Starlin Castro is in a funk. A deep one. A long one.

His overall numbers this year are the worst of his career – he’s now hitting just 272/.301/.414 for an OPS+ of just 92. Worse, his line is an ugly .243/.278/.393 going all the way back to May 11.

What’s going on with him? Should we be worried? Is this just a temporary blip? Well, it’s kind of both things.

Earlier this year, I noted a change in Castro’s approach at the plate (a more open stance), which seemed to be in service of pulling the ball more, and theoretically generating more power. It seems that the change has worked – Castro’s IsoP this year (.142) is the highest of his career – but has apparently left the rest of his game at the plate wanting. Is that the reason for his struggles?

Folks like to offer narratively pleasing, but ultimately unconvincing, explanations for his dip – the mid-season change in hitting coaches, for example – but the truth is that it’s probably as much an expected statistical variation as anything else.

You know where this is going: BABIP.

Over the course of their careers, hitters tend to put up the same batting average on their balls in play (BABIP), and outlier seasons in BABIP tend to regress back to the mean the next year. It’s become the vogue stat to point out hard luck years and too-much-luck years, but that’s largely because it’s predictive ability has been pretty strong.

So, let’s look at Castro’s BABIP story.

In 2010, his BABIP was .346, which is fairly high, but about what you’d expect for a player who hits so many line drives and who has decent speed. In 2011, then, as expected, his BABIP was .344. Two seasons isn’t quite enough for a baseline, but those figures square with his minor league BABIP, and give us a decent idea of the range you’d expect to see in 2012.

But his BABIP has actually been far, far lower – just .303. Your bad luck sirens should be blaring. Worse, his BABIP has dropped every month of the season, from .380 in April, to .333 in May, to .310 in June, to .238 in July, to .056 in August. That slide coincides perfectly with his increasing “slump.”

In other words, a huge part of the reason for Castro’s struggles is probably mere bad luck.


That all said, his dip in BABIP does nothing to assuage concerns that he remains too aggressive at the plate. To the contrary, an overly aggressive approach could lead to a drop in BABIP, particularly with someone who is as supremely good at making contact as is Castro – if you’re consistently making weak contact on pitches outside of the zone, you won’t hit as many hard liners, which are more likely to result in hits than anything else.

Just look at Castro’s O-swing percentage, which measures the rate at which he’s swinging at pitches outside of the strike zone. In 2010 and 2011, Castro swung at about 32/33% of pitches outside of the zone, and made contact with them about 72/73% of the time. This year, Castro is swinging at 40(!) % of pitches outside of the zone, and making contact with them at the same rate. That huge leap isn’t good.

Further, his K-rate this year (16%) is the highest it’s been in his career (previously he was around 13.5%), and his BB-rate (3.9%) is the lowest it’s been in his career (previously he was around 5.3%). Those issues have nothing at all to do with the drop in BABIP.

Clearly, Castro’s slump is a mix of bad luck, and a bad approach. He’ll probably hit better over the next month and a half just positive regression, but he’s got work to do. And it sounds like his manager, Dale Sveum, agrees.

“Until he’s willing to make some adjustments, it’s going to be a constant battle to be a consistent hitter,” Sveum said. “Everything’s in between – the leg kick, all the hand movement and everything. There has to be some adjustments to center baseballs on a consistent basis.”

Hopefully Castro gets things on track – and takes the next step forward – in time for 2013.

It’s important to keep in mind as we have these discussions, though: at 22, there’s plenty of time for him to fulfill his promise, even if it doesn’t arrive according to our preferred schedules.

  • Matt

    Is Dale implying that he currently isn’t willing to make necessary changes? Interesting….

    • Scotti

      No. He said it. He didn’t imply it. That said, the vast majority of MLB hitters are not “willing” to make “necessary changes.” Someone is always trying to tinker with your swing. The dude who changes every time someone suggests it is never going to succeed.

  • Engine 78

    Good write up Brett.

    His hitter profile over at BP validates the line drive/balls hit hard concerns too. In the 2nd half, he’s been pretty abysmal at driving the ball and making sold contact. Especially in the top third of the strikezone. He’s made contact and had it result in a line drive just ONCE since the Break in that part of the zone. On the flip side, he’s whiffed over 70% of the swings he’s taken up there. It’s mostly hard stuff too, not breaking. That’s more concerning than anything to me given just how good of a fastball hitter he’s been.

  • BetterNews

    I think Castro’s stance is part of the problem. However, I think Sveum is correct in pointing out all the hand movement. I don’t recall the Castro of past bouncing the bat around as much as he does today and he used to hold the bat much more vertical.

    • johnny kelroy

      Most Major League hitting coaches will tell you that stance means almost nothing. It’s really all about where your stance is when you are “loaded” just before you pull the trigger to swing. If you look at pics or videos of almost 90% or better of MLB players today they almost all look the same at this point. I think Brett’s article of free swining out of the zone and making weak contact is probably the most single problem he is having. I personally think he just has to become more patient…..even his takes do not look good right now.

      • BetterNews

        I would disagree. Stance is very important, I am not talking about about distance , or spread of the feet. I am talking about the front foot being way out of parallel with the back foot in relation to the plate. Very few successful players use this approach.

        • johnny kelroy

          I’ll agree to disagree. If stance is so important, then why do you see so much variation in the games best hitters? From Youkilis to Sheffield to Dawson, all different styles and stances, but look at all three when they are “loaded” (which is when their front foot has landed, or pushed) they are all similar at this point. I agree that maybe his problem in his stance is causing him to load up wrong, or is putting him in a bad loaded position, but I think much more of the problem Castro is having is because of his horrible pitch selection. In an at bat the other night, he swung at the first pitch out of the zone, took a fastball right down the middle for strike two, and then had to fight and foul off ugly pitches until he struck out. Much more of his issue is that he doesn’t even look confident on pitches he doesn’t swing at, when a hitter is confident, they look confident, and he looks to me that he has no idea what he is doing right now.

          • BetterNews

            I agree with everything you wrote. I did not wish to stress stance as being “very” important in my comment, but as important nonetheless. if I was to stress something about Castro’s swing it would be his bat position going into the swing. Watch Rizzo for example. He will rock his bat(for feeling and timing) but, he sqares up and his bat is almost vertical(few degrees from vertical) when he swings!

      • Drew7

        Your 1st three sentences (not saying the others are wrong) are on the money. While no twp have the exact same *batting stance*, I would be inclined to say even more than 90% of all hitters are in the same *hitting positon* once the front leg lands and the hips take over. There is no certain way you should stand or hold the bat before you’ve started getting into the hitting position, it is purely a matter of comfort.

        Coaches from the 1970’s would tell you different, but they’d also say things like “watch the ball hit the bat”, “take a level swing” and “get your arms extended”

  • J R

    I wish Starlin would just keep everything the exact same he did the first 2 years. Sure he wasn’t perfect and too aggressive, but he was still extremely valuable at the plate. Let the kid be himself and the power will contiue to progress. I think he’s being told way too many things..

    • BetterNews

      It could be he is getting conflicting advice. To me, it seems more a matter of over-confidence.(He believes he can hit anything because of his early success in his career) I’m not saying he has trown fundamentals out the window, but I do think he needs to knock the dust off and bring them out of the closet.

    • Hee Seop Chode

      I get where your thinking is J R, but isn’t making adjustments critical to MLB hitting? As recently as last year Jeter was making major changes to his swing:

  • clark addison

    The genesis of his slump coincides pretty well with Jaramillo’s firing. Coincidence? Maybe.

    • TWC

      Um, did you read the article?

      Folks like to offer narratively pleasing, but ultimately unconvincing, explanations for his dip – the mid-season change in hitting coaches, for example – but the truth is that it’s probably as much an expected statistical variation as anything else.

      • J R

        TWC, you seem like a hard core individual. You’re always going off on everyone. I am not hating on it, as I think it’s funny. I normally do agree with your thoughts as well.

        • TWC

          What?! ::feigned shock:: I’m a peace loving hippie. Just ask around.

          • Fishin Phil

            This is true. We start talking on the Message Board about WAR, and the first thing out of his keyboard is “Make Love, Not WAR” or “WAR, what is it good for?”

          • J R

            I can picture you now TWC; smoking some good vine, with you tie-die shirt, listening to the Grateful Dead waiting for anyone on the BleacherNation comments to cross the line, and going after them!

            • TWC

              I’m trying to come up with some sort of response to this, but I’m too busy looking around the room for the hidden cameras you placed.

              • FromFenwayPahk


            • Ogyu

              I always figured TWC was a robot operated by The Weather Channel.

        • Drew7

          He speaks truth – I saw him wearing Birkenstocks and a hoodie to work.

          TWC = Hippie

          • TWC

            Birkenstocks? Again with that nonsense? In the words of Perry Farrell:

            Oh, there’s that same asshole again. I thought it would never come to this. But the guy [wears] Birkenstock! I mean this guy is a real moron. He doesn’t even understand fashion!

            • Drew7

              I know, it’s just much funnier with the Birkenstocks included.

              • TWC

                I’m. Not. Laughing. ::glares at Drew::

                • bluekoolaidaholic

                  Ummmmmm, like WAR man, uhhhh like ummmm Dude, like wow man really far out WAR and BABIP mannnn.

  • ron

    I hate BABIPBS, it is used and described as “luck” when in reality it is a statistical measure of change. In this case it is a measure of a change in approaches that causes the perceived “luck”. I think luck is really a cop out and hate it.

    • Mike

      Not to be a jerk, but it sort of sounds like your problem is that you really don’t understand BABIP.

      • ron

        No offense taken but scientifically speaking it is BS. Statistics, like science only have value with a constant. If you change the equation or “stance/approac” the constant is gone and therefore makes BABIP worthless. Look at Ian Stewart, was his BABIP statistically power do to bad luck or a change in the equation in this case injury that limited his ability for consistent solid contact. When all things remain the same then BABIP has value. I see it more as statistical proof of inconsistency in approach that is all.

        • DocPeterWimsey

          As a statistical scientist, I can tell you that what you wrote is complete nonsense. Probability is about underdetermined situations: i.e., where myriad factors in addition to the one in which you are interested (here, Castro) are involved.

          • ron

            So you are saying that if he changed to a left-handed hitter and his BABIP dropped it would be due to luck?

            • DocPeterWimsey

              No, it would reflect the difference between his LHB and RHB skills. Switch hitters commonly perform differently (particularly for power numbers) from one side or the other. It’s one of the few breakdowns that ever actually holds up under scrutiny.

              • ron

                I was using an extreme example to prove a point. Like changing to a much heavier bat then just blaming it on luck.

                • DocPeterWimsey

                  You failed to prove a point. Altering a stance does not alter the batter’s basic tools. Ultimately, there is a simple test: if what you write is true, then performances should be drawn from a uniform distribution (i.e., random). They are not: they are drawn from binomial (or multinomial) distributions, with variations in those rates themselves following distributions (gamma, lognormal or Dirichlet) reflecting the variation in other factors (quality of opponent pitching, injury, weather conditions, ball parks, quality of opponent fielding, etc.)

                  • Ron

                    Altering a stance does not change the tools but does change how a the tools are applied. For example, Castro may have opened his stance to generate more power in his swing and has more homeruns. The tools are all there and are the same but he made a change that will effect things other than just power. (I am not pretending to know what the problem is by the way) What I am saying BABIP is a better measure of inconsistency and not merely “luck”.

                    • DocPeterWimsey

                      The biggest component of inconsistency is luck. What is happening is that a smaller proportion of Castro’s June and (especially) July GB and flares are getting through for singles than were his April and May GB and flares. Given his consistent power numbers, it’s not likely that he’s hitting more GB or flares: indeed, that should result in more singles, not fewer.

                    • Hee Seop Chode

                      wow, you guys are both pretty smart.

                  • BetterNews

                    “The biggest component of inconsitency is luck.” While I agree with 90% of the time, I am am having a really hard time swallowing that one Doc.

                    • DocPeterWimsey

                      Well, let me rephrase this with a question. Suppose a guy is a magical “constant” 0.300 hitter in that, regardless of the opposition or the conditions, he has a 30% chance of getting a hit in each and every AB.

                      Suppose that he gets 100 AB a month. How many months do we expect him to have a BA 0.25 or lower? How many months do we expect him to have a BA 0.320 or higher? Given 10 such players, how many of them will have a month batting 0.230 or lower?

            • Drew7

              The point all you “saber-haters” are missing here is that, if Castro’s problem was anything BUT luck, why would only his singles-rate plummet while every other rate remained largely unchanged?

              • Turn Two

                I am neither a saber-hater nor a strong supporter- i think its fun to look at but ultimately explains things after the fact much better than predicts.
                Castro can lay his bat on anything. Its always been his strength, the kid can flat hit. Therefore a change in his BABIP, correct me if I’m wrong here, could mean he simply is not hitting balls as hard as he does when he is hitting well, thus making them less likely to fall for hits. Whereas most guys would strike out in a slump like his, a talent like he has allows him, even when he is off, to make contact (albeit weak) Thus the lower BABIP- not luck- simply not hitting the ball very hard.
                Drew you say why is his singles rate the only thing thats down. I think he is still squaring up some balls well and pounding them when mistakes are made, its when mistakes aren’t made where he is struggling. The kid led the league in hits before he turned 20, those singles that allowed that to happen aren’t happening because he is taking the junk pitches that he used to impossibly find ways to smack for singles and they are turning into weak outs.
                Think of an aging QB who when he was younger used to rocket passes inbetween defenders and find his receiver. QB gets older, cant throw as hard, still thinks he can make those passes but its a hare slower and its picked off. Castro hasn’t gotten older but hes gotten out of his groove, he still thinks he can hit those pitches out of the zone but he cant, at least not with any thump on them for those line drive base hits of the past.

                • JMick

                  Best, and most logical, response yet. ^^^ Also, similar to what Brett was saying in the write up, if I’m not mistaken.

                • Drew7

                  You wrote quite a bit there, so I’d like to do my best to summarize first.

                  What you are saying is, this *slump* has caused him to not hit the ball hard, but only when the pitch is out of the zone? Also that, since his LD, GB, and FB rates have remained the same, he’s hitting the same proportion of those outcomes (roughly 2.5 GB, 1.5 FB, and 1 LD per 5 AB’s) when the pitch is out of the zone, but just not as hard?

                  • Turn Two

                    Not that cut and dry- but Castro has proven to be a solid “bad ball” hitter and if he is slumping that could be the logical area it declines. Information was presented stating his extra base hits were consistent with his history, so it is safe to assume he is still sizing up some pitches. Typically those pitches are going to be in the zone. Therefore, if we are looking for a BABIP change it would be logical that he is not hitting as many singles. It may be the bad pitches that he turned into the league lead in hits that has declined, thus lowering his BABIP.

              • BetterNews

                Drew 7-Brett himself said strikeout rate is up and walk rate down and he is swinging at a whopping 40% of balls outside the strikezone! Are we to believe it is simply bad luck?

                • Drew7

                  If you would take the time to read before you respond, I said “largely unchanged”; his K-rate did go up, but not outside of what is considered normal fluctuation. His BABIP-drop is much more drastic than the increase in K’s (which, by the way, has come with an increase in power).

              • Bric

                The problem I have with some of the sabermatricians is that (in my mind) either 1 + 2= 3 or it doesn’t. There’s no such thing as 1+2 equaling 4 on a “lucky” day.

                Doc can (and has) explained the luck factor and how 1+2 does in deed equal 4 sometimes but to me if your sabermetrics can be used to prove any number of different outcomes the entire nature of the methodology lacks fidelity and parsimony, and thus it’s usefulness and accuracy should be questioned.

                • Drew7

                  There are no absolutes in the game of baseball. There are, to quote Doc, “a myriad” of factors to consider. Thus, 1+2 doesn’t always equal 3. There isn’t any single stat that gives you an absolute answer; its mearly 1 piece of a big puzzle that, along with other measurements, help point you in the right direction.

                  I can’t tell people what to believe in, but I can tell you this: I havent always been keen to these sort of measurements, but have always been a “why behind the what” kind of person. If you understand why these statistics exist, how they are constructed, and what they are designed to tell you, you can make them very, “accurate and useful”.

                  • Bric

                    Well said. Unfortunately I don’t personally have access to the equations used to generate these statistics.

                    Moreover, many different websites use their own matrices in order to compile stats such as the hugely popular “WAR”. However, by their own admission, they all use different matrices in order to compile the number which results in a total a lack of mathematical fidelity. I think most statistical scientists should have an issue with that. I know I do. So I guess we just agree to somewhat disagree.

                    • DocPeterWimsey

                      There is a stats programming package called R that is freely available. It’s a little clunky compared to “real” programming in C, C++, Java, etc., (or Cobol if you are a Kobbold), but you can get exact probabilities of X events given Y trials and a rate of Z, and then quickly compare sets to see if you really need more than one Z for all of your examples.

                      (This applies just as well to baseball as to medical research, industrial accident rates, etc.)

                      You actually can calcualte a lot of these things in Excel, too: it will do binomial and multinomial probabilities. It’s a little clunkier, but almost everybody has it available.

                    • Drew7

                      “Moreover, many different websites use their own matrices in order to compile stats such as the hugely popular “WAR”. However, by their own admission, they all use different matrices in order to compile the number which results in a total a lack of mathematical fidelity”

                      I absolutely agree, which is why I’d never use WAR as standalone proof of someone’s ability.

                      However, there is a HUGE difference between WAR and BABIP, IsoD, OPS, wOBA, etc. To lump them all together and say they ALL lack fidelity is silly.

                • DocPeterWimsey

                  It’s not that 1+2=4 so much as that if the expectation is 3, then the probability of 3 events is high, but the probability of 2 or 4 events also is high. P[0] is not too low, either.

                  And then there are the other factors. That’s why statisticians do not just use Poisson distributions: they use Gamma distributions to model the expected shifts in Poisson rates due to other factors. So, we do not look at just rates (which reflect distributions of outcomes), but distributions of rates!

                  And that’s how we can deal with the fact that a batter has an easy schedule one month and a tough schedule the next, the fact that his wrists feel better/worse one month and worse/better the next, etc., etc.

                  • Turn Two

                    Everyone realizes you understand statistics, you have nothing to prove there- layman’s terms. What are you saying?

                    • DocPeterWimsey

                      It wasn’t so much to drop the lingo as to give people an idea of how statistical scientists approach these questions. Given that we expect a batter to get X hits per Y AB, we have a lot of different ways to describe how X/Y itself is going to vary because of other factors. In baseball, a 0.300 hitter has >30% chance of getting a hit off of a bad pitcher and <30% chance of getting a hit off of a good pitcher. So, the distribution of good and bad pitchers affects the distribution of X/Y even if the core skill of the batter remains constant. All of the other factors (fielders, ballparks, injuries) vary. So, if our expectation is 3 hits in 10 AB, it's not simply that the batter will often get 2 or 4 hits in 10 AB: we expect the expectation to drift up and down owing to the competition, playing conditions, etc.

                  • BetterNews

                    Doc-I’m gonna use a different analogy(getting away from baseball). I am at a shooting range shootng at a target with a rifle. I am a marksman and have shown I can hit my target within a specified area 98% of the time. I have sighted the rifle, taken wind speed and direction into consideration, very rested, etc.(in other words I am prepared. People would expect me to hit the target with the same consistency regardless of the day of year, the speed/direction of the wind, overcast or sunny, etc. Now there is a slim chance the wind might kick up and blow a grain of sand in my eye and I miss the target. Or I might hickup while in the process of shooting. My point here is luck certainly is not the major component in my success. The major component clearly would be my preparation and routine which I control, not dumb luck.

                    • Ogyu

                      I don’t think that’s a very good analogy. A batter is not swinging at a stationary target, but at a pitch thrown by another skilled player who is trying to get the batter out. Shoot your gun at a target that is moving around and trying to make you miss, and luck will have a lot more influence on the outcome.

                  • Bric

                    I’ve only used SPSS and Excel but remember Cobol from college. Anyway, again, that’s not the point.

                    Doc, we all know your expertice in this area. Turn Two’s right. The questions was what do the numbers suggest about Castro’s slump? in layman’s terms if you want to make a prediction or offer a correction. They may very well be reading.

                    I’ll make a totally unrelated prediction- Shark is going to get rocked in his next three starts and lose at least 2 of them.

                    • DocPeterWimsey

                      What the numbers suggest is that Castro cannot buy a single these days. However, he’s just as apt to get an extra-base hit, draw a walk or whiff as usual.

              • Flashfire

                So, I realize I should stay out of this, but I’m kind of stupid this way. My hope is that this will help, and not hurt.

                I’m an economist by trade, and economists are obsessed with probabilities and determining the outcome of somewhat random events — especially events associated with theoretically unpredictable human behavior.

                The key tool here is regression analysis. (Probably a probit regression, to be specific, in the case of at-bats.) What we’re trying to find there is the likelihood of a specific event, given a set of underlying circumstances. In this case, the event would be the probability that he gets a hit given that he put the ball in play.

                We observe that his actual BABIP has dropped from the last couple years. One side attributes that to luck. In the model, that means the “error term” — just random statistical noise, that has an average of zero — has, just by the luck of the draw, been negative each time he puts a ball in play. Thus, even though on average he gets a hit, say 36% of the time he puts the ball in play (BABIP of .360), this year due to dumb luck it’s actually 30% (BABIP of .303 — quick and dirty Internet search which may be wrong). This is not at all impossible given the model. What you could do — and I don’t have the time or inclination to test this — is find out how likely it is that this assumption is true. An economist would accept or reject the underlying assumption (that Starlin “should” have a BABIP of .360) based on the available evidence.

                The other option here is that the theoretical BABIP of .360 is wrong. What does that mean? That BABIP is going to be the result of several factors. Some of them are listed near here: LD%, GB%, walk rate, etc. Are those all of the factors? I genuinely don’t know — though I would assume there are others related to coaching and approach that impact it. I trust the coaching staff and front office to take actions to maximize BABIP in those factors. However, it remains possible — note I said possible — that something has changed which fundamentally shifts Starlin’s “theoretical” BABIP. If that’s the case, it isn’t bad luck causing this — it’s a change to one of the input factors. If that’s the case, it needs to be changed back as quickly as possible.

                Despite what everyone seems to assume about my feelings towards Starlin, I’m agnostic on which of these is the primary cause of his problems. That said, I am hoping the coaching staff does have a handle on it. The change in his place in the lineup suggests they think it’s more than pure bad luck. We’ll see what happens in the next couple seasons though — not being in on the minute-by-minute coaching and advice he’s receiving — it’s possible we’ll never know exactly what causing this slump.

                • Pat

                  Well said. The other thing to keep in mind is that it is probably not one factor causing the variance. It is probably several of the hundreds of factors that exist. Most likely though, those smaller factors do have the correlation that, for instance, line drive rate does.

                • Ron

                  That was great, I to should have stayed out of this as it is certainly outside my realm of knowledge/expertise. I guess I fall into the “it’s a change to one of the input factors” camp. I just have a hard time accepting a prolonged slump in this case as just bad luck.

                • Scotti

                  My thoughts on Castro? The smartest guys in “this” room are in the front office (and, by extension, the field staff). Are they saying that Castro is hitting into poor luck and that he’ll snap back with better luck or do they want him to make some changes?

                  My thoughts on BABIP? Like a great many things Sabre, often it gets applied incorrectly.

                  • Drew7

                    “My thoughts on BABIP? Like a great many things Sabre, often it gets applied incorrectly.”

                    I agree that tends to be the case, but is it so in this case? I hear that from so many people here, yet it is always left at that – with no example or explanation.

                    We won’t know the FO’s stance on this, so I suppose thats that.

                    • Scotti

                      We won’t know the FO’s stance on this, so I suppose thats that.

                      Not true. If the FO believed that this was just luck then Dale wouldn’t be saying that Castro needs to make changes. If it’s just “bad luck” (something even grizzled baseball folks believe in) then you want Castro to keep on doing what he’s doing. If he’s doing something wrong then you want him to change what he’s doing. They want him to change. Clearly the FO and staff do not see this as chance/luck/randomness.

                      “My thoughts on BABIP? Like a great many things Saber, often it gets applied incorrectly.”

                      I agree that tends to be the case, but is it so in this case? I hear that from so many people here, yet it is always left at that – with no example or explanation.

                      Well, typically you aren’t going to be getting a lot of detail from me if I’m posting using my Nook. I’ve switched over to the PC so I can go more into detail.

                      Yes, BABIP is being misaplied. In general the Saber Community has a long history of getting myopic on an idea and not letting go regardless of the facts (drafting college guys, “protection,” closers, etc.)–ironically (or hypocritically) it’s exactly like the “old school” guys being dug-in to their opinions that Saber guys historically complain about. BABIP, or rather chance/luck/randomness, is exactly one of those instances.

                      Certainly BABIP exists as a stat. But so does “doubles to GIDP ratio” or “HR to HBP ratio.” Anything can exist as a stat! The question is what purpose do any of those serve? What purpose are they supposed to serve? Even batting average has a theory (which is actually a fascinating story but I won’t get in to it here).

                      BABIP was originally created to show that pitchers have no/very little control over “outcomes” that don’t result in HR, Ks, BB or HBP. While that theory has since been debunked (that was a two-year study that drew conclusions that proved inaccurate once an 80-year study was done a couple of years later), the premise was that hitters and defense DID control those outcomes. Yet BABIP has somehow roughly morphed to batters not having control of outcomes outside of HR, BB and Ks. Again, in the original construct of the stat, batters were, in large part, supposed to control the outcome…

                      But EVERY time a batter slumps, someone somewhere BABIPs all over themselves. “OMG, Castro’s been playing with a broken foot!” “No, he’s just been hitting into bad luck.”

                      And even for those who do find good value in BABIP (chance/luck/randomness), the fact remains that, in Castro’s case, chance/luck/randomness are sure acting a lot more like a slump than chance/luck/randomness. Chance/luck/randomness comes back to the mean, no? Well, we’ve been talking about this slump for well over a month and no mean. There is nothing random about that. Over his last 142 AB he has only 28 hits (.197). That is well south of any other such stretch in his career and, again, not something that randomness “predicts” (again, randomness “predicted” weeks ago that he would be reverting to the norm–he isn’t). To be at his (prior) career average, he would have needed 15 hits to “randomly” drop that didn’t. I doubt that he has it 15 balls that would randomly drop for hits if half the defense were blindfolded. The dude is slumping.

                      Tangentially, many argue, “Well, players who hit .300 for their careers often have years where they hit .270 or .330. It all is random and it all evens out–proof of chance/luck/randomness.” Uh, players also have seasons where they have injuries and slumps, too. And a player can also get dialed in where it takes the league a while to figure out what he is doing. .270 and .330 simply do not prove randomness any more than they prove slumps.

                    • Drew7

                      First, I’d like to thank you for thoroughly explaining your stance on the subject. That doesnt necessarily mean I agree with all of it, but you’ve provided a really good explanation.

                      That said, the portion of your writing that includes “anything can be a stat” is spot on. In fact, I believe I wrote something very similar to what followed that line a little bit ago. You must understand what stat you’re using and why it is used, otherwise it has a great chance of being misused. I agree with you there.

                      My disagreement lies in the following sections, where you seem to dismiss the idea of this being a case of variance. Even if we consider his recent decline as a sufficient sample, its hard to argue that, given the fact that he is still driving the ball, taking walks, and K’ing at similar rates as before, a few GB, just havent gotten through. Even if a half-dozen did so, people wouldnt be talking about him “tanking” or being a “bust”.

                      I wont dismiss the idea that Castro may feel like he cant buy a basehit right now, and he may feel a bit mentally drained. I also stand by my opinion though, that he, the FO (who Im still not convi

                    • Drew7

                      …convinced are all that concerned), or fans to really be worried about.

                    • Scotti

                      Not a fan of the limits on being able to reply… Some day I’ll probably just take my lemonade stand down to the message board and stay there…

                      …given the fact that he is still driving the ball, taking walks, and K’ing at similar rates as before…

                      Correct me if I’m wrong but what we’ve seen posted here are similar rates year-to-year and not rates from this current stretch. If the rates from this 142 AB stretch showed the same rates across the board then I would be more tempted to agree that this is randomness.

                      Again, randomness means that it wont continue and yet it has. Randomness “forecasts” regression to the mean and “slump” forecasts that he continues to struggle until he has made the correct adjustments–it could go on indefinitely. Randomness simply doesn’t allow for that.

                      Even if a half-dozen did so, people wouldnt be talking about him “tanking” or being a “bust”.

                      This is the Internets here. A series of tubes. Lots of people with bizarre beliefs live in these tubes. Castro could go 20 for 20 and you’d still get people saying he was a bust because he didn’t walk (dude on a different site argued that Colvin’s AA record for consecutive hits was mere randomness because he wasn’t walking even though he was POUNDING the ball).

                      That said, a half dozen hits added to Castro’s last 142 AB would give him a .239 average during that time–that’s still a serious slump over 142 AB. That happens to players over the course of a year to be sure but not because of randomness. If all of those slumps were randomness then there wouldn’t be any actual slumps. Players do get into “funks” and funks are not random. Sure, players can be hitting the ball well and have nothing to show for it but not over 142 AB.

                  • BetterNews

                    I would diagree for sure. A drop from a .300+ avg. to.243 since May is more than mere bad luck. There are fundamental issues that need to be addressed, and they should be addressed with utmost expediency or management is not doing their job. It’s that simple.

                    • Drew7

                      Thanks for the insight, bud.

          • ron

            Is Stewart low BABIP just bad luck then also. I say no, it is lower because of injury. What I am saying is there has been a change and the change is mostlikely the cause not luck.

            • Pat

              Oftentimes change is used after the fact in a narrative sense to try to explain variance. It is convenient, because there is no way to measure approach and therefore no way to disprove such theories.

              I will agree that too often variance is referred to as luck, which can confuse the conversation sometimes if it is not clear they are being used synonymously.

              Look at Castro’s OPS+ numbers since he has been in the league. 100, 110, 92. There is nothing at all odd about the variance in those numbers. It shows a league average hitter with one up year and one down one. I think part of the problem is that many people believed that due to Castro’s age “he can only get better”. The truth is he could get better, remain the same, or get worse. Usually over a large sample it will be a combination of the three.

              • Mike

                Yep. That’s exactly what I was getting at below. Even if you take a “true” .300 hitter, he’s going to do better or worse than .300 sometimes, even with a full season’s worth of at-bats.

                And with less than a full season, the swings are even bigger. That’s the main reason I almost always hate “monthly” splits.

              • Drew7

                I mostly agree with this. I think the main problem is, while people are ok to assume a 22 yr old hitter will improve, they expect that improvement to be linear. Instead, they need to understand that, not only should a season like he’s having not shock anyone, you should expect it to happen (your explanation of *variance* above).

  • matt

    Another thing to keep in mind is that his line drive rates are pretty much the same.


    2010: 19.5%
    2011: 20.1%
    2012: 19.6%

    He’s still making solid contact, just not finding any holes.

    • DocPeterWimsey

      Is there a month by month breakdown of this? What would be the most informative data would be GB, FB & LD numbers to go with the K & BB numbers over the first 4 months of the season.

      If you look at outcomes, Castro has been constant in terms of extra base hits, walks and K’s. It’s Castro’s singles rate that has plummeted.

  • Steve

    I just hope its not because the New regime is trying to change Castro’s approach by making him take more pitchers or what have you hope its more him than others trying to change his approach to much, I believe in adjustments but you can change a natural talent.

    • J R

      I think Castro has been in between most of this year. He’s trying to rush the power (which may be from a move to the middle of the linup). He’s probably being told to wait for pitches in certain zones (which isn’t him at all). And he’s probably being told to take more walks (which absolutely isn’t him). Let the dude be himself..

  • Brian

    Age 22 might be the most important stat in the article.

    • ron


  • Derrek

    Well put Brett.

    There could also be a lack of motivation. Fore the past couple of seasons Bob Brenly has called the team a “dead-ass team”. Needless to say those teams had a lot of dead weight holding them down but I still feel that this team blows so many opportunities. Most of the players are so aggressive, especially with runners on base. It’s fine if you swing for power but you at least need to read pitches and adjust to the situation.

    Perhaps it would be unfair to use the term “dead-ass” to describe this team, but at the end of the day it certainly feels that way.

    • RY34

      agreed Derek, not to mention, this team just isn’t very smart in the baseball iq department; fundamentals have sorely lacked for years with the cubs, none of our players bunt worth a damn, all our pitchers walk way too many hitters, none of our hitters walk enough, our hitters don’t shorten their swing with two strikes, cannot inside out swing to advance runners, are horrible at going from first to third on base hits, lead the league in force outs when we should be turning double plays, i mean the list goes on. whatever we can fuck up we will and do fuck up. who was the last great leader of the cubs, i mean a guy that kept everyone on their toes and if they weren’t playing up to their capabilities dropped the hammer down?? i cannot remember a single guy of that caliber over the last 20 years.

  • Mike

    Sample size is always a problem too when looking at splits. To illustrate, I ran a pretty simple simulation:

    Give a guy a .300 “true” batting average. i.e., for every at bat, if a random number between 0 and 1000 is less than or equal to 300, count a hit. If it’s greater than 300, no hit. Give your made up batter 100 at bats a few hundred times, and I got a distribution of batting average over that time frame as follows:

    0.190 0.5%
    0.210 0.5%
    0.220 2.0%
    0.230 2.5%
    0.240 4.5%
    0.250 4.0%
    0.260 8.0%
    0.270 10.0%
    0.280 10.0%
    0.290 8.0%
    0.300 9.0%
    0.310 11.0%
    0.320 9.0%
    0.330 5.5%
    0.340 7.0%
    0.350 1.0%
    0.360 3.5%
    0.370 1.5%
    0.380 0.5%
    0.390 0.5%
    0.400 0.5%
    0.410 1.0%

    You can see that you’re still MOSTLY clustered around his “true” average, but there’s some wide swings in there.

    Now, expand that to 650 at bats (a full season) and this is the distribution you get:

    0.25 1%
    0.26 2%
    0.27 6%
    0.28 12%
    0.29 20%
    0.3 23%
    0.31 15%
    0.32 14%
    0.33 4%
    0.34 4%
    0.35 1%
    0.36 1%

    Still some variation, but MUCH more clustering around the “true” average. A lot of what we call “hot streaks” or slumps are just a function of random distribution.

    • Aaron


  • Andy

    For whatever it’s worth, his xBABIP is .325 this year compared to an actual BABIP of .303. So he really hasn’t been that unlucky. I think a big reason for that is his infield flyball rate has risen. That’s a huge BABIP killer. Plus his strikeout rate is at a career high 16.0% which is obviously going to hurt his average.

    But I think it’s equally important to point out the things that Castro is doing better this year. Like the post said, his power has never been better with an ISO of .142 and a HR/FB of 9.5%. People love to talk about his walk rate but he’s actually gotten better at that since June.

    BB% by month:
    April: 3.1%
    May: 1.7%
    June: 4.3%
    July: 5.8%
    August: 7.7% (obviously small sample)

    If it weren’t for that pesky batting average he could be playing the best baseball of his career right now. Just a reminder that the first slash can be as important as the last two.

    • dabynsky

      The walk rates are fascinating because there has been a huge jump in walks during his slump (under Jaramillo it was 2.7% and heading into last nights game it was over 6% under Rowson). If he could keep this approach and have his average return to normal levels we would have a very good offensive shortstop. I am hopeful because it really does seem like only the rate at which he is getting singles has dropped. I just hope that he can continue with his patience and driving the ball more once he gets things figured out to bring the average back up.

  • Drew7

    All of Castro’s numbers are in line with where they should be +/- sampling error (I’m sure, but Doc will almost certainly confirm or refute shortly): LD%, GB%, K-rate, BB-rate…all within that range.

    So that leaves us with BABIP: Why don’t we look to see what Castro’s numbers would look like if he were having the same luck as last year –

    If we add 16 singles to Castro’s season, we end up with a BABIP of .347 – actually a bit lower than his career rate. With his luck adjusted back to his career norm, I show a stat-line of: .309/.340/451 – or DAMN good for a 22 yr old, above-average defensive SS.

    Of all the concerns regarding this team, Castro is one that shouldn’t be included.

    • DocPeterWimsey

      As I noted above, what I’d love to see are the raw numbers: LDs, GB, FB, K & BB. Those are the “true” outcomes. A high proportion of LD go for hits, with a high proportion of those being extra base hits. An OK proportion of FB go for hits, but much more commonly of the XBH variety. About the same proportion of GB go for hits, but almost exclusively singles.

      Now, given that Castro’s XBH rate is the same, that suggests that his LD and FB rates have not drastically altered. However, it is possible that those rates have dropped and that he’s actually been lucky on picking up a greater proportion of XBH per LD or FB to compensate.

      Still, the incredibly low rate of singles per AB with contact really says: “bad luck.” A huge proportion of singles are grounders and flares. If Castro was hitting the ball poorly more frequently, then I would expect his singles rate to go up and his XBH rate to drop.

      • Drew7

        LD – 72 (19.6% – 19.8% career)
        GB – 177 (48.2% – 49.3 % career)
        FB – 118 (32.2% – 30.9% career)

        • Jack Weiland

          Two things I would note:

          1. The career averages here include this year.
          2. FB% includes infield flys, which are VERY different than Doc described outfield flies above. His infield fly rate is 7% higher than last year.

          • Drew7

            1. You are correct
            2. Yes, but his IFFB% seems pretty insignificant to me (it makes up 5% of his total FB over his career). He’s hit 26 of those in his career. I’m not sure what the “average” big league rate is, but surely it fluctuates wildly with everyone.

            Nervertheless,you’re right: my data was flawed

            • Drew7

              BUT….to be technical –

              “what I’d love to see are the raw numbers: LDs, GB, FB, K & BB.”

  • Cheryl

    Brett, Good writeup per usual. I think Matt raises an interesting question about Castro’s willingness to make adjustments. If you have ever done any writing and have to do revisions it can get frustrating. I wouldn’t be surprised if Castro is at that point in baseball now.He may be uncomfortable and frustrated in revising his approach to the point where he pushes back and tries to do his own thing.

  • Hee Seop Chode

    This is a well thought out, well researched, well written article. This is the stuff that makes BN my first destination on the web. Just imagine if the Trib had the confidence in its readers to hire Brett instead of, or in compliment to, the negative/always making himself laugh Sullivan.

    The success of your site is due to its superior content, Brett. Thanks.

    • willis

      I like your name.

      • Internet Random

        Whatchoo talkin’ ’bout?

    • Brett

      Appreciate the compliment, HSC.

  • Spencer

    How about the explanation that the longer Castro is in the league the more at-bats he has, and teams develop more detailed and advanced scouting reports on where the holes in his swing are. I suppose that would mean he has to alter his approach to re-adjust to the adjustments the league is making.

    • DocPeterWimsey

      Coaches position fielders based on where they hit the ball most frequently. As most hit balls are outs, that means that fielders are positioned based on where the outs are. In a sense, hits are always the exception even without any clue from scouting.

  • Myles

    It’s also very important to note that he’s probably still going to contribute more WAR to the team this year due to the MASSIVE improvement defensively.

    FWIW, I’ve got a stat degree from Purdue as well, but you don’t need one to understand that if you swing at as many pitches outside of the zone as he does, all of your plate appearances are going to get much more difficult.

    Here is the full list of qualified players that swing at a higher percentage of pitches outside the strikezone than Castro:

    Josh Hamilton
    Delmon Young
    Jeff Francoeur
    AJ Pierzynski
    Dayan Viciedo
    Adam Jones

    and THAT’S IT.

    Even Alfonso Soriano swings at fewer (full disclosure: he’s next in line behind Castro).

    As you can see, you can be productive if you swing at those pitches; for instance, Josh Hamilton is the leader in this category running away. The problem is, you can’t be that successful IF YOU MAKE CONTACT WITH THOSE PITCHES. You can swing at them as long as you miss (or foul them off), because (unless you’ve got 2 strikes) you’ve got an opportunity to do positive things later with a much greater probability then if you put that bad pitch in play. Josh Hamilton only makes contact with 53% of pitches out of the strike zone, which is way, way under the average (about 64%). Castro hits them 71.7% of the time. I can’t tell you how many of those are fouled off, but not all of them for sure.

    This also makes me sad: there are only 2 players that have a smaller difference between swings on pitches in the zone and swings on pitches outside of it. 2! He swings at way more pitches outside of the zone than most people but FEWER pitches inside the zone than the average person. I can’t even pretend to know Castro’s heat maps or whatever, but I have a hard time believing he’s taking advantage of this information.

    Best thing to take from this is that he can fix those things, though.

  • KyleNovak

    An excellent article!

    This is exactly the kind of the kind of thing you see at FanGraphs in the form of a “What wrong with Former All-Star X?”, or “Borderline Bench Player Y is hitting well, can he sustain his perfromance?” type article.

    Until Hit F/X and Field F/X data become released and unified, we can only look at what we have now. Has there been a change his swinging habits both in and out of the strike zone? Have his K-rate and BB-rate changed dramatically? What about a change in his approach (Pitches per P/A, First Pitch Swing %). The discussion here clearly touched on all of those pieces of data.

    HR/FB and LD% may be more uniform between sources like Fangraphs and B-R, but ground balls are not created equal at all. On each ball in play did the fielder make a routine play, not get to a clearly easy play, or make an incredibly different one? Until we can say with confidence that a ground ball hit 17 mph on the ground hit two feet to the right of the shortstop’s starting position, when compared to every other ground ball hit in that park analyzed down to the minutiae we won’t have completely accurate take on luck (or defensive skill, which ultimately means dWAR, and total WAR for that matter.)

    BABIP is a term that gets tossed around by tons of people who say “This is a high or low number and it WILL normalize to an average,” almost as if it was a predestined event. It’s much more complicated and is subject to a small-sample size (Reed Johnson’s high BABIP stayed high this year despite many people automatically assuming regression because he is a part-time platoon player with a smaller number of PAs and his physical skills didn’t drop off a cliff), injury (just ask Ian Stewart about that lagging wrist injury that he continued to play with for a long portion of the year), or factors we can’t yet quantify.

    Even Fangraphs put out an article saying, “Yeaaaah. . . so it turns out we were a bit . . umm. . wrong about our assumptions on pitcher and hitter BABIP and the reasons behind the certain fluctuations. Sorry about that. Especially since some of our writers tend to call people out as being “ignorant” if they don’t agree. Our bad.”

    We all learn new stuff, even smart people.

    Good discussion.

  • Frank

    While I’ve never been much of a believer in the power of one hitting coach over another, no matter how highly touted they were, it’s hard to ignore the fact that Castro’s initial struggles at the plate began a few weeks after Rudy Jaramillo’s departure. Jaramillo was fired June 12, at which time Castro had a line of .302/.314/.751. Between 6/12-6/16 he dropped all the way to .291/.306/.724. He then had another surge, and by 6/24 he was at .308/.321/.451, but the current slump then set in, causing him to plummet to where it is today.

    When it comes to veterans, especially those who have had success for years, even if they wanted to its hard for them to change their approach when a new hitting coach comes in. They’ve done things one way for years, and gotten results so it becomes habitual. However, Starlin Castro is not a high paid veteran with years of big league experience. He’s 22 years old and in his 3rd big league season, and Rudy Jaramillo, though many questioned his aggressive methods, was the only big league hitting coach he ever new.

    Should it come as a surprise that when a new guy came in, likely with his own methods and instruction that a young players like Castro is taking time to adjust? My other theory is that seeing as his main shortcoming is his OBP that knowing that the season is long since over, similarly to how they give guys certain things to work on in Spring Training or minor league rehab assignments resulting in ugly numbers, they’ve been asking to work more on his plate discipline and pitch recognition than hitting itself.

  • Kevin

    “Until he’s willing”

    Dale hit the nail on the head. Starlin’s head is the problem.

  • Sal T

    How about a few days off? I don’t think I’ve seen anyone else start or even come in late in games to play SS.

  • Toby

    Check out the big brain on Brett!

    Very nice post. Thanks.

    • Internet Random

      That is a tasty burger.

    • Brett

      Thanks, Tobes.

  • Brandon

    Plan and simple, about one month into the season Len Kasper did an interview with starlin and he said I look at my batting average everyday and I know I am suppose to be a .300 hitter. Every hitter I have ever coached that always looks at his batting average goes into crazy long hot streaks and even longer cold streaks. It was only a matter of time

    • Drew7

      I think that applies to every hitter…ever, regarless of what he’s looking at, regardless of the cause (be it luck, mechanical, or whatever). I think THAT is what we all need to realize.

  • someday…2015?

    Great story and numbers Brett. I really believe that if Castro closed his stance up and stopped trying to drive balls to left, he would see immediate improvements. I have also noticed Castro trying to take more pitches which is resulting in him guessing at the plate. For a kid with his quick hands and plate coverage he should not be up their guessing. Like you said he is 22 and has plenty of time to adjust and improve. I have no doubt Castro will find his bat again… Its only a matter of time for the youngest player on the team.

    • Jared

      i agree with everything you said, starlin has been trying to pull the hell outta everything. which is why is power is up and average is down. he will figure it out. he will figure out that there are times when he needs to turn on it and pull the ball and hit homeruns but there are certain pitches where he will have to go back to his old ways and go right back up the middle or go the other way with. and the coaches def got him workin on a lot of shit right now….i think thats why starlin did so well with rudy jaramillo….because he let starlin be himself at the plate. now i agree with both methods. starlin will soon find a median between super agressive and seeing alot of pitches and he will be just fine, hes just gonna have to go through the bad times to get to the good

  • Can’t think of a cool name

    Nice job by Sveum to throw Castro under the bus. Second time this year he’s talked negative about him to the press. Only two guys on this team with an OBP over .350 and no one hitting .300. Maybe Sveum is the problem.

  • FromFenwayPahk

    And this, friends, is why I read BN.

    Great post, Brett. I enjoyed it, although I remain more concerned with the progress of our shortstop-Castro than our top-of-the-order-hitter-Castro (mostly because I take solice in the anticipated regression to the BAPBIP-mean you made so clear).

    The Ron/Doc tete a tete that followed was also instructive. Ron, I was wondering the same bits about BAPBIP (e.g. why wouldn’t an new, open stance or new scouting by opposing pitchers move BAPBIP more than chance). Thanks for bringing them up.
    Doc, (and others) thanks for trying to get guys like me (more given to narrative) to understand the story as you read it.

    • Brett

      Thanks, FFP.

      • BetterNews

        Brett, what about me? I love your site, you know that, and I opologized in the most sincere fashion for comments that were misconstrued.

        • MichiganGoat

          You should just be happy that Brett even allows you to have this 4/5th chance to be in here after all you’ve done regardless of any apology you have given.

          • BetterNews

            You will never change, apparently.

            • MichiganGoat

              After all you put this site through (constant baiting, fruitless arguements, that stunt you played over at Boys of Spring, your racist “jap slap” comment, and your repeated antics) yeah I’m not going to trust or believe that your apology was heartfelt or serious. You bring down this site every time you are on. Goodnight.

      • BetterNews

        Brett- I am not a jerk, troll, imposter, or any one of names that commenters have made. I have followed the Cubs for forty+ years. I know Mg is your friend. He just needs to plain cut it out. All the jokes are just that. Just in jest. Nothing more. Don’t take away the fun.

  • baldtaxguy

    Very nice, thanks for the Castro food for thought.

  • Pingback: Starlin Castro Needs to Feed His Family and Other Bullets | Bleacher Nation | Chicago Cubs News, Rumors, and Commentary()

  • mike

    I think part of the problem is that everybody keeps saying he will eventually hit for more power. Thats probaly put it into his head to go out and try to hit for more power. That messed up Scot Thompson and messed up Matt Murton. I think its time to just STFU about what someone MIGHT do 5 years from now and just let the guy develop on his own.

    • DocPeterWimsey

      If this idea was true, then we should see Castro K’ing more than he has been. However, his K rate has benn constant this year. The only thing that has changed is the proportion of batted balls that go for singles. The simplest hypothesis predicting this is bad luck.

  • Pingback: Dale Sveum: Starlin Castro is a Six, Seven Hitter in a Good Lineup | Bleacher Nation | Chicago Cubs News, Rumors, and Commentary()