How Concened About Xander Bogaerts' Defense?

derekson

Member
SoSH Member
Jun 26, 2010
6,243
The people who are blithely saying "range doesn't improve" should take a look at the UZR numbers from Drew when he was in his early 20s and see how shitty his defense was. The very guy that you're comparing Bogaerts to had the same exact issue when he first came up to the majors. After a few years Drew figured it out and played good defense when he was healthy (he had a few more bad years around his ankle injury).
 
Drew's UZR/150 by year (age):
 
-12.6 in 480.1 innings (age 23)
-13.9 in 1283.1 innings (age 24)
-17.7 in 1294.1 innings (age 25)
3.0 in 1142 innings (age 26)
10.1 in 1259.1 innings (age 27)
 
As noted by Savin Hillbilly, Pedroia's history tells a similar story (though he didn't start out nearly as poorly). There is more to defense and even to range than the athleticism of youth. 
 

jscola85

Member
SoSH Member
Jul 14, 2005
1,305
Seems like Bogaerts' big problem with range is not footspeed but rather reaction time.  He seems to get bad jumps.  That is absolutely something that can be corrected.  I don't expect him to turn into Ozzie / Andrelton Simmons, but with time and experience, he will improve that reaction time.
 

Lose Remerswaal

Experiencing Furry Panic
Lifetime Member
SoSH Member
wolfe_boston said:
Yes, he's not going to improve his range and he's never going to be a GG SS, but maybe with some work he could improve to an average level which would be good enough.  The bottom line is that the Sox are likely to give WMB and X a full season before considering any changes.
 
You're judging his potential at SS based on 32 games played there during the regular season?

Do you know how you would have been judged based on 32 posts here?
 

nvalvo

Member
SoSH Member
Jul 16, 2005
21,510
Rogers Park
In light of this discussion, I was very interested to read over at OTM that Bogaerts is open to an extension. 
 
(Caveat: I like that blog fine, but since when do they do original reporting? Take this with grains of salt.) 
 
If true, this is against type (but not unprecedented) for a player represented by Scott Boras. But if we're cynical, it might suggest that Boras and Bogaerts are interested in locking in his salary while he's still seen as a SS. 
 

WenZink

New Member
Apr 23, 2010
1,078
nvalvo said:
In light of this discussion, I was very interested to read over at OTM that Bogaerts is open to an extension. 
 
(Caveat: I like that blog fine, but since when do they do original reporting? Take this with grains of salt.) 
 
If true, this is against type (but not unprecedented) for a player represented by Scott Boras. But if we're cynical, it might suggest that Boras and Bogaerts are interested in locking in his salary while he's still seen as a SS. 
 
I'm certain we'll see a lot this in Boston, even for players that don't have the pedigree of Bogaerts.  If Bradley or Middlebrooks show that they have the ability to stick they'll get offers (not as much as Bogaerts, but something to tempt them.)   And that will lead to some anxiety, much like in Chicago, last year, when Starlin Castro suddenly slumped after signing a long, pre-emptive offer.  It promises to add stability to aid future planning, but you'll end up eating a few deals to kids like Dan Bard.
 

joe dokes

Member
SoSH Member
Jul 18, 2005
30,249
jscola85 said:
Seems like Bogaerts' big problem with range is not footspeed but rather reaction time.  He seems to get bad jumps.  That is absolutely something that can be corrected.  I don't expect him to turn into Ozzie / Andrelton Simmons, but with time and experience, he will improve that reaction time.
 
Right. What ends up being lumped in with "reaction time" is anticipation, which is a function of experience. Know the hitters, (especially) know the pitchers on his own team, and his range will look better as he builds up information on what his pitchers do and what hitters do with certain pitches from certain pitchers.
 
The coaches position them, but eventually, Bogaerts will take that positioning and fine tune it a step either way based on his own processing of what's going on that day with this pitcher and this hitter and what the next pitch is (which they may not even know in the dugout).
 

WenZink

New Member
Apr 23, 2010
1,078
One positive thing about Xander's field presence, is that after an error, at least his jersey is neatly tucked in, so he doesn't look exactly like Edgar Renteria.
 

Mighty Joe Young

The North remembers
SoSH Member
Sep 14, 2002
8,402
Halifax, Nova Scotia , Canada
joe dokes said:
Right. What ends up being lumped in with "reaction time" is anticipation, which is a function of experience. Know the hitters, (especially) know the pitchers on his own team, and his range will look better as he builds up information on what his pitchers do and what hitters do with certain pitches from certain pitchers.
 
The coaches position them, but eventually, Bogaerts will take that positioning and fine tune it a step either way based on his own processing of what's going on that day with this pitcher and this hitter and what the next pitch is (which they may not even know in the dugout).
This could be describing the career of Cal Ripken at SS. He was a tall gangly and somewhat awkward guy at first. But he became one of the best fielding SS in baseball .. Primarily through the extraordinary anticipation he had. Bill James had a long essay on this in one of the early abstracts (don't ask me which one).
 

Al Zarilla

Member
SoSH Member
Dec 8, 2005
58,909
San Andreas Fault
WenZink said:
One positive thing about Xander's field presence, is that after an error, at least his jersey is neatly tucked in, so he doesn't look exactly like Edgar Renteria.
That might be because Edgar was kind of fat that year and that's what happens when you're fat.

You're right though, I remember that about Edgar.
 

Plympton91

bubble burster
SoSH Member
Oct 19, 2008
12,408
nvalvo said:
In light of this discussion, I was very interested to read over at OTM that Bogaerts is open to an extension. 
 
(Caveat: I like that blog fine, but since when do they do original reporting? Take this with grains of salt.) 
 
If true, this is against type (but not unprecedented) for a player represented by Scott Boras. But if we're cynical, it might suggest that Boras and Bogaerts are interested in locking in his salary while he's still seen as a SS. 
Another aspect is to get a long term deal done that still makes him a free agent at 27 or 28 years old, given the newfound reluctance of teams like the Red Sox to pay players in their 30s.
 

Saints Rest

Well-Known Member
Lifetime Member
SoSH Member
BCsMightyJoeYoung said:
This could be describing the career of Cal Ripken at SS. He was a tall gangly and somewhat awkward guy at first. But he became one of the best fielding SS in baseball .. Primarily through the extraordinary anticipation he had. Bill James had a long essay on this in one of the early abstracts (don't ask me which one).
I was just going to post this comp to Ripken, but I was going to cite George Will's book on baseball, Men at Work.
 

Doctor G

Member
SoSH Member
Jan 24, 2007
2,331
Bogaerts might have to be a shortstop who  plays the game on his feet. This is the opposite of the way Pedroia plays defense which is almost  an on his hands and knees style on balls he has to range for. Bogey has to be more of a Hardy Jeter style defender. This might limit his range but should cut down on the throwing errors he seems prone to when he hits the dirt and tries to bounce back up to make a throw. Use the plant and throw from the 3rd base hole and the spin and throw from behind the 2nd base bag.
his problems up till now seem to be a product of trying to do too much.
 

Reverend

for king and country
Lifetime Member
SoSH Member
Jan 20, 2007
64,053
This thread has turned pretty damned informative with respect to SS development and how different the reality is to some of the perceptions.
 
I appreciate the efforts of those who have made it so.
 

kieckeredinthehead

Member
SoSH Member
Jun 26, 2006
8,635
Butterfield talks about the importance of getting used to the different fields. It'll be interested to pay attention to whether Bogaerts looks better at Fenway over the next month or two than when they're on the road. 
 

ScubaSteveAvery

Master of the Senate
SoSH Member
Jul 29, 2007
8,329
Everywhere
Here is an update on Bogaert's defense courtesy of Joon Lee at WEEI.  The whole article is worth a read, but here are some snippets:
 
The sight of Bogaerts and Butterfield working in the infield has been common throughout the first month and a half of the season, as the shortstop tries to work on transitioning defensively to the big league level, a task that has been difficult for the 21-year-old. Following an April where balls frequently squirted past the ranging glove of Bogaerts, May has been a month of marked improvement for Bogaerts in the defensive transition to the big leagues according to Butterfield.
“[He's improved] a lot,” Butterfield said. “I don’t use a timeline, but since day one of spring training, you’re looking to get better during the course of spring training and during the course of the year and sometimes the slope is a little bit steeper than others but right now, I’m really pleased where he’s at. His feet are getting better every day and the key for any young shortstop is his feet, glove action is good. There is always little things that every player is trying to iron out, especially young players [at] demanding positions like shortstop. There’s things that he’s still working on but I like where his curve is going.”
 
 
The veteran presence of second baseman Dustin Pedroia provides a second eye on the field to position the infielders as best as possible pre-pitch. Pedroia’s assistance in the teaching process of Bogaerts has helped the shortstop get used to a lot of technical aspects on the field quicker.
“The sooner that he gets used to everything that is going on as far as the mental part, the over shifts, where he’s supposed to be and all of our team defense becomes second nature to him, then he can let his hair down and he can play like he’s used to playing and as a kid in the minor leagues,” Butterfield said. “Now, there is just a lot more added responsibility than he’s ever had in his life.”
Something that Butterfield says is important is to keep expectations in check for Bogaerts in the field, stating that the transition defensively for any young player is challenging. According to FanGraphs, Bogaerts currently sits 17 among 27 qualified shortstops in the majors in defense. In addition, Bogaerts has a plus/minus of -3 (three fewer plays than the average shortstop) and a -3 in runs saved according to John Dewan’s Fielding Bible. However, because defense is relatively difficult to quantify concretely with statistics, Butterfield tries to point out areas of importance and focus for each player he is working with.
 
 
One of the big takeaways is the mental aspect and risk of information overload.  Bogaerts mentioned that in the minors he just went out and played because there are so few scouting reports on players' hitting tendencies. But now he has to take in a ton of information about hitting tendencies and shifts combined with a much faster game defensively.  It is easy to see how this messes with one mentally and causes second guessing and rushing.  
 
On the topic of range, Butterfield mentioned that Bogaerts is getting better at pre-pitch anticipation, which helps his range because he gets better jumps.  This is one aspect where Pedroia is probably the best in the game at second base.  Pedroia isn't fast, but he gets such good reads that his range is huge.  If Bogaerts can take cues from Pedroia (which the article states that he is) then I expect his range to increase to at least average.  
 

ALiveH

Member
SoSH Member
Apr 23, 2010
1,104
sounds like he's already a league-average defensive SS and getting better quickly, so don't worry be happy!
 

iayork

Member
SoSH Member
Apr 6, 2006
639
In general, though, I think that players don't tend to improve defensively as they get older and gain experience.  At most positions -- especially shortstop and outfield -- players enter the league at their peak, and decline from there.
 
Note: This was a surprise to me, and I would not be at all surprised if I'm completely wrong.  Hopefully if I am someone will point out the stupid mistake I've made.
 
Here's what I get when I look at UZR/150 for players relative to their debut (edit: This is UZR/150 as a percent of each players' peak UZR/150). 
 
More explanation how I got this here and here.  
 
 

O Captain! My Captain!

Member
SoSH Member
Mar 3, 2009
3,532
iayork said:
In general, though, I think that players don't tend to improve defensively as they get older and gain experience.  At most positions -- especially shortstop and outfield -- players enter the league at their peak, and decline from there.
 
Note: This was a surprise to me, and I would not be at all surprised if I'm completely wrong.  Hopefully if I am someone will point out the stupid mistake I've made.
 
Here's what I get when I look at UZR/150 for players relative to their debut. 
 
 
More explanation how I got this here and here.  
 
you should look at fielding as a % of debut UZR/150 (with an innings limit to avoid cups of coffee etc), I think, though it will probably show mostly the same effect. i'm also not sure how you'd deal with survivorship bias, which is probably the worst at the most defensively difficult positions.
 
the magnitude of the effect is pretty small,  though. it looks like from peak to full decline shortstops lose 10% of their defensive value. you can see that's way less of a decline than something like the variation between defensive value between individuals
 

kieckeredinthehead

Member
SoSH Member
Jun 26, 2006
8,635
By including all data in a linear regression, you're basically anchoring the results to the very bad final years of a player's career. Probably best to do as a repeat measures ANOVA, looking in particular at years since entering the league. But an easier one would be to just look at pairwise comparisons between the debut year and a few of the years after.
 
Or just shift analysis entirely and look at the distribution of peak defensive year vs. year after debut. Eyeballing your data, it appears that there as many fielders who "peak" at year 11 than year 1, with mode more like years 2-5.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
O Captain! My Captain! said:
 
you should look at fielding as a % of debut UZR/150 (with an innings limit to avoid cups of coffee etc), I think, though it will probably show mostly the same effect. i'm also not sure how you'd deal with survivorship bias, which is probably the worst at the most defensively difficult positions.
 
Clarifying: That's what I did, expressed everything as a percent of each player's peak level.  Straight UZR/150 (i.e. not expressing it as percent of peak) looks like this (and the survivorship effect is very obvious):
 
 

iayork

Member
SoSH Member
Apr 6, 2006
639
kieckeredinthehead said:
By including all data in a linear regression, you're basically anchoring the results to the very bad final years of a player's career. Probably best to do as a repeat measures ANOVA, looking in particular at years since entering the league. But an easier one would be to just look at pairwise comparisons between the debut year and a few of the years after.
 
Or just shift analysis entirely and look at the distribution of peak defensive year vs. year after debut. Eyeballing your data, it appears that there as many fielders who "peak" at year 11 than year 1, with mode more like years 2-5.
 
I tried some higher-order trends, and they actually look as if they fit worse overall. Second-order looks like this:
 
 
Edit: I'll try removing a player's final year from the analysis and see if that changes the effect.  I don't think it will significantly, but it's a good idea.
 

kieckeredinthehead

Member
SoSH Member
Jun 26, 2006
8,635
It's not just about the final year, though. By including all of the years of data in your analysis, you're asking the question "What is the trend in defensive ability over the course of a player's career?" Not surprisingly, your best fit is (first order) linear: players get worse as they get older. Quantifying your data as percentage of peak also totally destroys the assumption of normality, so really any parametric regression (or ANOVA) is going to be problematic.
 
edit: You really do just want to do a repeat-measures ANOVA on the raw UZR/150 data, something like:
 
UZR/150 ~ year.in.league + age + randomeffect=playerid
 
If you want to throw your data up somewhere, happy to run it for you.  
 

threecy

Cosbologist
SoSH Member
Sep 1, 2006
1,587
Tamworth, NH
In regard to projecting a player's future fielding ability, I think some of the trending may be influenced by what tools brought the player to the majors.  There are certainly some players whose ticket to the show was their glove, with hopes their bat would catch up.  In this instance, it could make sense that their peak defense would be attained early in their career, especially in light of injuries, reduced playing time due to bat not catching up, etc.
 
On the other hand, it might be interesting to break out players who made it to the big leagues because of their bat.  Especially if they are intelligent and hard working, it's not surpsing if they improve as time passes.  One example might be Wade Boogs, who was arguably below average in the field during his early years, and didn't hit his stride defensively until his mid 30s.

 
 

iayork

Member
SoSH Member
Apr 6, 2006
639
threecy said:
In regard to projecting a player's future fielding ability, I think some of the trending may be influenced by what tools brought the player to the majors.  There are certainly some players whose ticket to the show was their glove, with hopes their bat would catch up.  
 
 
For what it's worth (and it's not quite the same question) there's not much correlation between UZR/150 and wOBA.  I had expected to see a reverse correlation, with good batters compensating for bad gloves and vice-versa, but it's practically zero.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
( ... deleted while I figure out what's going on
sonuvabitch, the csv files look like they got jittered on import ...
OK, not actually a big deal, just the scale got counted twice by SQLite; the actual patterns are identical.
Will correct in a little while)
 

kieckeredinthehead

Member
SoSH Member
Jun 26, 2006
8,635
Okay, so here's the repeat measures model. The basic idea here is that you calculate the mean UZR/150 for a particular player (the random effect), then you look at what their UZR/150 was in a given year after their debut in relation to the career mean (with year as a fixed effect). Take the mean of all of those across all players, and you get a set of estimates for the effect of year since debut on UZR/150. The horizontal line can be interpreted as the mean UZR/150 for the entire career. Obviously, no error bars on this, and variance increases as time increases because there are way fewer SS at year 10 than year 1. 
 

 
It looks to me like the first 7-8 years are a defensive player's best, with a steady decline after that. Year 3 is actually the peak defensive year, and the only one that's significantly better than 0/average. There are fewer data points for SS, but if anything there may actually be some improvement in the later years as they learn the position. I didn't see age in the dataset, which would be the next thing to control for - shortstops may come up earlier, so the average shortstop may be 28 in their 5th year, whereas other fielders are 32 in their 5th year (or whatever). In that case, it would make sense that shortstops still have the physical gifts to improve "later" in their career. And of course, as multiple folks mentioned, survivorship problems. 
 
edit: thanks again for sharing, that's a cool source of data
 

iayork

Member
SoSH Member
Apr 6, 2006
639
kieckeredinthehead said:
It looks to me like the first 7-8 years are a defensive player's best, with a steady decline after that. Year 3 is actually the peak defensive year, and the only one that's significantly better than 0/average. There are fewer data points for SS, but if anything there may actually be some improvement in the later years as they learn the position. I didn't see age in the dataset, which would be the next thing to control for - shortstops may come up earlier, so the average shortstop may be 28 in their 5th year, whereas other fielders are 32 in their 5th year (or whatever). In that case, it would make sense that shortstops still have the physical gifts to improve "later" in their career. And of course, as multiple folks mentioned, survivorship problems.  
 
That fits much better with my expectations.  Is that R? Can you share the script?
 
I updated the cdv file with the correctly-scaled UZR/150 (SQLite applied a /150 twice without me noticing, too much of a rush), but it doesn't change any patterns, just the numbers on the Y axis.
 

kieckeredinthehead

Member
SoSH Member
Jun 26, 2006
8,635
Here you go.
 
require(lme4)

# Load table
data <- read.csv("uzrNorm.csv", header=TRUE)

# Adjust UZR/150, reclassify year since debut
data$uzr150 <- data$UZR*150/(data$Inn/9)
data$years <- round(data$Post.debut)
data$years <- as.factor(data$years)

# Subset SS
ss <- subset(data, Pos=="SS")

# Mixed efects model
all <- lmer(uzr150 ~ years + (1|bbrefID), data=data)

# Messy hack to calculate relative mean UZR/150 vs. intercept
alldata <- all@fixef
alldata2 <- alldata[2:21]
alldata2 <- alldata2 + alldata[1]

plot(alldata2[1:20], ylab="UZR/150 above/below mean", xlab="Year Since Debut", main="All Fielders")
abline(h=0)

# Same, but for SS
ssaov <- lmer(uzr150 ~ years + (1|bbrefID), data=ss)
ssdata <- ssaov@fixef
ssdata2 <- ssdata[2:14]
ssdata2 <- ssdata2 + ssdata[1]

plot(ssdata2, ylab="UZR/150 above/below mean", xlab="Year Since Debut", main="Shortstops")
abline(h=0)
 

iayork

Member
SoSH Member
Apr 6, 2006
639
I know this is beating a dead horse, but whatever.  I realized that (1) instead of years post debut I should look at number of years playing a particular position, and (2) I should toss players who only played one or two years at a position (since they'll automatically peak in the first couple years and force a downward trend on everything else).  Doing that, most positions show relatively little effect from experience, except for outfielders, whose UZR/150 absolutely plummets like a rock after their second year or so.  This shows up in plain old peak vs. time plots, as well as in the repeat measures model kieckeredinthehead came up with.  
 
UZR/150 as a percent of a player's peak, per position, by years playing a position:
 
Repeat measures UZR/150 by years playing position, for shortstops:
 
For outfielders:
 
If you, like me, are wondering about the sudden jump at the end of the repeat measures outfielders plot: Since 2002, when UZR/150 started up, there have been two players who played outfield for 11 years; they are Ichiro Suzuki, who had a UZR/150 of 17.8 in 2013, and Carl Crawford (14.2 in 2013).  Both of those players were well above average, so that little dot at the end is accurate, albeit entirely small sample size and survivor effect.