Is JD Martinez a good defensive outfielder?

[The Gray Eagle]

Speier in the Glob takes a deep dive into the numbers with Jeff Stern from TruMedia Networks, an analytics firm that uses Statcast data and consults with major league teams.

"With Detroit, Martinez had graded as well below-average. After his trade, UZR and DRS offered favorable numbers for his work. Martinez recalled a conversation with Arizona head of analytics Mike Fitzgerald in which he discussed the disparity.

“My [UZR] has always been a negative in the outfield,” recounted Martinez. “Then I go to Arizona and my [UZR] is positive. I go, ‘What’s the difference?’ He said, ‘Now we’ve got you positioned right.’ ”

According to a major league source, Detroit ranked among the bottom few teams in the majors in outfield defensive positioning last year. Arizona was among the best. The difference was reflected in the defensive numbers Martinez delivered with the D-backs."

Another possible explanation for why Martinez could have rated so poorly last year with Detroit, but much better with Arizona later in the same season:
"The start of Martinez’s 2017 season was delayed by a sprained foot ligament that diminished his speed.
“I couldn’t run, but they needed my bat,” Martinez said. “When I was in Detroit, they were like, ‘We need your bat. We don’t care if you get to those balls.’”
Thus, Martinez played, and his numbers may have taken a hit."

Interesting note on sample sizes for outfielders:
"Most plays for an outfielder are either completely routine — made something close to 100 percent of the time — or completely impossible. Over the course of a full year, there are only about 50-100 plays that are neither routine nor impossible. As of the All-Star break, TruMedia had Martinez involved in just 14 outfield plays that didn’t involve those two extremes, with such a small sample creating a risk of distortion."


So what's the deal? Is JD a good outfielder, or what?
TruMedia says he is solid overall, better in RF than LF:

"That said, among those 14 plays, TruMedia — which graded his defense as below-average in 2016 and 2017 — had his overall body of work as being average to slightly above-average this season. While Martinez has failed to make a couple of relatively high-probability plays, he’s offset those with a couple of very good plays (the one in Toronto, another in Texas) in which he ranged back into the gap. Interestingly, TruMedia graded him as above-average in right — where he’s played for most of his career — while characterizing his work in left as slightly below average."

Overall, a really interesting article that brings to light some stuff about defense and publicly available stats that should be of interest to SOSH.

 

[Pozo the Clown]

Thanks for sharing, Gray Eagle.

Having watched almost every Sox game this year, using solely the eye test, I'd been underwhelmed by JD's play in the field (to say the least). Of course, the cards were stacked against JD when the eye test was also applied to his OF teammates on the Sox. That said, the article provides insights that the eye test alone would never reveal. A worthwhile read for sure.

 

[Savin Hillbilly]

Cool stuff. Some of us have questioned the JDM-is-a-butcher meme from early on. As Speier points out, the majority of outfielders are within spitting distance of average, and JDM appears to be one of those guys: on the outfield bell curve, he's maybe a C student--and while that's not impressive, it's a passing grade.

The one thing I missed in Speier's article--at least I didn't see it on a quick read-through--is an acknowledgment that while plays that are neither routine nor impossible for outfielders tend to be infrequent, they also tend to be momentous, with a significantly higher runs allowed/prevented impact than a typical 50/50 infield play. So the margin separating a C- from a B+ outfielder is small in sample, but large in consequence. It seems kind of like the difference between driving a car and piloting a passenger jet.

I'm also somewhat intrigued as to whether this article is an early warning sign of a JBJ or Benintendi trade for a pitcher.

 

[nvalvo]

14 plays before the All Star break, yes, but each OF play is worth more than a run: the linear weights value of a double plus the unmade out. So while 14 plays may not sound like a ton, if you think about it as a 20 run swing, you see how it makes a big difference.

The stuff about positioning is very interesting, in conjunction with some of the things we've been talking about in the JBJ thread. You can see how it would really make the comparative goal of stats like UZR challenging. Just as Bradley's OOZ plays suffer because the Sox choose to have him back up his talented corner outfielders on balls in the gap, JDM's terrible numbers in Detroit may have been exacerbated by Detroit's coaching staff.

That said, the new Statcast-derived Outs Above Average stat isn't wild about JDM, either: he's accrued -4 outs in a relatively small number of innings. Of course, as with all new defensive stats, I have questions about how it does for Fenway's LF, and those questions aren't quelled when I see that Benintendi is *also* at -4. (JBJ +7, Betts +5). I suspect Beni may have accrued some of that "value" in CF, where I've never thought he looked all that good, but it still raises questions given the eccentricities of the ballpark.

 

[The Needler]

14 plays before the All Star break, yes, but each OF play is worth more than a run: the linear weights value of a double plus the unmade out. So while 14 plays may not sound like a ton, if you think about it as a 20 run swing, you see how it makes a big difference.

Where does the value of a double come from? I would think the great majority of plays that “should have been made” fall in for a single that the OF doesn’t get to.

 

[Plympton91]

This has been a really interesting discussion.

I think it can be said that many plays in the infield are pretty routine too. There is the added degree of difficulty of throwing to first base.

That insight is why I’ve always thought fielding should get a significantly lower weight than offense. If About 1/3 of outs are K’s and another 1/3 are routine, then by definition fielding is less important than offense or pitching.

This is a long way of saying, if these numbers are right, then playing JD in the outfield and trading JBJ or Benintendi for an upgrade elsewhere is more viable.

 

[OCD SS]

While this might make the case that JDM is not Manny-level terrible (and maybe that Manny himself wasn’t as bad as the numbers said), I would be really surprised if this mindset was enough to put an OFer on the trade block:

First, his best position is RF, which is a lot different for the Sox playing in Fenway than most teams. If he becomes a starting OFer for most of his games, it would still likely be in LF for half of them.

With his opt out, the years of control are pretty limited for the Sox. That makes it tougher to trade Benni, and since Mookie isn’t going anywhere, that leaves what they could get for JBJ, which is probably not enough to warrant discussion.

 

[Reverend]

Speier in the Glob takes a deep dive into the numbers with Jeff Stern from TruMedia Networks, an analytics firm that uses Statcast data and consults with major league teams.

"With Detroit, Martinez had graded as well below-average. After his trade, UZR and DRS offered favorable numbers for his work. Martinez recalled a conversation with Arizona head of analytics Mike Fitzgerald in which he discussed the disparity.

“My [UZR] has always been a negative in the outfield,” recounted Martinez. “Then I go to Arizona and my [UZR] is positive. I go, ‘What’s the difference?’ He said, ‘Now we’ve got you positioned right.’ ”

According to a major league source, Detroit ranked among the bottom few teams in the majors in outfield defensive positioning last year. Arizona was among the best. The difference was reflected in the defensive numbers Martinez delivered with the D-backs."

Another possible explanation for why Martinez could have rated so poorly last year with Detroit, but much better with Arizona later in the same season:
"The start of Martinez’s 2017 season was delayed by a sprained foot ligament that diminished his speed.
“I couldn’t run, but they needed my bat,” Martinez said. “When I was in Detroit, they were like, ‘We need your bat. We don’t care if you get to those balls.’”
Thus, Martinez played, and his numbers may have taken a hit."

Interesting note on sample sizes for outfielders:
"Most plays for an outfielder are either completely routine — made something close to 100 percent of the time — or completely impossible. Over the course of a full year, there are only about 50-100 plays that are neither routine nor impossible. As of the All-Star break, TruMedia had Martinez involved in just 14 outfield plays that didn’t involve those two extremes, with such a small sample creating a risk of distortion."


So what's the deal? Is JD a good outfielder, or what?
TruMedia says he is solid overall, better in RF than LF:

"That said, among those 14 plays, TruMedia — which graded his defense as below-average in 2016 and 2017 — had his overall body of work as being average to slightly above-average this season. While Martinez has failed to make a couple of relatively high-probability plays, he’s offset those with a couple of very good plays (the one in Toronto, another in Texas) in which he ranged back into the gap. Interestingly, TruMedia graded him as above-average in right — where he’s played for most of his career — while characterizing his work in left as slightly below average."

Overall, a really interesting article that brings to light some stuff about defense and publicly available stats that should be of interest to SOSH.

Has @Buzzkill Pauley seen this? Because I missed it.

I would love his thoughts.

 

[teddywingman]

That insight is why I’ve always thought fielding should get a significantly lower weight than offense. If About 1/3 of outs are K’s and another 1/3 are routine, then by definition fielding is less important than offense or pitching.

This is a terrible simplification of... everything. I don't even know where to start. Maybe with you're joking, right? But since you probably aren't, I'll point out that your little assertion would make more sense if pitching and fielding were completely independent of one another.

 

[EricFeczko]

An alternative idea would be to weight fielding as an aggregate team measure, rather than an individual player measure.

Given how much positioning plays a role in fielding success, I wonder how much an aggregate team measure would be tied to the manager...

This has been a really interesting discussion.

I think it can be said that many plays in the infield are pretty routine too. There is the added degree of difficulty of throwing to first base.

That insight is why I’ve always thought fielding should get a significantly lower weight than offense. If About 1/3 of outs are K’s and another 1/3 are routine, then by definition fielding is less important than offense or pitching.

This is a long way of saying, if these numbers are right, then playing JD in the outfield and trading JBJ or Benintendi for an upgrade elsewhere is more viable.

Benintendi is currently producing at 145 wRC+, good for 15th best in the league. He's 24 and not even in arbitration - -he's under team control until 2023.

We're not trading Beni.

I'm not even sure what a package would look like for benintendi. Maybe bellinger/Buehler from the dodgers? Then Bellinger slides into 1B when mitchy leaves, and buehler replaces one of our starters? Perhaps Baez/Contreras from the Cubs? Beni would help them drop heyward -- but I'd suspect Theo would want to get rid of him in the process...

 

[nvalvo]

We're not trading Beni.

You are correct. Fangraphs has him at #14 in trade value, between Ozzy Albies and Corey Seager.

 

[RIrooter09]

He’s been the 4th most valuable OF in MLB according to fWAR. We are not trading him.

 

[Reverend]

An alternative idea would be to weight fielding as an aggregate team measure, rather than an individual player measure.

Why are we weighting anything here?

Assess run prevention. Assess run creation.

RC-RP=Outcome. Or vice versa. How did "weights" even get involved in the conversation?

That said, I agree that it is obvious that fielding/run prevention should be considered as a team metric-- @Buzzkill Pauley has been killing it in this area, imo.

 

[Plympton91]

Why are we weighting anything here?

Assess run prevention. Assess run creation.

RC-RP=Outcome. Or vice versa. How did "weights" even get involved in the conversation?

That said, I agree that it is obvious that fielding/run prevention should be considered as a team metric-- @Buzzkill Pauley has been killing it in this area, imo.

I agree with your last sentence absolutely and Fezko’s prior similar statement Which is why even the best of the pre-statcast individual numbers for defense are about as useful as plus/minus statistics in hockey. And thus you cannot “assess run prevention” by an individual player.

What we do know is that the center fielder plays no role in a strikeout. He also plays no role in a popup to 2B. This is why CF only have range factors of about 2.5. That’s how many outs they contribute. And we know that even a starting college CF will catch 99 percent of routine flyballs to CF, which inflate Range factors to even the 2.5. So how can a CFer’s fielding skills be anywhere near as important as their offensive skills, which are on display 4 times a game and aren’t replicable by an average college CF.

This shouldn’t be controversial, it’s not opinion, it’s mathematics.

 

[Reverend]

I agree with your last sentence absolutely and Fezko’s prior similar statement Which is why even the best of the pre-statcast individual numbers for defense are about as useful as plus/minus statistics in hockey. And thus you cannot “assess run prevention” by an individual player.

What we do know is that the center fielder plays no role in a strikeout. He also plays no role in a popup to 2B. This is why CF only have range factors of about 2.5. That’s how many outs they contribute. And we know that even a starting college CF will catch 99 percent of routine flyballs to CF, which inflate Range factors to even the 2.5. So how can a CFer’s fielding skills be anywhere near as important as their offensive skills, which are on display 4 times a game and aren’t replicable by an average college CF.

This shouldn’t be controversial, it’s not opinion, it’s mathematics.

A few years ago, I would have agreed with you.

But MLB's StatCast has shown that JBJ, at times, has a negative reaction time.

I wouldn't have thought that was a thing either. Like, in ninja books maybe, but not in real life. But yes, JBJ often reacts before contact and this is observable. Not only will this be in a batter's head depending on what he's trying to do with the ball, he's also, like, right there in center behind the pitcher doing whatever the hell his mutant head causes him to do.

 

[Devizier]

A few years ago, I would have agreed with you.

But MLB's StatCast has shown that JBJ, at times, has a negative reaction time.

I wouldn't have thought that was a thing either.

The conclusion would be that "reaction time" is not, in fact, measuring reaction time. It is measuring an aggregate of anticipation and reaction time.

As for the discussion as a whole, I tend to agree with Feczko's alternative suggestion of primarily focusing on measurements of team defense, acknowledging the noise inherent in that approach. At least we wouldn't be under the impression of false precision.

 

[Reverend]

The conclusion would be that "reaction time" is not, in fact, measuring reaction time. It is measuring an aggregate of anticipation and reaction time.

They use video. But yeah, I guess that’s what that means.

The point though is he moves before contact. The rest is words we use to define what he does.

Have you seen Gattaca by any chance?

 

[teddywingman]

Also, how do we value the positioning of Betts and Benintendi? They can almost play the line, knowing that JBJ can cover the gap.

Several times this year Mookie has made catches on the line or in foul ground, a ball that didn't seem catchable upon contact, but then you watch the replay and see where Mookie was positioned to begin with.

How do you measure JBJ's value then?

 

[williams_482]

What we do know is that the center fielder plays no role in a strikeout. He also plays no role in a popup to 2B.

That's true. It's also irrelevant, because any modern defensive metric worth a damn is already presenting results in terms of total runs saved (or outs saved, which for outfielders are functionally the same thing because the linear weights difference between a double and an out is almost exactly one run), so assessing relative value of offensive and defensive skill is completely pointless.

A +10 fielder and average hitter is just as good as a +10 hitter and average fielder (with the obvious caveat that defensive numbers have larger error bars).

Finally, although outfielders have a relatively small number of contested plays, the difference between success and failure on those plays is quite large, and the difference in catch probability between really good and really bad outfielders on those plays is much larger than the difference between a really good or really bad hitter in a single plate appearance. Raw volume of opportunities is only part of the equation.

 

[Buzzkill Pauley]

Why are we weighting anything here?

Assess run prevention. Assess run creation.

RC-RP=Outcome. Or vice versa. How did "weights" even get involved in the conversation?

That said, I agree that it is obvious that fielding/run prevention should be considered as a team metric.

I just read and re-read the article, which was a really interesting and detailed account of the new analytics since ballparks got fitted for Statcast data. Major kudos to Speier for researching and writing a great piece of actual investigative journalism rather than the opinionated schlock that usually gets printed on the sports pages. Very, very well done.

My main takeaway from the article is that when measuring one player's (in this case JDM's) defense, it's impossible to separate the "team" aspect from the "player" aspect. Statistically speaking, JDM went from a butcher to an adequate defensive corner because both his injured foot improved (player), and his positioning in the field improved (team). Without either improvement, he might be expected to remain a generally poor defensive outfielder. However, since both a.) his healed ligament now allows him to run faster, and b.) his post-Detroit coaching staffs now has him positioned better, then c.) he's now okay...not really good, but average enough defensively that he's not going to kill your pitchers.

Now, the other more conceptual takeaway of the article, IMO, is based on Speier's statement that decisions on JDM's playing time in LF/RF "appear to be based chiefly on the belief that the outfield alignment of Benintendi in left, Jackie Bradley, Jr. in center, and Mookie Betts in right is elite, rather than based on a view of Martinez as defensively inadequate."

Basically stating the obvious, Speier writes that decisions about JDM's defense are not based entirely on him as a player, but how he fits best into the team's defensive system. To take it one step farther, though, my eyeballs suggest the Red Sox have changed how their defense operates since 2016 (remember, Statcast only got up and running in 2015...). While the Red Sox have dramatically increased their use of infielder shifts over the same period, the team's also changed the way outfielders appear to be assigned ball-hawking responsibilities.

Using the back of my napkin here, I'll just say that the Red Sox appear to defense batted balls with left fielder, right fielder, and off-shifted infielder primarily responsible for lateral movement (relational to home plate); the center fielder and all three shifted infielders are primarily responsible for vertical movement. I suspect this is because Statcast has provided them with precise exit-velocity and hang-time data showing this to be an optimal arrangement for suppressing SLG.

How does JDM fit into that scheme? Just okay, because he's a just okay defensive player. But in comparison to the killer-B's outfield, just okay doesn't cut it. They can do better.

Also, how do we value the positioning of Betts and Benintendi? They can almost play the line, knowing that JBJ can cover the gap.

Several times this year Mookie has made catches on the line or in foul ground, a ball that didn't seem catchable upon contact, but then you watch the replay and see where Mookie was positioned to begin with.

The Red Sox prepare cards for the outfielders to use to have proper positioning, which is sometimes seen on TV broadcast. These are no doubt batter- and pitcher- specific.

Having said that, though, the Red Sox appear (to me) to usually do just the opposite of what you mention. Instead, they seem to more often position Betts and Benintendi (fleet-footed corner outfielders who also get good jumps and run good routes) closer to the LCF and RCF gaps than to the lines. This leaves JBJ with a narrower-width zone, but a vertically deeper one. My suspicion is that there’s less average hang time on flyballs hit in the air to the gaps than flyballs hit to the lines (until full Statcast data is public, I can’t confirm this).

And leaving aside jumps and routes, Beni and Betts have fast enough sprint speed to chase down balls hit into the corners, so they're able to still hold those sure-doubles down the lines to sure doubles instead of playing them into triples. Meanwhile, some sure-doubles-or-possibly-triples hit into the gaps are more often turned into outs than might otherwise be expected.

Which brings me to weights. Weights are useful to assessing possibilities, when data doesn't actually exist. I think they have a place in the conversation because of the way that baseball is changing with Statcast data. Where an infield shift is generally meant to turn singles into outs, outfield defense should be able to generally minimize XBH or turning them into outs outright. Because the run-expected weight of XBH are much higher than singles, optimizing outfield defense should be way more valuable to suppressing runs.

And, because I'm a firm believer in the team concept of baseball defense, I think that defense-expectancy weights will someday be used by fans (and probably already are by teams) to compare Player X's defensive value to Player Y's, within the context of a team's defensive strategy. My expectation is that this would use analysis of Statcast's raw numbers on each individual outfielder's jumps, route efficiency, and sprint speed.

Statcast defensive numbers should allow teams to project how Player X will impact the runs-against environment in comparison to Player Y; as well, it should allow a more accurate expected runs-scored comparison. But IMO using more and simpler combinations of raw defensive stats will better allow teams (and fans) to better compare defensive options than by using OAA, UZR, or DRS. Just like using a batter's triple-slash line instead of OPS better allows teams to better compare specific offensive options.

 

[teddywingman]

Good post. I'm interested in more closely watching the outfield positioning, it's just hard to do on TV. It was great to see it from the center field bleachers a few weeks ago.

 

[williams_482]

I just read and re-read the article, which was a really interesting and detailed account of the new analytics since ballparks got fitted for Statcast data. Major kudos to Speier for researching and writing a great piece of actual investigative journalism rather than the opinionated schlock that usually gets printed on the sports pages. Very, very well done.

My main takeaway from the article is that when measuring one player's (in this case JDM's) defense, it's impossible to separate the "team" aspect from the "player" aspect. Statistically speaking, JDM went from a butcher to an adequate defensive corner because both his injured foot improved (player), and his positioning in the field improved (team). Without either improvement, he might be expected to remain a generally poor defensive outfielder. However, since both a.) his healed ligament now allows him to run faster, and b.) his post-Detroit coaching staffs now has him positioned better, then c.) he's now okay...not really good, but average enough defensively that he's not going to kill your pitchers.

Now, the other more conceptual takeaway of the article, IMO, is based on Speier's statement that decisions on JDM's playing time in LF/RF "appear to be based chiefly on the belief that the outfield alignment of Benintendi in left, Jackie Bradley, Jr. in center, and Mookie Betts in right is elite, rather than based on a view of Martinez as defensively inadequate."

Basically stating the obvious, Speier writes that decisions about JDM's defense are not based entirely on him as a player, but how he fits best into the team's defensive system. To take it one step farther, though, my eyeballs suggest the Red Sox have changed how their defense operates since 2016 (remember, Statcast only got up and running in 2015...). While the Red Sox have dramatically increased their use of infielder shifts over the same period, the team's also changed the way outfielders appear to be assigned ball-hawking responsibilities.

Using the back of my napkin here, I'll just say that the Red Sox appear to defense batted balls with left fielder, right fielder, and off-shifted infielder primarily responsible for lateral movement (relational to home plate); the center fielder and all three shifted infielders are primarily responsible for vertical movement. I suspect this is because Statcast has provided them with precise exit-velocity and hang-time data showing this to be an optimal arrangement for suppressing SLG.

How does JDM fit into that scheme? Just okay, because he's a just okay defensive player. But in comparison to the killer-B's outfield, just okay doesn't cut it. They can do better.



The Red Sox prepare cards for the outfielders to use to have proper positioning, which is sometimes seen on TV broadcast. These are no doubt batter- and pitcher- specific.

Having said that, though, the Red Sox appear (to me) to usually do just the opposite of what you mention. Instead, they seem to more often position Betts and Benintendi (fleet-footed corner outfielders who also get good jumps and run good routes) closer to the LCF and RCF gaps than to the lines. This leaves JBJ with a narrower-width zone, but a vertically deeper one. My suspicion is that there’s less average hang time on flyballs hit in the air to the gaps than flyballs hit to the lines (until full Statcast data is public, I can’t confirm this).

And leaving aside jumps and routes, Beni and Betts have fast enough sprint speed to chase down balls hit into the corners, so they're able to still hold those sure-doubles down the lines to sure doubles instead of playing them into triples. Meanwhile, some sure-doubles-or-possibly-triples hit into the gaps are more often turned into outs than might otherwise be expected.

Which brings me to weights. Weights are useful to assessing possibilities, when data doesn't actually exist. I think they have a place in the conversation because of the way that baseball is changing with Statcast data. Where an infield shift is generally meant to turn singles into outs, outfield defense should be able to generally minimize XBH or turning them into outs outright. Because the run-expected weight of XBH are much higher than singles, optimizing outfield defense should be way more valuable to suppressing runs.

And, because I'm a firm believer in the team concept of baseball defense, I think that defense-expectancy weights will someday be used by fans (and probably already are by teams) to compare Player X's defensive value to Player Y's, within the context of a team's defensive strategy. My expectation is that this would use analysis of Statcast's raw numbers on each individual outfielder's jumps, route efficiency, and sprint speed.

Statcast defensive numbers should allow teams to project how Player X will impact the runs-against environment in comparison to Player Y; as well, it should allow a more accurate expected runs-scored comparison. But IMO using more and simpler combinations of raw defensive stats will better allow teams (and fans) to better compare defensive options than by using OAA, UZR, or DRS. Just like using a batter's triple-slash line instead of OPS better allows teams to better compare specific offensive options.

Great post.

I'm reminded of this recent post by tangotiger, which looks like a "sneak peak" at an early version of the defensive metrics you are describing.

Something else I found interesting: we have had good component/peripheral statistics for batters for over a century now, but (excepting early visionaries like F.C. Lane and Allan Roth) we didn't know how to translate them into runs and wins until relatively recently. Defensive statistics, meanwhile, were at first largely useless, and once we had genuinely useful ones, the internals were so complicated that presenting them in any form other than as outs created or runs saved wasn't really practical. It is only now, with Statcast and the like, that we are finally starting to tease out peripheral numbers for defensive statistics that can tell us what a player is actually doing to create value on defense.

 

[Plympton91]

That's true. It's also irrelevant, because any modern defensive metric worth a damn is already presenting results in terms of total runs saved (or outs saved, which for outfielders are functionally the same thing because the linear weights difference between a double and an out is almost exactly one run), so assessing relative value of offensive and defensive skill is completely pointless.

A +10 fielder and average hitter is just as good as a +10 hitter and average fielder (with the obvious caveat that defensive numbers have larger error bars).

Finally, although outfielders have a relatively small number of contested plays, the difference between success and failure on those plays is quite large, and the difference in catch probability between really good and really bad outfielders on those plays is much larger than the difference between a really good or really bad hitter in a single plate appearance. Raw volume of opportunities is only part of the equation.

Those large error bars you acknowledge are precisely because the defensive metrics, even the most advanced ones, are no better than plus / minus ratings in hockey.

The statcast stuff Reverend mentioned is very useful for telling us the answer to the question, “Which outfielder has the best range?” And that’s useful information. But if someone is +10 in DRS or whatever likeminded stat you prefer, that means nothing until you tell me who was flanking him and how good the team’s advance scouts are.

Likewise, I’ll stipulate that a great play by an outfielder can often save 2 or 3 or even 4 bases. Isn’t it likewise true that a great hit by a good offensive player can generate 2 or 3 or 4 bases? A great 10 pitch at bat can cause wear and tear on an opposing pitcher the same way a great catch can save pitches for your pitcher?

And a hitter gets 3-5 at bats every single game while a defender gets maybe one opprtunity a game to make a non routine play.

Conclusion: offense is more important than fielding. Full stop.

 

[Reverend]

Those large error bars you acknowledge are precisely because the defensive metrics, even the most advanced ones, are no better than plus / minus ratings in hockey.

The statcast stuff Reverend mentioned is very useful for telling us the answer to the question, “Which outfielder has the best range?” And that’s useful information. But if someone is +10 in DRS or whatever likeminded stat you prefer, that means nothing until you tell me who was flanking him and how good the team’s advance scouts are.

Likewise, I’ll stipulate that a great play by an outfielder can often save 2 or 3 or even 4 bases. Isn’t it likewise true that a great hit by a good offensive player can generate 2 or 3 or 4 bases? A great 10 pitch at bat can cause wear and tear on an opposing pitcher the same way a great catch can save pitches for your pitcher?

And a hitter gets 3-5 at bats every single game while a defender gets maybe one opprtunity a game to make a non routine play.

Conclusion: offense is more important than fielding. Full stop.

Surely things like the Law of Diminishing Returns and issues of Comparative Advantage and the like come in to play, no?

 

[Plympton91]

Surely things like the Law of Diminishing Returns and issues of Comparative Advantage and the like come in to play, no?

Laws of diminishing returns could come into play, sure. You can’t have 3 JD Martinez’s in the outfield (or Jose Canseco in RF with Kevin Mitchell in LF)

And it’s possible that Bradley is such an outlier defensively that even using my weighting scheme he is a net benefit over a much stronger hitter.

But that decision would still be mostly informed by art, and would have to include unmeasurable stuff like pitchers’ confidence knowing anything hit to CF might be an out, in addition to appropriately discounted and heavily regressed pseudo-science of defensive runs saved.

 

[Cesar Crespo]

And a hitter gets 3-5 at bats every single game while a defender gets maybe one opprtunity a game to make a non routine play.

Conclusion: offense is more important than fielding. Full stop.

And for 4.5 of those at bats, the good hitter will end up with the same results as the bad hitter. Will the defender have a better success rate than 10% to make a difference in that one non routine play (although I think 1 non routine play a game is a pretty high number)?

The difference between a good hitter and a bad one is probably a lot smaller than you think on a per PA level and 5 of them added together may very well add up to less than 1 non routine play. If that number is more like half a routine play, maybe it shifts back to hitting. Again, I think 1 is an incredibly high number. Even a half seems high.