Earlier this month rumors flew around that the Philadelphia Phillies were interested in interviewing Mets hitting coach Chili Davis for the same position. Shortly after those rumors leaked out, Andy Martino reported via Twitter that the Mets had reached multi-year deals to retain Chili Davis and assistant hitting coach Tom Slater in their current positions.

While the Mets were far from an offensive juggernaut – they ranked 11th in team wOBA – they did have plenty to be happy about in terms of increased production from the lineup in 2019. In the simplest of all stats, the Mets scored 791 runs during the 2019 season after plating on 676 in 2018. But it would be foolish to simply say that the Mets scored 115 more runs this season than last, and Chili Davis deserves all of the credit for that.

But it is important to figure out how much Chili Davis might have had to do with this improvement, and along with it to determine if the organization made the right move in retaining him as hitting coach on a multi-year deal. It is nearly impossible to pinpoint precisely how big of a role his coaching played, but the purpose of this analysis will be to look at the issue in a broader sense.

Methodology

To determine how much credit Chili Davis should get for the Mets increased offensive production, I will look at two main statistics – wOBA and HR/FB%. The reasoning for picking these two over all other stats is simply that wOBA provides us with one offensive number that fully encapsulates hitting, and HR/FB% to provide context within the long ball explosion that MLB has experienced.

This is a statistical approach to examining Chili Davis’ impact on Mets hitters, but it is far from a holistic one. As such it will be inherently incomplete. The analysis to come does not factor in the hundreds of daily tweaks and meetings between hitting coach and hitter that take place throughout the season because it cannot. Only Chili Davis and Mets hitters can speak to that. Embedded in these data to come are the fruits of “good process, bad results” and also “bad process, good results.”

Chili Davis may have an impact on team chemistry that is impossible to measure through these numbers, but in a year with plenty of turmoil with manager Mickey Callaway could have been integral to the team not falling apart. Unless we were in the locker room, there is no way to truly know. All of this is to say that the following analysis will only paint part of the picture of what Chili Davis brings to the Mets as hitting coach, but it is the only side we can objectively examine.

I will use these data on both a team-wide basis and on an individual level in an attempt to see how much better the Mets improved relative to Major League Baseball, and to parse out how much of that improvement came from team newcomers in comparison to improvements made by returners. The goal is to weed out the noise of the juiced baseball and new player acquisitions as best as possible to get at the heart of Chili Davis’ contribution to the Mets in 2019.

Analysis

Examining batted ball statistics for the Mets as a team and on an individual basis shows a demonstrative change in the team’s offensive production from 2018 under Kevin Long and 2019. In terms of wOBA, the Mets offense, including pitchers, saw a jump from a .305 clip in 2018 to .325 in 2019. The 20-point increase was the third-largest jump in MLB, surpassed by only the Minnesota Twins (34-point increase) and the Houston Astros (28-point increase).

Of note, during the offensive explosion of 2019 MLB-wide wOBA increased five points, from .315 in 2018 to .320 in 2019. This means the Mets went from being 10 points worse than league average to five points better¹.

Table 1. – Chili Davis-coached teams and year-end wOBA ranking

Team wOBA MLB wOBA Rank
2012 Athletics 0.311 20th
2013 Athletics 0.327 3rd
2014 Athletics 0.312 13th
2015 Red Sox 0.321 6th
2016 Red Sox 0.346 1st
2017 Red Sox 0.316 20th
2018 Cubs 0.321 11th
2019 Mets 0.325 11th

One of the driving factors in the Mets’ increased offensive production in 2019 was the large jump in HR/FB%. In 2018 the Mets had a HR/FB% of 10.9%, 14.2% less than the league average rate. That rate spiked in 2019 to 16.8%, which was 9.8% better than the MLB-average rate. The 5.9% jump represents a 54% increase in HR/FB%, which is the third-largest in baseball this season, trailing only the Twins and Chicago Cubs.

For good measure, as a team the Mets had an average exit velocity of 88.4 mph, better than the 87.5 league average and another significant increase from the 87.2 mph team-wide mark in 2018 according to Statcast data pulled from Baseball Savant.

But how much credit does Chili Davis actually deserve for this? The honest answer is that it is impossible to truly tell. It is undeniable that that newcomers to the lineup in 2019 – particularly Pete Alonso, J.D. Davis and Wilson Ramos – had a meaningful impact on the overall production of the lineup. Removing Alonso from the Mets lineup alone drops the team’s HR/FB% for the season from 16.8% to 14.9% – the difference between being 9.8% better than the league to 2.6% worse.

Despite the disappointment of Robinson Cano and Jed Lowrie, General Manager Brodie Van Wagenen’s acquisitions during the offseason on the whole made the Mets a better offensive team. Though Ramos’ numbers took a huge dive (wOBA fell from .361 to .327) he was still an upgrade at catcher and J.D. Davis’ breakout season provided a lift at third/left field over the various players the Mets used in 2018.

The Sandy Alderson administration and the Mets minor league player development staff (and the juiced baseballs) deserve the credit for bringing Alonso’s record-setting rookie season to fruition. Those factors explain a good chunk of the Mets improvement on offense, but then there is Chili Davis. He was hired as the Mets hitting coach because he preached a style of hitting that broke from the somewhat recent trend of launch angle-oriented hitting.

The hope was that a different message and approach would help Mets hitters, and therein lies at least part of our answer of how much credit Chili Davis deserves for the increased offensive production. To seek the answer, I analyzed offensive statistics for the 10 position players who suited up exclusively for the Mets at the MLB level in both 2018 and 2019².

Of that population only Brandon Nimmo saw his HR/FB rate decrease from 2018 to 2019, a drop from 17.5% to 16.3%. His attempt to play through injuries and his extended time on the Injured List this season likely have far more to do with this than his new hitting coach does. Nimmo was one of two players, along with Juan Lagares to exhibit a drop in wOBA from year-to-year (45 points and 71 points, respectively).

But with those two negatives is a whole lot of positive and meaningful improvement from the other eight hitters in the population. The population in 2018 posted a .325 wOBA and 12.5% HR/FB%. That exact same population in 2019 had a .334 wOBA and 14.8% HR/FB%, with the cumulative numbers in both years weighted by plate appearances.


1 – The standard deviation in the rates of change in wOBA from 2018 to 2019 was 10 points, meaning the Mets were two standard deviations better in 2019 than in 2018. From a statistical standpoint we can be 95% confident that the Mets superior offense in 2019 is not the work of random variation, but actual, legitimate improvement.
2 – The players: Amed Rosario, Brandon Nimmo, Dominic Smith, Jeff McNeil, Juan Lagares, Luis Guillorme, Michael Conforto, Todd Frazier and Tomas Nido.

Table 2. New York Mets wOBA and HR/FB%, returning players

2019 Season PA wOBA HR/FB   2018 Season PA wOBA HR/FB   HR/FB inc wOBA inc
Amed Rosario 655 .318 10.3%   A. Rosario 592 .290 7.0%   3.3% .028
Brandon Nimmo 254 .340 16.3%   B. Nimmo 535 .385 17.5%   -1.2% -.045
Dominic Smith 197 .368 22.4%   D. Smith 149 .288 13.2%   9.2% .080
Jeff McNeil 567 .384 15.4%   J. McNeil 248 .368 3.8%   11.6% .016
Juan Lagares 285 .258 7.9%   J. Lagares 64 .329 0.0%   7.9% -.071
Luis Guillorme 70 .299 5.6%   L. Guillorme 74 .242 0.0%   5.6% .057
Michael Conforto 648 .358 20.5%   M. Conforto 638 .342 19.7%   0.8% .016
Todd Frazier 499 .328 14.3%   T. Frazier 472 .302 12.9%   1.4% .026
Tomas Nido 144 .227 13.8%   T. Nido 90 .193 6.7%   7.1% .034
Weighted Avg.   .334 14.8%   Weighted Avg.   .325 12.5%    

While the HR/FB% in the population increased by 18.4%, MLB-wide that stat was up by 20.4% in 2019. In terms of wOBA, the population experienced a 2.8% bump while MLB experienced a 1.6% greater wOBA this season. It looks good that the Mets improved offensively and outperformed baseball at large, but the nine-point wOBA increase for the population represents less than one standard deviation of improvement and is not statistically significant.

If you want to cherry-pick data and remove Nimmo and Lagares from the population, the returning players wOBA jumps to .341, a 31-point increase over that same eight-player population in 2018, which is statistically significant. I’m torn on whether that is fair to do or not since every team experiences injuries so it makes the comparison to league-wide data more suspect, but it’s not fair to allow injuries and more playing time for a lesser player to make the improvements made elsewhere seem less significant.

Conclusion

Even the staunchest Chili Davis supporters must concede that the team’s increased offensive output was heavily influenced by to Van Wagenen’s acquisitions and Alonso’s mammoth rookie year, but that doesn’t mean that he doesn’t deserve a ton of credit for the Mets’ improved offense. It is entirely possible that the new players were as good as they were because of Chili Davis, but it is equally as possible that his tutelage might have held them back from bigger seasons. But we can never objectively know that, and we are searching for objectivity.

Dominic Smith and J.D. Davis, among other Mets anecdotally spoke incredibly highly of working with him and gave him a lot of credit for their performance this season. But the problem is simply that when dealing with only one year of data – a few hundred plate appearances per player – a lot of random variation can get in the mix. As much as I attempted to weed that out, it is impossible to do so without a much larger sample.

As Table 1 shows, year-to-year wOBA performances can be fickle, and just because a team finishes with a good number one year doesn’t mean they will be good again the next, and vice versa. There exist simply too many other factors that can cloud the data. That being said, the Mets did achieve some statistically significant improvement from 2018 to 2019, and Chili Davis is at least somewhat to credit for that development. The Mets did the right thing in retaining him as hitting coach moving forward.

Takeaways
1. The Mets had the third-most improved offense in Major League Baseball in 2019.
2. A portion of that improvement came from the contributions of newcomers J.D. Davis and Wilson Ramos and rookie Pete Alonso.
3. Chili Davis oversaw a significant improvement in the hitting performance of players who returned from 2018.

7 comments on “Did Chili Davis actually help the Mets offense?

  • Joe Vasile

    I didn’t mention this in the article, but I was surprised at how much the Mets improved this year offensively under Chili Davis. I went into researching this topic believing that Pete Alonso and the offseason acquisitions were most to credit for the better offense, but the numbers suggest that Davis had a positive impact.

  • John Fox

    From the article: “offensive production from 2018 under Kevin Long and 2019. ”

    Kevin Long’s last year as hitting coach for the Mets was 2017, Pat Roessler was hitting coach in 2018.

  • Brian Joura

    Joe – good to read you here again! I enjoyed this piece because it made me think. Here are some things that jumped to mind reading this:

    For the team-wide increase in runs from 2018, along with the addition of Alonso and J.D. Davis, there’s also the decrease in PT for guys who stunk. By my count, the Mets had 909 PA go to non-pitchers with a sub-.700 OPS in 2019. The year before that number was 2,830.

    Was it Davis as hitting coach or not giving loads of PT to Kevin Plawecki and Jose Reyes and Austin Jackson and Adrian Gonzalez?

    As for the direct comparison to guys who played for the team in 2018 and 2019 – there are too many variables to contribute all of the success to Chili Davis. There are young guys improving with more MLB experience, some BABIP risers, some guys being healthier than a year ago.

    I feel like you have to demonstrate something that Chili actually did to credit him for any increase in production that may have happened. Otherwise it’s just crediting the rooster for the sun rising because it happened on his watch.

    The narrative was that Chili was going to emphasize hitting to all fields over the pull the ball, go for power mindset that existed previously. Given that, I have a really hard time crediting Chili for any increase in HR/FB — unless you can show that the increase came about due to opposite field homers. And what does the wOBA increase look like if you remove the HR?

  • TJ

    Brian,
    Lots of high math there, nice. I like that math, it makes sense and is reasonably objective. Even teams like the Mets use it as a tool, but my gut tells me that hitting coaches live and die more by perception and vibe than by empirical numbers. The Mets improved and raked better in 2019, Chili rules. If they are hitting .220 and 10 back in May, he’ll be a bum. Lol. That said, I like his approach and glad he’s back.

  • José

    I (and other self-righteous snobs) do not consider statistics to be “higher math” or any kind of math for that matter. I (and others) consider it to be “mathematical science”

    Just thought I’d chime in with a comment which can’t be considered unrelated to the topic of discussion

    Is being snarky in violation of posting rules?

    • Mr_Math

      You seem like a nice guy, José, but you’re trying too hard to demonstrate your relevancy. Also, your statement about being a self-righteous slob I interpret as you being a mathematician, something with which I can relate.

      I’ve been looking over my grad school transcript, and I found 4 graduate level prob/stat courses which I took:

      Math 5723 Intro To Mathematical Statistics
      Math 5773 Advanced Applied Regression Analysis
      Math 5793 Advanced Applied Statistics
      Math 5803 Advanced Probability

      Further, I noticed I received grade of “A” in 3 of these, and a “B” in 5723. Now, you think it would be valuable to have an expert in Probability and Statistics regularly visiting this here blog.

      Alas, I hardly remember any of it, probably because I found it of little interest. If, however, anyone has questions about Galois Theory (the endpoint in a year-long Abstract Algebra course), I can definitely answer these since that is my area of expertise.

      So, inevitably, I’ll just return to being the dull child in this here sandbox, and I can’t come close to expressing the delight I feel in having that status!

  • Chris B

    Hey Joe welcome back to M360 and what a great reunion article! Thank you for your diligent research and explanation of your process. I agree with Brian that this piece certainly made me think on a few different levels.

    I think that your isolation of the 2018 players is sound and I agree with your gut in leaving Nimmo and Lagares in the sample group. That’s how you combat Brian’s point above re: young guys, babip, etc.

    Clearly another season of data points will either support or denounce the article’s key takeaways. But then we will be able to restructure our thinking on the matter and adapt. For the information at our disposal and the small sample size, this is a well constructed set of analysis.

    Things to look out for next year:
    – Players who played all three seasons ’18-’20 (Mcneil, Conforto, etc)
    – The “new” Mets (Alonso, Davis, etc) and relationship of ’19 and ’20
    – Batted ball profile consistency of ’19 to ’20

Leave a Reply

Your email address will not be published. Required fields are marked *

The maximum upload file size: 100 MB. You can upload: image, audio, video, document, spreadsheet, interactive, text, archive, code, other. Links to YouTube, Facebook, Twitter and other services inserted in the comment text will be automatically embedded. Drop file here