we praise open-minded team officials for pursuing that extra edge and considering as much information as possible when it comes to most analysis. Perhaps, even when it concerns the traditional position battle, we should praise front offices for keeping open minds into spring training. Even when the subject at hand is a decision about a starting role that has probably been pre-ordained, and the extra data available are those same denigrated 50 plate appearances – even then it’s all new data, and it should all figure into the final decision. As Tom Tippet of the Boston Red Sox said at the MIT Sloan Sports Conference, “It’s all scouting.”

Print This Post

Graphs: Baseball, Roto, Beer, brats (OK, no graphs for that...yet), repeat. Follow him on Twitter @enosarris.

27 Responses to “There’s No Such Thing as a Position Battle”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. MikeS says:

    It kind of depends on whether you believe in players performing differently under pressure. I know it’s really difficult to prove but maybe some players will perform differently (better or worse) when they know there is a roster spot on the line as opposed to when they are just getting into shape knowing for sure they are either going to AAA or have already made the club. Everybody has an experience when the adult supervision is gone and YOU are the one making the decisions. It feels different. Even though It is hard to prove it for ballplayers, why should they be any different?

    But yeah, it’s a real small sample size. I never understand it when some player makes the big league club, starts off with a couple of bad weeks and gets sent down. What exactly did you learn in those two weeks that you didn’t know before and made you change your mind?

    Vote -1 Vote +1

    • Barkey Walker says:

      Lots of things could be happening there. A batter could get upset and change their stance / start swinging at garbage pitches, a pitcher could lose movement in their breaking ball. Pressure, and trying to “catch up” can do bad things for a player.

      Even top athletes can have problems and start hacking when their BA dips. The difference is their bad days are better than the other options good days, so you keep them around.

      Vote -1 Vote +1

  2. test says:

    Some position battles result from guys trying new positions. Plate appearances aren’t the only data available from spring training – defensive information is probably better. It doesn’t who is hitting the groundballs/liners/flyballs during games, and you can more realistically practice/show skill at defense outside of a game situation. In this sense, teams may actually have a lot more and better information about defense at the end of spring training than they do anything else. Hitters get uneven pitching and small samples, pitchers often “try new things” or are just stretching out their arms.

    Vote -1 Vote +1

  3. frugalscott says:

    The thrust of this post is somewhat legitimate in that teams often have a favorite going into any ‘position battle’, but even when there is no declared battle sometimes a player impresses and changes minds.

    Honestly, I believe in numbers (old, new and emerging) as much as anyone, but I am concerned when we get overly attached simply to numbers generated and how they can be manipulated. So often, we miss both the context in which those numbers are generated and the other important factors that make up what a player really is.

    It is true that much scouting is done on players and that players get only a limited amount of time on the field in the spring. It is also a time, however, for extra time in the cage, extra defensive work, learning to throw new pitches or refining the ones you currently have. Some guys will put in this work to the point of exhaustion and some will avoid it like the plague. Given mostly equal ability, the more coachable player will likely win out. If a player has made a mechanical adjustment (either in his hitting or pitching approach), his history is no longer as relevant so what he does with the new approach will have an impact.

    I’m sure you’ve watched a lot of spring games. I know I have here in Florida and it’s not just the small sample size. There are a ton of factors that make spring ‘numbers’ unreliable. The wind blows like crazy here in the spring and the closer to the beach, the stronger the wind. Defenses are not always the strongest. The fluctuating level of opposition influences numbers as well. These are just a few of the context factors that make spring numbers dicey.

    As such, it is important for teams trying to improve and having any question marks to see the big picture. In that picture, pure stats are a single tree in the forest.

    The point is that teams really do have position battles, especially in their rotations, and they use much more than just spring stats to make their final decisions. More often that you seem to think, they make decisions that come as at least some surprise.

    Vote -1 Vote +1

  4. Nate says:

    Biggest audience is the players.

    And there are plenty of times where the talent level between combatants is sufficiently close to make the actual determination as to the better one an unknown, especially for positions like 5th starter or last guy on the bullpen or bench, when neither of the contestants are great players.

    At lot of times saberists give the impression that a player’s value is known to infinite decimal places, when a lot of times the level of uncertainty can be nearly as large as the measurement itself, again especially for the marginal candidates.

    Vote -1 Vote +1

    • pft says:

      Some of the followers of “popularized” sabermetrics accept these numbers as gospel, and it becomes a matter of faith to them. Many sabermetricians, at least the “popular” sabermetricians, tend to avoid using words like uncertainty.

      UZR for example is reported to 1 decimal point despite it’s author claiming that UZR has little value due to the uncertainty. While he at least acknowledges the uncertainty, he makes no effort to estimate it or avoid reporting his numbers to 1 decimal place. WAR is of course based on 1 year of UZR, so is obviously uncertain.

      Park adjustments for offensive numbers, also used in WAR, are also uncertain. For example, Ichiro and Damon both had the same OPS (within 2 points in favor of Damon) in 2010. However, their park adjusted OPS+ was 113 for Ichiro, and 106 for Damon. Folks who trust these numbers absolutely would say Ichiro was a more valuable hitter last year.

      Comerika park is a bad park for LHB’ers who hit 380 ft HR’s, but is a RHB’ers paradise. SAFECO is death to power hitters, especially to LF, but for a guy like Ichiro who slap the ball around, has little impact. Yet because SAFECO is considered overall to be a tougher hitters park for an average hitter, Ichiro, who is not hurt by SAFECO gets a big boost, while Damon who played in a park that does hurt him does not.

      Vote -1 Vote +1

  5. Curious George says:

    “Sometimes they have completely revamped their idea at the plate, like Jose Bautista did last spring”

    Bautista revamped his spring at the beginning of September in the 2009 season, not during spring training 2010.

    Vote -1 Vote +1

    • Eno Sarris says:

      yes, good point. I think it might have become clear in ST that the changes he made had stuck, but good point.

      Vote -1 Vote +1

      • Larry Yocum says:

        Every season there is somebody that surprises us though. Another example would be Andres Torres moving from AAAA journeyman to power hitting, gold glove worthy, leadoff extraordinare for the WS champs. Last season he wasn’t even worthy of a “competition” against Aaron Rowand and was only able to get playing time early due to injury. Without him going all +6 WAR on us, the Giants don’t even come close to the postseason.

        This spring he is also not worthy of being in a competition with Rowand, but for different reasons. That seemed to be a clear example of a “position battle” where Torres just needed to show that last season was not a fluke. So far this spring he is hitting .333 and Rowand made his first start in left field yesterday. Competition over.

        The real battle in Giants camp though is for longman. They brought up Suppan, but so far Ryan Vogelsong has been the talk of camp as he has outpitched everyone else for that last roster spot and has shown increased velocity. Good teams need to keep an open mind if they want to find those journeyman diamonds like Torres, or a few years ago, a Nelson Cruz, that was placed on waivers and passed on by every team in the majors.

        Vote -1 Vote +1

      • merizobeach says:

        Larry, many of us who really noticed Torres in ’09 already believed he deserved the job, hands down over Rowand, before last season began. It was Sabean & Bochy’s initial inflexibility that allowed Rowand (and Molina and Wellemeyer) to open the season as starters, which also had the team on the verge of non-contention two months into the season.

        Vote -1 Vote +1

  6. JCA says:

    In some organizations, there can be a split in opinion over which player deserves to start. That apparently happened in Washington, where Mike Morse was the favorite of a number of the staff, but both the manager and the GM favored the more natural outfielder, Roger Bernardina. Botht he GM and the manager seem to have a preference for athletes, and Bernardina running and jumping ability, as well as occasional power, seemed to impress Riggs and Rizz more than Morse’s consistent OBP and average over his career and improving power consistent with his aging. Morse backed up the contingent that did not view his performance last year as a fluke and more or less took the job. I suppose this one is an example of a close call where neither player was an established starter (or even an established major leaguer), so it would fit into the small subset of legit spring positional battles, which can be swayed by a performance that confirms one view of a player rather than establishing a new view.

    Vote -1 Vote +1

    • Eno Sarris says:

      This split opinion idea is also relevant to the Mets battle it seems. It isn’t something I necessarily considered and is a good point.

      Vote -1 Vote +1

  7. GiantHusker says:

    Eno, you came as close as anyone possibly could to nailing this topic. Spring Training performance is close to 100% worthless, but it is psychologically impossible to ignore it–especially for we fans, who are starved for baseball after our winter hibernation and will gladly eat anything we can get.
    We Giant fans have, in desperation, gotten excited about the trivial differences among the left-field candidates and the trivial competition among very few candidates for the last 2 or 3 roster spots.

    Vote -1 Vote +1

  8. Richard Gadsden says:

    I think when a player is trying out a new position, you want the defensive data – which is going to be only a hundred innings at most, so will be mostly scouting rather than statistical – to make sure that he has reached an acceptable level.

    Bear in mind that even though we have positional translations for fielding stats, those are for players who have practised the new position and have learnt the particular skills relating to it, not for a player tossed into a position they haven’t played before with no opportunity to prepare.

    When a player is moving up the defensive spectrum (even when you expect they can do it, like Youkilis moving back to 3B for this season) you want to be extra careful.

    Vote -1 Vote +1

  9. gonfalon says:

    @ Larry Yocum:

    “so far Ryan Vogelsong has been the talk of camp as he has outpitched everyone else for that last roster spot and has shown increased velocity.”

    I’ll root for Vogelsong to have success as much as the next guy, but good luck if you’re expecting anything from him. after a good Spring Training in 2004 for the Pirates, his horrid performance when games counted (6-13, 6.50 ERA, 1.617 WHIP) truly earned him the nickname “White Flag”.

    Vote -1 Vote +1

  10. Chris says:

    Really can’t stand this article.

    I hate how fangraphs thinks that if something cannot be solved by statistics than it cannot be solved. And the one area where they really have not established themselves is projections. I absolutely believe that every manager should take into account every player’s previous 5 year stats, but if 2 of those years were at southwest missouri community college and in high A and double A, vs. a 35 year old whose last 5 years have extensive data. There’s just absolutely no way to effectively conclude based on the data alone. And I’m confused has to how just watching spring training and judging it could be hurtful.

    I think when it comes down to a lot of this, I guess I have a question. If there is not a large sample size to base a decision, isn’t a small sample size better than no sample? Sure the sample will definitely be wrong many times because of random fluctuations, but if they believe that samples are normally representative of the actual talent, then (in the absence of other data), it would be better to use the small sample size than not use anything.

    Plus defensive numbers? Come on….I might dare say that a scout is a more effective judge of defense over a season than any of the defensive mechanics.

    Vote -1 Vote +1

    • Eno Sarris says:

      I’m saying that the observations in spring are a small sample compared to all of those numbers and observations you cite. I didn’t say a single thing about defensive numbers in particular.

      And I said exactly what you asked is true. Yes, if a player is new to the organization or new to the position, then the small sample is better than no sample.

      Perhaps you gave too much weight to my sensationalist headline.

      Vote -1 Vote +1

  11. pft says:

    The only data worth a darn in ST is observational (scouting) for reasons.

    Most teams have a good idea when a top prospect is ready to be handed a MLB job, regardless of their ST stats. When they sign a FA with guaranteed money, they already have decided the player has a job.

    The battles for the available MLB players for a given position tend to involve players coming off injury or bad seasons, or marginal players. They tend to be signed to low dollars or minor league deals, and may have options. A bad season could be due to decline, poor talent, or bad luck. In the latter case, regression to the mean may be expected.

    Observation in ST helps determine if a player is truly healthy and assess his current tools/skills. Taken together with past performance one can figure out his potential upside or downside. Other considerations are age, contract, ability to be optioned, available depth, etc.

    Such judgements as to who gets the job to start the season are of course imperfect. Player who win the battle in ST may not necessarily hold the job until the end of the season.

    Vote -1 Vote +1

    • Eno Sarris says:

      I agree with most of this. I think they have to put ST performance in the context of the players’ career, and weigh it similarly. So a breakout has to be seen in the context of their recent performance, their data, the opinion of scouts…

      It just seems that ST work is the smallest part of this puzzle.

      Vote -1 Vote +1

  12. pft says:

    Meant “The only data worth a darn in ST is observational (scouting) for OBVIOUS reasons.”

    Vote -1 Vote +1

  13. Adamsternum says:

    props for the kabuki theatre reference

    Vote -1 Vote +1

    • Nate says:

      Heh, I assumed everybody was completely familiar with kabuki theatre since they go through it every time they fly on an airplane.

      Vote -1 Vote +1

  14. merizobeach says:

    “serious fans need something to ponder as they drive to work.”

    There are many layers of subtlety here. I like one of the deeper, more subtle premises: “the sheeple shall be bred for labor (as well as its taxation, and war)”.

    Vote -1 Vote +1

  15. Dave says:

    Aren’t most battles between “proven” vet vs. upcoming youngster? In the team’s eyes, a known commodity (good or bad, but reliable) vs. the unkown (again, good or bad, but with uncertainty)?

    Vote -1 Vote +1

  16. Greg W says:

    Interesting that nobody has mentioned injuries as a reason to roll guys out there for spring training time. Really, a good portion of what we call ‘battles’, bullpen and rotation especially, are determinations of the health of players right now, not overall ability. Matsusaka is throwing at 86-88? Hmmm, maybe we better find out who the 6th starter really is…. and if the 6th guy is experiencing shoulder issues already, who’s the 7th guy? I think a lot of what goes on at camp is about seeing if the offseason has been kind to veterans, or a great leap forward for rookies. Health, though is the overriding issue. Just ask Scott Downs, his new contract, and his borken toe.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *