Archive for August, 2015

Analytics Are Good, But Psychometrics Can Make Them Great

This is not about a relief pitcher resting horizontally on a comfy couch as he spills his deepest darkest secrets to a furrowed, bearded psychologist, nor is this about prescribing medication to a team’s severely depressed kicker who just missed the game-winner. We’re talking about sports psychology, but not the kind of stereotypical psychology you’re used to. Instead, we’re talking about psychometrics – how to measure the ways that a player’s psyche (thoughts, feelings, opinions) relates to the most important thing imaginable for sport teams: performance.

Seeing is believing

Counting the yards that a running back gains after contact or the runs prevented by pitching independent of defense are advanced numerical methods of breaking down a player’s performance. Most of the traditional analytics work the same way; a player’s previous performance is charted, observed, and dissected to make a projection about how that player will perform in the future. A team’s forecasted performance is usually the sum of the individual players’ projected performances. This is (generally) the state of analytics in a nutshell.

Not only have analytics shown that previous performance predicts some level of future performance, it also just makes sense. Watching a player hit a 3-point shot, scoring pad-side against the goalie, and hitting a home run are visible to everyone; it’s what makes sports, sports. You know that Mike Trout is a good baseball player because you can see his performance. You can see him make ridiculous plays in the outfield and then watch him hit a home run into a fishing net in the center-field bleachers. You can check the box score the next day and you can see the numbers immediately reflect his awesomeness. You can visit FanGraphs and read about a sabermetric stat that further corroborates Trout’s awesomeness, and then you can use that same stat to find out about another obscure player’s performance and realize he’s kind of awesome as well. Analytics makes sense because most of it is overtly visible – above the surface, leaving everything else that can’t be seen as “intangible”.

What lies beneath

 Even if analysts were to measure more “intangible” characteristics, like a player’s leadership, grit, or mental toughness, they don’t seem to amount to the same numerical accessibility as traditional performance metrics, nor do they seem to be relatable to future performance. However, with carefully designed tools, psychometrics can not only measure these “intangible” characteristics, but can help predict future performance in the same way as traditional analytics. Ideally, psychometrics from players and teams can help complement performance analytics that are now readily being used.

In fact, measurement of the human mind and behavior isn’t anything new – over 100 years of psychological research has shown that the human psyche is quantifiable in the same way that previous performance is quantifiable. Psychologists have measured and quantified aggression across different cultures[1], charismatic leadership in managers[2], intrinsic motivation in children[3], and team cohesion within collegiate and recreational sports teams[4]. What’s more, these numbers can even fit nicely into the same models, projections, and predictions that have been used with traditional analytics. Yet despite the depth and breadth of this research, professional sports teams have been slow to tap into this area of study, pooh-poohed by pundits as “intangibles,” unseen and unrecognized by professional sport team brass.

You won’t know unless you try

If the results of these measurements help to win more games, what do teams have to lose? Teams should not fear the minuscule amount of time that their players would spend filling out a carefully designed survey if it means understanding more about them – and, ultimately, understanding more about their team. Teams should not fear the analysis of dugout, sideline, team bus, or hotel conversations between players, all of which include rich amounts of data that can help to explain the relationships between players. Teams should not fear the measurement of a player’s comments, quotes, tweets, or posts, their spoken or written words might reveal hidden emotions or intentions. The analytics movement is far from over, and if teams are looking for more numerical insights, look no further than psychometrics.

 

[1] Ramirez, J.M., Fujihara, T., & Van Goozen, S. (2001). Cultural and Gender Differences in Anger and Aggression: A comparison between Japanese, Dutch, and Spanish students. Journal of Social Psychology. 141, 119-121.

[2] Conger, J.A., Kanugo, R.N., & Menon, S.T. (2000). Charismatic leadership and follower effects. Journal of Organizational Behavior. 21, 747 – 767.

[3] Marinak, B.A. & Gambrell, L.B. (2008). Intrinsic motivation and rewards: What sustains young children’s engagement with text? Literacy Research and Instruction, 47(1), 9 – 26.

[4] Carron, A.V., Colman, M.M., Wheeler, J., & Stevens, D. (2002). Cohesion and performance in sport: A meta analysis. Journal of Sport and Exercise Pscyhology. 24, 168 – 188.

 


Does the Home Run Derby Affect Batted Ball Distribution?

Last week on RotoGraphs’ The Sleeper and the Bust podcast, Eno and Paul briefly discussed the possibility that Todd Frazier’s second half swoon in 2014 and again here in 2015 might have something to do with his participation in the Home Run Derby. While de-bunking the Derby Curse has been a popular topic of many data-driven pieces in recent years, research has largely focused on outcomes. For example, looking at changes in first and second half OPS and HR% for participants. Eno considered that the effects of the Derby might reveal themselves in other more subtle manifestations like batted ball data. Looks like he was onto something.

Most of the research on the subject that I’ve read takes a binary approach to participation – comparing splits of those who participated to those of players who didn’t. However, the Derby Curse’s narrative is that dozens of max-effort and mostly pull-side swings ruin a player’s 2nd half approach at the plate. So why would Bret Boone’s 2003 zero-homer first round exit lead to a 6% decrease in HR/FB rate in the 2nd half? After all, his *cough* economical Derby performance required he take only the minimum number of swings possible. Could it be plausible that changes in batted ball distribution are correlated with Derby performance rather than mere participation?

To find out, I exported the 1st and 2nd half Batted Ball data from the FanGraphs leaderboards for all Derby participants dating back to 2002, the earliest that batted ball data is available. I then added a column for home runs hit by each participant and regressed changes in batted ball rates for each BIP type against the number of home runs hit in each Derby performance.

In doing so I found 3 statistically significant relationships: ΔOppo%, ΔMed%, and ΔHard%, with the first two negatively correlated with HR hit and the latter positively correlated.

Coeff R2 p-value
ΔOppo% -0.10531 0.06042 0.01149
ΔMed% -0.11301 0.04041 0.03977
ΔHard% 0.10106 0.03567 0.05365

As one might expect running only simple regressions, the R2 values are low, intimating that other factors explain the majority of the variance. And I’m not sure that even a great performance at the Derby that requires those repeated max-effort and mostly pull-side swings has that significant of an effect on the RoS batted ball data. That said, a participant who hit 20 HR could expect on average to see a 2% decrease both in Oppo% and Med Hit % and a 2% increase in Hard Hit %.

It’s interesting that if anything, the data suggests that a better Derby performance correlates to an increased Hard% in the 2nd half, although it seems to come at the expense of Med% not Soft%. Nevertheless, an increase in hard-hit balls runs contrary to the notion that success at the Derby leads to a second half swoon.

ΔMed%

 Derby HR vs. Change in Med%

ΔHard%

 Derby HR vs. Change in Hard%

And while there’s no statistically significant increase in Pull%, it’s worth noting that the opposite hit type, Oppo%, decreases for those who do well at the Derby. Is that because players think more about pulling the ball and favor the inside pitch post-Derby or is there some temporary loss of skill in going the other way? Perhaps looking at how Pull/Center/Oppo distributions and heatmaps change in the weeks following the Derby might shed more light on that.

ΔOppo%

Derby HR vs. Change in Oppo %

So while we may not necessarily have proved or disproved the existence of a Derby Curse, we at least discovered that an exciting Derby performance is, if anything, more likely to precede an increase in the amount of hard contact a participant makes in the second half. Unfortunately for Bret Boone, this news may have come 12 years too late.


A Theory and A Challenge

I love this site. It covers the full spectrum of baseball, from classical scouting all the way to the most esoteric of baseball analysis. At times I envy the analytical abilities of our writers, as well as their access to granular data, that I likely lack the technical competence to gather. Today, I would like to propose a a theory, as well as a challenge to the numerous writers on this site to put the theory to the test. It is also likely that this has been proposed before and answered before, in which case, point me in that direction please.

THE THEORY:

We can measure command by compiling a pitcher’s xISO and xBABIP based solely on where they locate their pitches, in the context of the hitter’s preference to location. In other words, the ability to “pitch to the corners” is only valuable if one is pitching to corners that the hitter can’t get to, which is batter-specific. An 80-command pitcher will be able to maximize the xISO of his pitches, simply by pitching to “cold” areas of the hitter’s strike zone.

There are a few of ways to approach this (I’m sure more than three, but I digress). The first question is what sample size to use to estimate the player’s preference within the strike zone? Evidence suggest certain players make rapid adjustments (Trout) which would indicate a SSS would be ideal, whereas other players exhibit strong long-term tendencies (Dozier? just a guess, not founded in data) that would indicate a LSS would be ideal.

The second axis would be to evaluate a player’s effective strike zone, i.e. if we looked at the hitter’s swing probabilities, what type of strike zone would we construct, given only data concerning the hitter’s propensity to swing. We could then tease out whether the pitcher is maximizing the player’s effective strike zone (pitchers only throwing balls to Vladdy Guerrero comes to mind). This analysis may be redundant, as this can probably be captured if we are able to incorporate the third axis:

What are the thresholds for considering a pitch well-located? I.e. if a pitcher throws a ball way outside, but the hitter swings, then this is a well-placed pitch, thus at what probability of swing% is a ball a well-commanded pitch?

THE CHALLENGE

Test it! (or show me where this has already been fully fleshed out.) I’ve always wondered if there was a way to build up a command ERA to see if a pitcher is able to put it where hitters have to swing but don’t want to and I look forward to reading about it.


Examining Three True Outcome Percentage

Take a look at Chris Davis‘s stat line in August: 11 games, 45 PA, 14 Ks, 7 BBs, 6 HRs. Nothing really jumps out; it’s pretty typical for Chris Davis. Looking deeper though, this selection of plate appearances is actually quite remarkable. 27 out of the 45, or 60% of them, ended with a strikeout, walk, or home run, known as the “three true outcomes” where the ball does not end up in play.

As Baseball Prospectus explains in its definition of TTO, the statistic actually gained relevance with the introduction of DIPS, FIP, and other pitching estimators that ignored the outcomes of balls in play. While still not commonly used, it’s certainly interesting to take a look at once in a while to see what players are taking luck into their own hands.

Chris Davis is actually not the most extreme three true outcome player. Despite his 60 TTO% August, his season-long percentage through August 13 stands at 48.9%, good for 5th in baseball of those who have at least 300 plate appearances. The rest of the top-10 leaderboard features both good names and bad. On the good side, we have Giancarlo Stanton, the only player to feature a HR% over 8% (his is 8.5% , and he actually leads second-place Nelson Cruz by 1.4%). Other names you might associate with quality players are Bryce Harper, Joc Pederson, and George Springer, all of whom have a K% under 30% and a HR% of over 4%. The players who might not be as happy to be on this list include the aforementioned Chris Davis, Chris Carter, Steven Souza, Kris Bryant, and Colby Rasmus, who all feature a K% of 31% or higher. Mike Zunino, who comes in at 10th, sports a walk rate and home run rate of just 5.6% and 2.8%, respectively, but more than makes up for it with a 34.2% strikeout rate, second only to Souza.

Now that we’re done with the fun facts, let’s get into what it really means. TTO players are swing-for-the-fence players, those who aim to hit the ball over the wall every time they make contact. This is the cause behind their multitude of strikeouts. It also accounts for their walks, with the reasoning that pitchers are simply afraid to throw them hittable pitches.

The real question becomes “Are these TTO players valuable?” Looking at a graph comparing TTO% to wRC+ over the past 15 years, there is little correlation. It seems as though it is slightly more productive to be a TTO player, mainly because of the home runs and walks. This is far from a correlation though, as many bad players have a high TTO% and vice versa.

If we split it up into its parts, we might get a better view. League average TTO% has risen over the last decade, from 27.3% in 2005 to 30.3% this year (with a high of 30.5% in 2012).

We know the overall percentage has risen, but what’s driving it? If you’ve been following baseball, you know that the quality of pitchers has improved in recent years. Predictably, this has led to a decrease in walk rate and home run rate.

 

If 2/3 of the TTO% has decreased, but TTO% has still increased, that must mean the change in the third category must be drastic. This happens to be exactly the case. While BB% and HR% have fallen approximately a combined 1% over the past 10 years, league wide K% has risen by 4%.

What this means is that nowadays, if you are a TTO player, it’s likely much of that is coming from your strikeouts. In fact, out of the top-25 TTO% players with at least 200 PAs, only Paul Goldschmidt has a K% under 20%. Does this make high TTO% players bad? As I said before, there really isn’t a correlation, You’ll see players like Bryce Harper and Mike Trout with a high TTO%, while Buster Posey has one of the lowest because of his low K%.

The reality is, there are many different kinds of players. Some have adopted this TTO mentality, but others have stayed with a more conservative contact-focused approach. Without further information, it’s difficult to say which strategy is better. As a fan of statistics, I prefer the TTO players because it’s much easier to predict their performance. I don’t think they care much about that though.

Also, if you were curious, here’s a list of the top TTO% players with 200 PAs, created using FanGraphs data through August 13.


Don’t Sleep On These Post Hypers

NL West Edition

We’ve all been there and done that, our dynasty/keeper league(s) haven’t gone as planned. Perhaps you went for it in the offseason, ditched your prospects for grizzled productive vets and it all went south from there. No matter your story, the rebuild can be difficult in the sense of valuing the players you want. You could fall into the “shiny new toy trap” and end up with a bust or broken player (envision a Joc Pederson type in an AVG league instead of OBP). In this upcoming series, I will be highlighting players based on positions and pointing out whether I’d go for them in separate leagues (NL/AL only) or mixed.

So without further ado, here’s the first segment.

Read the rest of this entry »


BABIP Aging Curves

At age 35, Albert Pujols is having somewhat of a resurgent season. Many wrote him off last year after he posted his second straight, for him, subpar season. This year, though, he has hit 30 home runs through 108 games with ZiPS projecting him to get to 40 on the season. But there remain two big differences between 2015 and prime Pujols. One, he is walking less, at 7.5% vs. his career average of 11.8%. And two, his BABIP is a minuscule .228, continuing a declining trend:

Pujols BABIP

It certainly makes sense that with a loss of footspeed, BABIP would decline as well. After doing a quick mental recall, I decided to look up Mo Vaughn as another power hitter who seemingly lost it overnight. And sure enough, he experienced a big BABIP decline late in his career as well:

Vaughn BABIP

He still put up a .314 BABIP in his last full season, but it was a step change from the average .365 (!!!) BABIP he put up from 25-30.

So, is this a larger trend that we should be paying attention to? Or are Pujols and Vaughn just confirmation bias. Thanks to FanGraphs’ excellently downloadable data, I expanded the datatset to include every season and every player. Grouping by age reveals:

BABIP by Age

Well seemingly a lot of nothing. The BABIP for all 20 year olds in that time was .301, while the BABIP for all 39 year olds was .295. Definitely a decline, but with a p-value of 0.7 is not statistically significant. So that’s disappointing for my thesis, but encouraging for all the old folks out there! Back to the drawing board.

Pujols and Vaughn were big, hulking guys. Maybe when they lost a step, it was a step that they could less afford to lose and the impact on their BABIP of a marginal slowing down was magnified. So what if we restrict the group to only power hitters? For this, I defined power hitters as players with career ISOs over .200. The results appear to support my hypothesis better:

BABIP by Age, Power Hitters

This is plotted on the same scale as the previous chart so we can appreciate the relative differences. For this sample, the BABIP for power hitters declined from .313 at age 22 to .296 at age 36. Interestingly enough, power hitters had higher BABIPs earlier in their careers than the general population (including the power hitters), which then dip lower than the general population later in their careers. Apparently hitting the ball hard does have some benefits.

This time, the science backs up the hypothesis! My engineering professors would be so proud. With a p-value of 0.0165, the difference in BABIP between a 36 year old power hitter and a 22 year old power hitter is statistically significant. Pujols and Vaughn were indeed the victims of a real trend.

There could be a number of factors behind this. The first one I highlighted is the loss of footspeed. Second, it could just be that as you get older you don’t hit the ball as hard. Looking at exit velocity or ISO by age would help us judge that. Finally, age and a loss of bat speed or reflexes could lead to a change in batted ball in a way that leads to less balls falling for hits. It would make sense that as his bat speed slowed, Pujols tried to hit more fly balls to recover some of the home run power. That is the next thing I will look at.


An Overview of Prospect Production by Minor League Plate Appearances

Prospects are the lifeblood of any baseball organization. They have the ability to provide large amounts of value for their team while making a fraction of what they could earn on the open market. This provides a huge competitive advantage for teams that have a superior player development system. Every organization has a different plan for their prospects and the purpose of this research was to attempt to determine which development plan yields the most production in a team’s cost controlled years for each group of players.

The Data

The first step in gathering the data was to find every hitter that debuted from 1995-2009. I stopped at 2009, because this covers most of the prospect’s cost controlled years. I chose to start in 1995, because it gave me a big sample size and I got to avoid the strike year of 1994. Next, I omitted anyone who debuted at the age of 29 or older. I did this, because players that are over 28 are usually not considered prospects and their clubs would not consider them to be future building blocks for their organization.

The final step was to eliminate anyone who did not exceed their rookie limits. I decided to omit these players, because any player that cannot amass 130 at bats in their career was probably never considered a serious prospect. If they were, at least one team would have given them more opportunities to earn a starting job.

Methodology

To determine a player’s production during his cost controlled years, I found when every player exceeded their rookie status and added the next five years of WAR to their total. If the player had previous major league experience prior to the season they lost their rookie status, I included those numbers as well. For a player’s minor league plate appearances total, I included all of their plate appearances from the start of their professional career up to and including the year they lost their rookie status.

I then broke up the data by player groups. I split up the data by players who attended college, American born players that did not attend college and international born players that did not attend college. Throughout the rest of this article, I will simply refer to these groups as college players, high school players and international players.

Next, I partitioned the data by minor league plate appearances. I decided to split the plate appearances into groups of 500. I chose this amount of plate appearances, because it is a nice proxy for a full season of production and it splits the data into a fairly even distribution of players among the groups.

Overall Performance

I’ll start by giving a simple overview of total player production over their cost controlled years. The table below shows the median WAR for each grouping. I decided to use median instead of average throughout this article, because the WAR measurement is right skewed instead of normally distributed.

Median WAR for All Players

View post on imgur.com

College Observations

As you can see in the table above, college players need the least amount of plate appearances to produce a high level of WAR, but there is a sharp decline in production when a college player amasses over 2500 plate appearances. It makes sense that this player group is the quickest to develop, because they have had several more years of amateur competition to help hone their skills for professional baseball. This should create a smoother transition period for these players and reduce the amount of plate appearances needed to become a valued member of the major league club.

High School Observations

Unlike their college counterparts, American high school players take an extra 500 plate appearances before they reach their peak value of 15.4 WAR. However, high school players also have a wider range of success than either college or international players. High school players also produce more than the other two groups of players. This result may seem counter-intuitive, since it is commonly accepted that high school players are riskier prospects than college players. It is important to remember that this process does not account for all of the high school prospects that never receive an at bat in the majors. We therefore create a selection bias where we only look at the players that were good enough to make it to the majors in the first place. This means that if a high school player is good enough to make it to the majors; he’s probably going to be a productive major leaguer.

International Observations

The international player group offers the least amount of production. I believe there are several factors that contribute to this result. One of the main factors could be that many of these players have not played as much organized baseball as their counterparts. I also think that there could potentially be a language barrier issue that makes it more difficult for an organization to teach foreign players as opposed to their English speaking teammates. Of course that conclusion is just pure speculation on my part, but I believe that it is a reasonable assumption to make.

Total Player Summary

As the table above shows, the longer a prospect is in the minor leagues, the less chance they have of making an impact in the major leagues. This makes sense, because if a prospect is outperforming everyone in the minor leagues, they will be called up much sooner to help the major league club than everyone else. This leads me to believe that this table may not be the most informative for every minor leaguer. Perhaps, if we segment the data between Baseball America’s top 100 prospects and every other prospect, we will get a more accurate depiction of minor league development. It is essential to remember that the more we split the data, the less accurate our individual values may be. Therefore, we should not take the numerical value of WAR for each grouping too seriously. It is more important to take an overall view of the values in the tables below before drawing any conclusions about player development.

Median WAR for Top 100 Prospects

View post on imgur.com

Top 100 Prospects Summary

Yet again, we see that college players develop the quickest and that high school players take a little longer to develop. College players also have a quick drop in production after 1000 plate appearances, but they still yield the highest production of the three groups. International prospects are a bit of a mystery here. There does not seem to be a pattern in their production. I assume this is because there are major differences in baseball development between South American prospects, Japanese prospects and Canadian prospects, and any other nation’s prospects you can think of. In the future I may revisit this issue, but for now I’ll have to make do with what I have.

Median WAR for Non-Top 100 Prospects

View post on imgur.com

Non-Top 100 Prospects Summary

As expected, we see a dramatic drop in overall WAR across the board. This means that Baseball America is usually correct when identifying the most impactful future major league players. Kudos to you Baseball America. We also observe that these groups of players develop a bit more slowly than their more heralded prospects. These college players continue to peak early, but they are still 500 plate appearances in development behind the top prospects. High school players take even longer to develop now with a peak of 2.8 WAR in the 2001-2500 plate appearances group as opposed to 15.4 WAR in the 1001-1500 plate appearances group for the top high school prospects. International players are much more consistent in this table than the previous one. Unfortunately, they also have the worst total median WAR of 0.1.

Conclusions

So let’s do a quick recap. Usually the less time a player spends in the minors, the more productive they will be in the majors. High school prospects offer the most production, while international prospects offer the least production and college prospects fall somewhere in-between. We also observed that college prospects develop the quickest, high school prospects develop a little slower and international prospects are a bit of a mixed bag. I attributed this to simply combining all foreign born players into one group instead of by nation or continent.  I hope this article has been informative and that it provides some guidance on when teams should consider calling up their most prized assets.


Stephen Strasburg Is Better Than You Think

To a casual baseball fan, Stephen Strasburg‘s numbers are not pretty. The owner of a 4.76 ERA and a 1.38 WHIP, Strasburg is clearly having the worst season of his career. But how bad has he been, really? Not as bad as you think. Take a look at these 2015 stats:

Player A: 3.48 xFIP, 22.8 K%, 5.5 BB%
Player B: 3.31 xFIP, 24.1 K%, 5.3 BB%
Player C: 3.18 xFIP, 24.9 K%, 6.0 BB%

Player A is none other than Johny Cueto, recently traded to the Kansas City Royals. 12th in ERA among qualified pitchers, Cueto is widely considered among the best, and perhaps deservedly so with five straight years of a sub-3 ERA. While he has consistently outperformed the above metrics, they are still indicative of general pitcher performance and should not be overlooked when comparing the quality of different pitchers.

Player B actually has the fifth lowest ERA among qualified pitchers and was also traded at the deadline. He’s been one of the most reliable pitchers over the past five years and has been an ace on every staff for which he’s pitched. Player B is David Price.

Player C is obviously Stephen Strasburg, and as you can see, his peripheral stats stack up against the best in the game. In addition to these 2 players, Strasburg also compares positively to others like Sonny Gray and Scott Kazmir, both of whom have better ERAs but a worse xFIP, K%, and BB%.  Strasburg is pitching like an ace, and xFIP shows that, so why have his results been so poor?

Well, first of all, there’s his .345 BABIP. Not only is this high compared to the league average (.296), it’s well above his career mark of .302. Considering he’s not giving up any more line drives or hard contact than usual, his BABIP should fall back to around the .300 mark and bring his ERA down with it.

Not only is his BABIP at an all-time high, his LOB% is at an all-time low. Currently at 65.3%, it figures to inch back up to his career 73.2% mark, or at least to the league average of 72.4%. Considering his strikeouts have not dropped off, there’s no reason for his drop on LOB%, and it can simply be chalked up to bad luck, something that he’s had plenty of this year.

Looking at these stats, there’s nothing that suggests Strasburg is anything but unlucky. However, as Jeff Sullivan pointed out here, Strasburg’s problem could stem from the injury he suffered in the spring. He had apparently adjusted his mechanics to compensate for the discomfort, and even though it appears as though he has fixed this, it’s possible that when pitching from the stretch and in higher leverage situations, he returns to this altered motion by default. When looking at the difference in Strasburg’s stats between pitching from the windup and the stretch, this is what we see:

K% xFIP
Bases Empty 30.1 2.73
Runners on Base 17.0 3.98

Evidently, this claim has some ground. Strasburg is clearly having some problems with runners on base, particularly in striking batters out. Before we deal with the strikeout numbers, let’s take a look to make sure that he’s not just getting killed during the at bats that don’t end in strikeouts.

GB/FB Batted Ball Velocity (mph) Hard Hit % Infield Hit %
Bases Empty .98 89 29.7 4.5
Runners on Base 2.05 88 28.7 12.2

Strasburg is actually generating more ground balls and weaker contact with runners on base. His infield hit percentage is triple what it is when the bases are empty, something that can be attributed to luck. With such weak contact, it’s safe to say this isn’t the problem. So it must be the strikeouts. If we take a look at his whiff rates, the results are intriguing:

2010-2014 2015
Bases Empty 20.1% 17.5%
Runners On Base 17.9% 8.6%

OK, so there’s definitely a problem here. With runners on base, he’s only whiffing batters at half the rate he’s done previously in his career, as well as half the rate that he does with the bases empty. So what’s the issue? Well, it’s not his pitch velocity:

4 Seam 2 Seam Changeup Curve Slider
Bases Empty 95.1 mph 95.4 mph 88.4 mph 81.3 mph 86.7 mph
Runners on Base 95.2 mph 94.9 mph 88.0 mph 81.5 mph 87.2 mph

Strasburg’s average velocity with runners on base is 91.5 mph, compared to 91.0 mph with the bases empty, so he’s actually throwing the ball harder when there’s runners on base. That can’t be the problem. He’s also not walking a significant amount more batters when there are runners on base, so it’s not like he’s sacrificing control for increased speed.

Without any numbers to provide a reason, it appears Strasburg’s struggles when striking out batters with runners on base are either based purely in luck or are completely mental. This is not necessarily a good thing, as we have no idea if or when he will sort it out. With his skill, Strasburg has the potential to be one of the best in the game. He just needs to get out of his own head, and maybe get just a little bit luckier.


Falling Starlin

He could be playing (Saturday). I’m not sure yet. I want to see how it plays today, but I wanted to be upfront with him and just let him know it’s not just a day off.

— Joe Maddon

And with those words on Friday, August 7th, the Castro Regime fell in Chicago. Starlin Castro has earned the pine, posting an abysmal .268 on-base, around 50 points worse than the MLB average. Power has been even more of a problem; Castro’s ISO of .068 is sixth worst in MLB among qualifying hitters. It is also the worst of Castro’s professional career. Maybe he contributes with speed? Nope, not since 2012, when Castro stole 25 in 38 tries. He’s had only 23 ineffective attempts since then. His defense, long and loudly criticized, hasn’t been all bad; the metrics differ on him, but add them all up (metaphorically, anyway) and he seems to grade out about average.

Castro is striking out only a bit more this year than he has in his career (16.8% vs. 15.7%), but he’s making weaker contact. His infield fly percentage is at a career high of 12.9%, a full 5% higher than his career average. It was high last year, too, but he made up for it with a line-drive rate of over 22%. The line drives are gone this year, with Castro hitting a career low of 15.8%, which is, like his ISO, sixth worst among qualifiers.

Castro isn’t obviously being pitched differently this year. He’s seeing a few more strikes, but that’s probably an effect of his power outage rather than a cause. It doesn’t seem that pitchers have found some sort of secret recipe to deprive him of hits. Rather, it appears that fastballs are simply overwhelming him. According to his PITCHf/x data, Castro’s done pretty well against most offspeed pitches, but he has a league worst -2.70 runs above average/100 against four-seamers, and he’s 4th worst against two-seamers (-2.74). There are some decent hitters who have struggled with one of those pitches this year, but no one has been as bad as Castro at both.

Castro has been known to travel with a rough crowd, and more recently there’s been some ADD speculation. The Cubs organization is hinting that conditioning is a problem, which would explain the loss of power and his inability to hit the fastball. Perhaps, but Castro is 10th overall in total plate appearances since 2012. Whatever his problems may be, durability hasn’t been one of them.

And it’s worth remembering that Castro plays the most difficult position in what is arguably the most difficult team sport. He’s still only 25 years old, and by the standards of young shortstops, Castro has done quite well so far. He’s 29th in career bWAR (8.1) in the divisional era for shortstops through age 25. There are some great players in the top 50, and some not-great players, but there’s only one real disaster: Bobby Crosby at #40. (Ok, Rafael Ramirez at #39 was pretty bad too.) So Castro could have a Crosby-Ramirez future, in which he rapidly descends into mediocrity and irrelevance. But the vast majority of players with achievements similar to his at age 25 did not.

This suggests that either patience or a change of scenery could help Castro, as Grant Brisbee suggested in refuting the ADD speculation in the post linked above. Patience would not, however, seem to be the right move for the Cubs at the moment. Theo Epstein correctly eschewed the splashy megamove at the trade deadline: the wildcard game isn’t worth surrendering prospects. But it makes sense to to take less costly steps to improve this roster for the stretch run, and Castro is easily the biggest hole on the 25-man roster, with arguable exception of the 5th starter slot, now filled (for the moment, at least) by Dan Haren. The Cubs have been more than patient with Castro, and the performance hasn’t been there. Maybe they can give him more at-bats if they fall out of contention, but right now the team’s immediate future matters more than Castro’s.

That said, maybe the Cubs could spend a few minutes rethinking their approach to Castro. He’s has had three different managers in the last three years, each using a different approach with him. Dale Sveum’s tough love didn’t work, and Maddon’s zany zen isn’t working either. It was Rick Renteria’s more personal approach that seemed to get the most out of Castro. The karmic wheel spins in unpredictable ways, and Castro’s collapse may simply be the earthly price the Cubs are paying for Renteria’s defenestration, but it also suggests Castro can be reached, because someone was able to do it. Maddon is intelligent enough to realize this, and flexible enough to recognize that the shtick that works for most players doesn’t work for all. If Castro’s benching is coupled with some creative efforts to get him re-engaged, the Cubs may still be able to get value out of the player.

Diets, workouts, Ritalin, and perceptive coaching will be for naught if Castro is in fact the second coming of Rafael Ramirez. At some point his relatively reasonable contract will begin to look like an albatross, and the Cubs will cut him loose or trade him for minimal return. It would be helpful if players came equipped with little red crystals in their palms that glowed when the player reached his ceiling, but that won’t happen until at least the next renegotiation of the CBA.  So yes, it is possible that Castro has plateaued, and neither he nor the Cubs have figured that out yet.

But the Cubs have a little time. They can jury-rig their infield until they’re ready to press Javier Baez (or even Arismendy Alcantara) into service. They can see how the rest of the season develops, and how Castro progresses as they attempt to rebuild him in place, much like they’re doing with Wrigley Field.  As many have observed, his trade value can’t get much lower, so it doesn’t hurt the Cubs to take a little more time to see what they have. Burning a valuable roster spot on an unproductive player is dangerous, but the biggie-sized September roster is nigh.

If I had to bet, I’d bet that Castro will be moved in the offseason in exchange for someone else’s disappointment (Jed Gyorko, anyone?). But it’s not impossible that, in the top of the 12th inning of Game 7 of the 2015 World Series, Castro comes in from the end of the bench to hit the game-winning homer.  On what started as a day off.


Two Infielders You Should Be Talking About

I wish I knew why Jung-ho Kang and Ben Paulsen seem to get so little respect. It’s baffling. Regardless, people should be talking about these guys and their production — both have very legit numbers, yet few seem to have noticed. More to my point: fantasy baseball players should pick them up from the waiver wire ASAP. I mean, right this second.

Kang, recall, is the stud the Pirates signed from Korea. An unknown for the better part of the season, Kang is making his presence felt in the middle of the Pirates lineup, having just earned honors this July for NL Rookie of the Month. Kang, with dual SS/3B eligibility, is owned in just 57.9% of ESPN leagues and is slashing a highly productive .291/.365/.446 and, based on what he did in Korea, his .809 OPS could prove to be low in the long run.

Kang went through a bit of a power drought in June, but he caught fire in July. He’s now hitting .291 with 8 HR and 35 RBI. Consider that in the last week of July, Kang recorded multiple hits in five out of eight games with 6 R, 2 HR, and 3 RBI in that stretch. In his next game, on August 1, he hit his 8th home run of the season, a ball that traveled 412 feet. In 2014, Kang launched 40 home runs in 120 games in Korea, while also hitting .297. The kid can flat-out rake. With Jordy Mercer on the shelf (and not very good when healthy), Kang continues to occupy the 4–6 holes in Clint Hurdle’s lineup.

As many hitters have said before: As the summer heats up, so do they. I suspect we’re going to see Kang launch many more home runs before season’s end. If nothing else, even if the power is merely moderate, the fact that he hits for average, steals a few bases, and slots in the middle of a very potent Bucs lineup makes him worthy of a pickup in leagues of any size.

Ben Paulsen. What’s not to love about a guy who: 1) plays half his games at Coors Field; 2) made minor league pitching look like little league; 3) hits for both power and average; and 4) absolutely kills right-handed pitching? Answer: Nothing. His numbers aren’t dissimilar from those of Kang (in fact, they’re nearly identical), with a .300 average, 8 HR, and 34 RBI. His average is a bit buoyed by a .363 BABIP, though ZiPS projects a .333 BABIP the rest of the way. The only knocks against Paulsen are playing time and his ugly platoon splits, which are obviously related. But as with guys I’ve discussed before, who cares if he’s not an everyday starter; he’d just tank your average anyway. Instead, bench him against the few lefties he’s allowed to face, and you won’t be disappointed.

FanGraphs had this to say about him before the season started; it’s like these guys are clairvoyant or something. But they’re also very much wrong in the when they say that Paulsen’s game is made for just NL-only leagues. It’s much better than that (keep reading). Per FanGraphs:

The Quick Opinion: If Morneau starts the year on the disabled list as he recovers from knee surgery, Paulsen could be a sneaky short-term option in NL-only leagues, but that’s about it.

Paulsen, actually, is now effectively an everyday starter in the mercurial Walt Weiss’ lineup, thanks to the demotion of Wilin “Baby Bull” Rosario. Justin Morneau’s concussion symptoms are persisting, and he may have played his final game in the big leagues. Thus, the gig is Paulsen’s to lose, and with Corey Dickerson on the DL again, Paulsen has also been playing some corner outfield when called upon.

And when the 27-year old Paulsen is called upon, the numbers are a thing of beauty — against RHP, anyway, who he’s torturing to the tune of a .308/.361/.535 triple slash. Paulsen’s OPS of .896 isn’t just ‘productive,’ it’s downright fantastic. Frankly, it’s more than a little weird that just 19.7% of ESPN players own him. I’m happy to say I’m one of them, though I missed out on Kang, much to my dismay (and totally because of my stupidity).

There will be more blogs to follow, with similar themes in mind: finding value where there seemingly is none. There always is, you just have to look hard enough.