Lucas Duda, Everyday Player
Over the last six months, Lucas Duda has probably heard a lot about last season’s .180 batting average against left-handers. A quick search reveals more than half a million hits for “Lucas Duda .180” and all the results in the first five pages were relevant. Fortunately for Duda, that unsightly .180 batting average in just 125 plate appearances is not relevant when it comes to evaluating him as a hitter. Duda had an excellent season in 2014 that saw him record 30 home runs and establish himself as one of the better hitters in the league, despite questions persisting about his ability to handle left-handed pitching. With two home runs yesterday off left-handed pitchers, Duda equaled his total from all of 2014 in one day. With four of his five home runs on the season coming off of southpaws complementing his power surge against righties from last season, Duda has helped cement his status as an everyday player and removed the prospect of a platoon from the equation.
After Duda’s poor results against left-handers last season, the New York Mets could have placed Duda in a platoon, benching him against lefties. Duda never had particularly good minor-league numbers against lefties, either, reinforcing the logic of such a move. After the signing of Michael Cuddyer, there was some talk of letting Cuddyer take plate appearances at first base against lefties and making Duda a straight platoon player. After the signing, Duda said all of the right things but maintained confidence that he could still hit lefties.
As far as the lefties go, I feel like in the past I’ve hit lefties a little bit better than I’ve shown…But if he has to spell me at first base versus lefties, whatever is good for the team, whatever is going to help us win, I’m all for. I’ll come off the bench or come in late in the game or whatever it is. Anything to help us win. That’s our main goal. We’ve got to accomplish it this year.
Making the decision about a left-handed hitter with good talent and solid overall hitting numbers is a tough one. Platoon statistics for hitters don’t stabilize until around 1,000 plate appearances, but waiting five to eight seasons to determine if a lefty like Duda will come around requires patience that is difficult when better options are immediately available. Since the beginning of 2011, there are 48 left-handed hitters, including Duda, with at least 500 plate appearances against lefties. The four best left-handed hitters facing lefties during that time have been Joey Votto, David Ortiz, Prince Fielder, and Curtis Granderson, each with a wRC+ of at least 120 over the last five years. Those players had varying degrees of success against lefties to start their careers.
Joey Votto was always good against lefties, with a 132 wRC+ in his first 619 plate appearances against lefties, and Prince Fielder was decent with a 105 wRC+ in his first three full seasons. Curtis Granderson did not fare as well, producing a poor wRC+ of 72 through his first 472 plate appearances. Even removing David Ortiz’s Minnesota Twins seasons and focusing solely on his Boston Red Sox career, David Ortiz had just an 81 wRC+ in 471 plate appearances during his first three years in Boston. Heading into this season, Lucas Duda has recorded a 75 wRC+ against left-handers in 469 plate appearances.
Focusing in on the .180 batting average, or even his overall batting line, from 2014 is too small of a sample, just as relying completely on his fast start in 2015 would be a mistake. So far in 2015, Carlos Gonzalez, Michael Morse, Mike Napoli, Chase Utley, and Melky Cabrera all have a sub-70 wRC+ in about the same number of plate appearances as Duda had last season against lefties. While those players might have some underlying issues causing their poor performance thus far, we would never look solely at their performance in 2015 and expect the same miserable production going forward.
After showing promise as a hitter in 2011, Duda slumped a bit during his sophomore season, but has posted better numbers in each of the last two seasons, leading Duda towards star territory last season. Since moving to first base, his WAR is no longer held down by his disastrous fielding, and over the last calendar year his 4.6 WAR is 26th among all qualified hitters. His 145 wRC+ ranks eleventh by that measure, and his 30 home runs place him tenth in that time. Among first basemen in the last year, he sits behind only Anthony Rizzo, Miguel Cabrera, and Jose Abreu and he ranks ahead of Freddie Freeman, Adrian Gonzalez, and Edwin Encarnacion.
His early-season power surge against left-handed hitters has also brought his career numbers against them to an acceptable level. Of the 48 left-handed hitters with at least 500 plate appearances against lefties since the beginning of 2011, Duda ranks right in the middle, as the table below illustrates.
| Rank | Name | PA | BB% | K% | OBP | SLG | wRC+ |
|---|---|---|---|---|---|---|---|
| 18 | Bryce Harper | 515 | 11.3 | 24.1 | 0.330 | 0.396 | 102 |
| 19 | Brett Gardner | 569 | 9.3 | 21.3 | 0.338 | 0.381 | 100 |
| 20 | Brian McCann | 624 | 4.3 | 18.1 | 0.293 | 0.436 | 100 |
| 21 | Jon Jay | 509 | 7.5 | 16.5 | 0.344 | 0.358 | 99 |
| 22 | Jay Bruce | 780 | 7.3 | 28.7 | 0.287 | 0.444 | 97 |
| 23 | Ichiro Suzuki | 708 | 4.1 | 8.8 | 0.327 | 0.373 | 95 |
| 24 | Lucas Duda | 500 | 8.8 | 30.6 | 0.315 | 0.360 | 94 |
| 25 | Ben Revere | 648 | 3.1 | 9.9 | 0.334 | 0.360 | 93 |
| 26 | Nick Markakis | 845 | 7.2 | 13.8 | 0.337 | 0.349 | 91 |
| 27 | Chase Utley | 693 | 9.7 | 14.3 | 0.315 | 0.362 | 90 |
| 28 | Kyle Seager | 806 | 5.5 | 19.1 | 0.289 | 0.381 | 89 |
| 29 | Eric Hosmer | 830 | 6.0 | 17.3 | 0.312 | 0.369 | 88 |
| 30 | Carlos Gonzalez | 663 | 5.9 | 23.8 | 0.311 | 0.433 | 86 |
His strikeout rate over 30% is still unsightly, but Duda’s numbers are actually better than many of his peers’ against left-handed pitching. Over the past few seasons, the highest collective seasonal mark for lefty-against-lefty matchups among non-pitchers has been the 90 wRC+ figure from 2010. Last season the wRC+ was 85; in 2013 it was 81; and it’s 90 in the early part of this season. He hasn’t made changes to his swing to better adjust to lefties, and his swing looks similar to the one Eno Sarris profiled at the beginning of 2013. Here it is on a walk-off home run last September.
He pulled that one down the right-field line, but in his two home runs yesterday, he used more of the middle of the field. Here’s the first one against Jaime Garcia.
Here’s the second one, against Randy Choate, who had given up just two home runs to left-handed hitters since the beginning of 2012 before yesterday, and has held lefties to a .187/.269/.268 line over the course of his career.
Both of yesterday’s home runs went towards the middle part of the ballpark, which could be random, although it is possible Duda has changed his approach slightly this season. In December, Duda mentioned perhaps being a little pull-happy against left-handers, and Keith Hernandez was getting some credit yesterday after spending some time with Duda in Spring Training. The numbers so far this season do lend some credibility to Duda’s pull-heavy theory, but we have very few plate appearances to go on. From 2011 to 2014 against left-handers, he pulled the ball 44% of the time. In a small number of plate appearances this season, he is pulling the ball 32% of the time, with most of the lost pulled balls heading toward the middle of the field.
Duda retained confidence in his ability to hit lefties, and a very poor 2014 against lefties was in stark contrast to his performance against righties, but early success this season is not due to a simple flip of a switch. Duda has had limited plate appearances in his career against lefties and it can take awhile for players to adjust to the difficulty of hitting without the platoon advantage against the best pitchers in the world. Joey Votto came to the majors ready to hit lefties, but it took awhile for Curtis Granderson and David Ortiz.
Fifty plate appearances with great results is not an indication that Duda is on his way to invincibility, but his career numbers do indicate that he might be competent against left-handed pitchers. This season will not be enough to make a determination on Duda’s ability to hit his way out of a platoon. He will need a few more years before we can be sure, but the Mets’ certainly looks smart in giving Duda the opportunity to be an everyday player. Fifty plate appearances do not mean much over the course of a career, but for Duda, changing perceptions in these 50 plate appearances after his 2014 season, even though it’s not sustainable, could mean the difference in Duda’s role for the rest of his career.
Craig Edwards can be found on twitter @craigjedwards.
Can we all just agree that this is Fangraphs and we know what a small sample size is?
This is a good article – but I felt awkward the last paragraph where the author needed to backtrack almost all of what he wrote because of the sample size.
We know it’s SSS, and we know you know it’s SSS, I wish we didn’t always have to devote column inches to caveats. This is a safe place.
And again, just in case the sentiment was lost — great article. Thanks!
The problem is, without SSS there’s literally nothing here to discuss. The SSS is the question the post is about — what to do in response to it and how to draw conclusions despite it. I agree that the disclaimer often feels redundant, but here it’s not a tangential footnote, it’s the main subject under discussion.
SSS issues aside, I’m confused by the author’s references to Duda having “acceptable” or “competent” career numbers vs LHP’s.
A career .299 wOBA / 90 wRC+ is not acceptable for a 1Bman. That’s below replacement. And his career .322 BABIP vs LHP’s is 31 points higher than his BABIP vs opposite-handed pitchers.
Over his career, Duda has been worth -6 runs vs LHP’s, in 519 PA’s. If you look at his overall career numbers and remove the batting component (leaving baserunning/fielding/position/league adjustment/replacement runs), you’ll find he’s been worth -15.7 runs below replacement– or -4 runs per 519 PA’s. Add his batting runs vs LHP’s back in (-6 runs in 519 PA’s), and you’ve got a player who’s been worth 10 runs BELOW replacement vs lefties for his career.
Some of the issue is roster construction as what is acceptable or competent for 125 PAs at first base isn’t the same as what you would want over 600 PAs. A lot of the negative value referenced above occurred because Duda was playing OF. The data isn’t complete but we have 2200 innings of Duda at first and he’s been average. Adding in base running, etc., that’s actually about 5 runs above replacement per year. He’s been six runs below average L v L in his career (4 since 2011), making him essentially replacement level as a first baseman facing lefties.
So with limited roster spots and resources, you might get 1-2 more runs by finding an average hitter to play first base in limited time. Backups that are at least average hitters are not easy to find, thus Duda’s output is acceptable. Adding in the potential to improve given Duda’s overall success as a hitter, and it doesn’t make a lot of sense to replace him against lefties.
Fair points. Thanks for the response.
I still definitely would not call a player around replacement level “competent”, even if that’s the only option a team has at that position.
does the change in the RF fence give him more confidence that his fly balls might go out of the park, and affect him mentally? both of those hr’s yesterday were homeruns this year or last, but I wonder if he was gunning for the line too much last year (the space between modell’s clubhouse and the foul pole), but now is more open towards all of RF.
*Fantasy Alert*
I targeted Duda in my draft as a cheap source of power, but I also drafted a back-up 1B that I would put into my lineup whenever the Mets faced a lefty. I’m not mentioning this because anybody cares about my team. I’m mentioning it because I agreed with the masses – he’s a good hitter overall, but useless against lefties. *Alert over*
With that out of the way, I have to give Terry Collins credit for giving him a shot this year. I read a couple months ago that he would not platoon Duda in the beginning of the season because he wanted to give him a chance to prove he can hit lefties. I thought he was bonkers, especially after they signed Cuddyer and Mayberry, but that has paid off so far.
SSS obviously applies and it’s not enough at bats to show that he can actually hit lefties, but I think it’s enough of a sample to suggest that he’s better than what he showed last year. After all, he has 18 hits off of lefties this year so far, and he only had 20 all of last year (in 75 more PA).
On the other hand, this seems to make the Cuddyer signing even worse. If I turned my head sideways and squinted enough, I could kinda see why they would sign him to play some LF and to start 20-30 games at 1B. Without the platoon issues, there’s less use for someone like Cuddyer – and certainly not enough to warrant a 1st round draft pick.
First off, I did the same thing with Duda in my fantasy leagues. Last year I paired him up with the similarly challenged Adam Lind, benching whichever drew a LHP or moving Duda to an OF spot when both match-ups were favorable.
I remember reading last year that he did most of damage on fastballs. I couldn’t quickly find the data but from watching him he seems to be doing a better job recognizing and squaring up on the straight stuff from LHP this season. All that said, the .467 BABIP, 36% %LD and 28.6% HR/FB against lefties all scream for regression.
Regarding the SSS comments above, why is that we always look at stabilization points? Wouldn’t it be better to be looking at confidence intervals. That way hitting .390 off of lefties for 50 ABs can be placed in a definite context. We seem to always focus on the sample size but not the performance. Looking at the CI can tell you if hitting .390 for 50 ABs is more or less impressive than, say hitting .280 against lefties for 150 ABs and show the exact error bounds for both situations.
Fantastic point Jon!
Yeah, I think it would make sense to see a ton more confidence intervals in baseball writing… maybe we need to popularize methods and online tools for it?
For any binomial variable it would be dead easy to provide a little calculator, user puts in rate stat and number of trials. OBP, HR%, etc.
For stats like SLG or FIP I’d have to learn how to do it… Simulation if nothing else.
Subtle but important point: CIs and ‘stabilization point’ are answering different questions, unless you generate CIs with a prior assumption about the distribution the player is drawn from.
That’s a big part of what the stabilization results are telling you: how wide the spread is in the population. Effectively being compared against the spread of a naive CI.
I foresee discussion. 🙂
Yes! I would love to see this type of work.
Just a quick calculation which is not entirely accurate but the type of thing that I’d find interesting:
Ignoring the rest of Duda’s career, and if we say batting average is strictly binomial, then a 95% lower bound for batting average given a sample of hitting .390 in 50 at bats is: average vs lefties >= .254 with 95% confidence.
If you did the same thing with Chase Utley’s atrocious season you’d find a 95% CI that he is under a .233 hitter. Now, this of course ignores previous history which isn’t entirely fair. And also I’m not sure what to use to compute the standard deviation. Maybe recent history’s BA could be used to incorporate that a bit, or you could use league averages, I’m not sure and I’d have to think about that a bit more. Just wanted to quickly make a calculation to see what it would even look like.
ANOTHER CRAIG EDWARDS METS PUFF PIECE.
Alternate title:
“Lucas Duda’s 180 Against Lefties”
That actually is good enough to deserve an edit!
This song should be played when we are reminded of SSS ..looking down upon our tiny brains
https://m.youtube.com/watch?v=dw9qqvm-LT8
Duda can hold his own against lefties and be league average to above average I think the Mets will be satisfied
Informative article, but please, please proofread before you lost. Noticed at least a half-dozen typos in there…gheesh.
Proofread before you lost?
The idea of a platoon coming into the season is a joke. If anyone deserves to be in the lineup every single game for the Mets, its Duda. Without him that lineup is pathetic.
I’m something of a hardcore-analytics novitiate, but isn’t the only pertinent question whether Duda will hit lefties better than Cuddeyer’s replacement in LF would do so? Like, what actual benchmark does he have to exceed for everyday-Duda to be a min-maxing choice right now?
That is one important factor. John Mayberry has hit lefties well in his career, and you probably would expect better production than Duda this year bit not with a great level of confidence. Duda’s potential and overall talent as a hitter weigh in Duda’s favor as far as getting him the majority of the playing time.
Off topic, but doesn’t this point to issues with defensive WAR? I thought the positional defensive adjustment was supposed to make it such that a player’s defensive WAR shouldn’t change appreciably if he changed positions. So shouldn’t Duda have the same defensive rating at 1B vs. in the OF? If not, doesn’t that mean our positional adjustment factors are off?
The adjustment is based on the offensive levels of production by position overall, and not directly tied to defensive talent. Andrelton Simmons likely couldn’t move from ss to first and save even more runs, most players if moved to catcher would make the game unrecognizable, and a catcher cogent move to short and be equally effective. Some players aren’t meant to be outfielders, and Duda suffered out there, but that doesn’t mean he isn’t well suited to be an average 1B or that the positional adjustment needs changing.
A good example of this is to think about what would happen if Texas moved Andrus to 1B and Fielder to SS. According to the positional adjustments, it would be a wash. But obviously their defense would be a mess if that happened. Fun to watch, but a mess.