Photo by Joshua Sarner/Icon Sportswire
I should preface this by saying there’s a good chance I’m completely wrong here. I consume fantasy baseball articles and podcasts like I chug milk after eating hot wings, and I have not heard any of the analysts, who are far more knowledgeable than me, say that there’s any reason at all to be concerned about Nolan Arenado. I feel like Ian Malcolm giving a warning in Jurassic Park, except without the nerd sex appeal of Jeff Goldblum, and also in this case, my colleagues are more savvy mathematical minds. Maybe Titanic is a better analogy, since over the latter half of this decade, Arenado has been an unsinkable ship and has seldom changed course.
If I named this article “Why I Am Slightly Concerned About Nolan Arenado,” I doubt it would be any less controversial. After all, he has been among the most consistent all-around players in the major leagues both offensively and defensively. Sure, Mike Trout has been better overall, but Arenado has been more consistent since his breakout 2015 season, with batting between .287 and .309 and hitting between 37 and 42 home runs every year since.
In addition to his elite hitting, he just won his 6th consecutive Gold Glove and his second consecutive Platinum Glove for the best overall defender in the National League. A recent Hardball Times article also claimed that according to Baseball Prospectus’s new all-encompassing metric, DRC+, Arenado may actually be underrated. His 146 DRC+ in 2018 was 5th-best in baseball, and since 2012, even with park factors accounted for, Arenado’s DRC+ over that span suggests he has been only a smidgen worse than Mike Trout. Not bad company to keep. No wonder why he’s about to earn the highest player salary in arbitration, regardless of whether he gets the $30 million he’s asking for, the Rockies’ figure of $24 million or somewhere in between. When I say I am concerned about him, I am absolutely factoring in the sky-high expectations currently placed on him.
In 2018, Arenado’s stat line, while not his best season, was smack dab in the middle of his expected range of outcomes, hitting his 38 homers with a .297 Average, a career-best .374 OBP and a .561 SLG%, which was his lowest since before his 2015 breakout, but only slightly (he slugged .570 in 2016). Sure, his run production was down with his 110 RBI falling short of the illustrious 130 RBI he posted the past two years, but that was largely due to a weaker supporting cast. The experts generally expect a 2019 that is equivalent or better than his 2018, and his projections say the same. The most pessimistic projection calls for a .286/.357/.548 line with 36 HR, but that’s from Steamer, which tends to be among the most conservative. Despite the lack of stolen bases, his consistently elite power bat has owners penciling him in for first-round production and nobody is worried about Arenado. Nobody is ever worried about Arenado. One could argue that this his defining trait.
Then again, this is exactly the kind of player where a hidden weakness or chink in the armor would go unnoticed. And it seems I may have found one, and a rather big one at that.
Arenado’s xStats Anomaly
When you look beyond his career of excellence and beyond the surface stats and focus solely on his 2018 peripherals, something troubling emerges. His triple-slash of .297/.374/.561 with 38 HR is elite. However, his expected triple-slash (xSlash) of .275/.354/.480 with 26.8 xHR is merely above average. The last two numbers are of the most concern. Batting average could rise or fall 10 or 20 points and if the power remains the same, few would complain. But the difference between his actual and expected slugging is over .080%, and the difference in homers more than 11. A player who hits .275 with 27 Home Runs (I’m rounding up to be generous) is still a perfectly healthy line that any team would take, but not in the first, second or even third round in fantasy baseball. Among third baseman, the most similar offensive production to Arenado’s 2018 xSlash of .275/.354/.480 with 27 xHR is Matt Chapman. Also an elite defender, Chapman hit .278/.356/.508 with 24 HR in 2018. I happen to love Matt Chapman as a real life and fantasy player, but he is a player with a 2019 NFBC ADP of 107. Arenado’s ADP is 9.
Considering how impossible this seems, there must be some sort of explanation. Perhaps this hidden downward lurch was nothing more than a wrench in the gears in the xStats algorithm, since no projection system is infallible. I still remember back in 2012 when there was media buzz that scientists may have discovered a nanoparticle faster than the speed of light, which would obliterate Einstein’s theory of relativity. But then scientists at CERN learned it was only due to an error in the test results caused by a loose cable. The math world didn’t divide by zero because the scientists first assumption was that they screwed up somehow, and tested every possible screw-up first. Likewise, I will first assume the null hypothesis of “System Report: Everything is Fine. Nothing is Ruined.” Granted, that “System Report” was just a Post-it Note Homestar Runner slapped on the screen of Strong Bad’s computer that was definitely ruined, so let’s do our due diligence here.
Let’s investigate first whether the xStats algorithms have always been pessimistic on Arenado.
|Nolan Arenado Year||AVG||xAVG||SLG%||xSLG%||HR||xHR|
In every other year, Arenado had outperformed the expectations by a small, reasonable amount, except for SLG% vs xSLG% where the discrepancy is indeed larger. Despite the gap in SLG%, his amount of xHRs has been remarkably consistent since 2015: 37, 37, 37… and then 27. Even if we account for previous year’s discrepancies between his actual numbers and his xStats expected numbers, his 2018 was decidedly worse. If we attempt to nullify any previous xStats bias (by taking the mean of the 2015-2017 actual vs. expected discrepancies in each category and adding them to the 2018 xSlash), then the adjusted 2018 xSlash would come out to .283/.360/.521 with 29.5 HR. So while it’s fair to say that Arenado has historically been rather underrated by xStats, this still does not account for the massive dropoff in expected power production that occurred in 2018.
But perhaps there was something specific to the 2018 season that affected all Rockies hitters in this way. This could always be the case if there were an error in accounting for a year-to-year change in park factors. So let’s look at his top teammates who played a full season in 2018 and see how they stacked up.
|Player (2018 Season)||AVG||xAVG||SLG%||xSLG%||HR||xHR|
We wouldn’t be at all wrong to conclude that xStats didn’t give Rockies hitters enough credit for their home park, especially where xSLG% and xHR are concerned. However, the system wasn’t nearly as pessimistic on any of these players in 2018 as it was on Arenado. Trevor Story slugged the same as Arenado in 2018, but xStats saw Story’s hitting as far more legit, even giving him a higher xAVG than Arenado. Trevor Story! Going by their respective xSlash lines, Arenado was as about valuable as a Charlie Blackmon who never runs.
Still, maybe xStats is the only advanced indicator that had a personal vendetta against Arenado in 2018. Let’s look at Statcast, which has their own system of expected stats. I generally prefer xStats, as it takes more variables into account than Statcast’s version, but the other information Statcast provides should still help strengthen or dispel the theory that evil is afoot. Arenado’s Statcast 2018 XAVG is .268 and XSLG is .471, though I view those numbers with more skepticism as Statcast does seem to unfairly underrate him. Statcast’s system gave Arenado a career .271 XBA and .492 XSLG, which leads me to believe Statcast doesn’t factor in his bandbox of a home field. So it’s reasonable to toss these numbers as mostly irrelevant.
However, Statcast did note that in 2018, he had some other more legitimate signs of a power decline. Arenado only had 35 Barrels and a 5.2% Barrel/PA in 2018, which falls short of his 2015-2017 rates, which averaged 45 Barrels and a 6.6% Barrel/PA every year. Relative to the rest of the league, his barreling ability had actually already been inching downward, as his Barrel/PA among qualified players (150 BBE) from 2015-2017 was #39, #61, #75, respectively. Then in 2018, it plummeted to #135. Based on his 2018 launch angles, exit velocities and barrel rates, the player Statcast describes as most similar is Asdrubal Cabrera.
Here I should point out that, even in his best seasons, Arenado has never been a Statcast darling, the tongue-in-cheek nickname given to players with gaudy exit velocities and barrel rates that indicate the ability to hit homers far and hit them often. Rather, he has taken a different path to success. Behold the Arenado Five-Fold Path to Fantasy Enlightenment:
- Stay healthy all season.
- Hit an above-average rate of hard contact and exit velocity.
- Hit a very high percentage of fly balls.
- Play in an extreme hitter-friendly park.
- Maintain a low strikeout rate (14-17%).
For Your Health
The good news is that Arenado was healthy enough to amass 673 plate appearances, the 4th consecutive year in which he’s logged 665 or more, which is quite rare nowadays. In April, he played through some knee soreness and was fine, and in August, he tweaked his shoulder but was back in the lineup in a few days. It doesn’t seem that health was an issue, but that also takes away one more excuse for what caused his drops in contact and fly-ball rate.
Another sign of his health is that his exit velocity itself was A-OK. His overall eV of 90 mph and FB/LD is even slightly better than 2016 and 2017, and his FB/LD eV of 94 mph is the same as his 2015 rate. At first glance, it would even appear he even had a hard contact breakthrough at a career-high 43%, though much of these gains are nullified by the 4% hard contact increase across the league from 2017 to 2018. Long story short, no problems in Five-Fold Path departments #1 or #2.
Nolan, Your Fly is Down
Which brings us to #3, the main answer to why Arenado’s barrel rate likely declined- He just didn’t hit enough fly balls. After averaging 45% fly balls from 2015-2017, which consistently ranked in the top-20 among all qualified hitters, in 2018 his fly rate fell just short of 39%, ranking 44th. Here is where we arrive at a missing puzzle piece, why Arenado still looked like a Mustang despite the Honda Fit engine. In 2018, Arenado’s HR/FB was a career high 20.7%, over 3.5 points higher than his 2015-2017 average HR/FB of 17.2%. Beep beep! Yeah that’s right, I put my Honda Fit clown horn in the Mustang too.
No, a HR/FB over 20% is not automatically unsustainable. Christian Yelich had a 35% HR/FB last year… now that one is unsustainable. While Arenado’s 20.7% mark is 20th-highest on HR/FB and he does play half his games in Coors, most of the players above him on the HR/FB leaderboard produced markedly better Barrel/PA rates and FB/LD exit velocities. Even though Arenado had a career-best FB/LD eV of 94.4 mph compared to the league, that still only ranked 74th.
Arenado’s fly ball game has always been more about quantity than quality, and the quality is unchanged but the quantity is down. Based on his career rates, his HR/FB should regress back to the mean, and if he can’t bump his fly ball rate back up to his typically extreme levels, his home run total will likely be closer to 30 than 40. If you’re skeptical of this, just take his average 2015-2017 HR/FB and multiply it by his 2018 total fly balls… that’s 31.6 home runs. That may not seem like a big enough deal to write a somewhat long-winded article about it, but when power is the attribute that makes him a no-doubt first rounder and recipient of the richest arbitration salary in history, it does matter.
As I mentioned earlier, this is Arenado’s final season under club control, and he and the Rockies are likely to headed to arbitration. While some trade rumors have swirled about, some including the Yankees, Rockies GM Jeff Bridich has kept mum. He says the club is content to “keep things under wraps, under our hat for now” and they will not “set any deadlines or ultimatums,” as they attempt to work out a 2019 salary, and maybe even a long-term deal, with their star slugger. While the odds are long that he gets traded, it’s still worth considering the possibility, as his getting traded to virtually any other team could have a substantial negative impact on his fantasy value.
Although the Humidor and raised walls in 2016 curtailed the video game numbers of previous years, the thin Colorado air still led to a HR Park Factor of 1.28 (2nd highest) and Runs Park Factor of 1.271 (2nd highest). It’s no surprise that Colorado players especially do better at home than on the road, but Arenado’s home/road split is especially extreme. In 330 PA at home, he hit .347 with 23 HR, while in 343 PA on the road, he hit just .248 with 15 HR. The Rockies team batting average is .287 with 119 HR at home and .225 with 91 HR on the road, but Arenado’s Home/Road splits for both categories are nearly double that discrepancy. It’s probably in everyone’s best interests to keep him Arenado in that park, as he and Coors are a perfect match. “Arenado” even rhymes with “Colorado”.
Still, it’s worth noting that the run production decline was largely due to a weaker overall offense, and the Rockies front office has yet to do much about it than to sign Daniel Murphy after letting DJ LeMahieu walk. From 2015-2017, the Rockies offense averaged .271/.330/.444 with a .331 wOBA, but the offense slashed just .256/.322/.435 with a .325 wOBA in 2018. While the GM’s passivity could be an omen of a mid-season trade for Arenado, it also hints that even if Arenado were to sustain his present real-life production, the 130 RBI seasons of yore may forever be a thing of the past.
It’s not uncommon for players to have skill changes in their career, and it would be forgivable if Arenado had been sacrificing some power for improved contact, so let’s see if he’s made improvements there.
This is the equivalent of receiving a consolation gift box that contains thousands of live spiders… assuming you’re not an entomologist. The most concerning decline is that of his Z-Contact%, which was has slightly declined every year but the rate of decline has only accelerated. In 2015, Arenado’s 90.9% Z-contact% was excellent. Francisco Lindor had a 90.9% Z-contact% this year. But the 84.4% Z-contact% Arenado posted in 2018 is actually slightly below the 2018 league average of 85.5%. That mark is tied with 2018 Edwin Encarnacion, and of those two, Arenado had the higher whiff rate rate (SwStr%), since he also had a slightly below average chase rate (33% O-Swing%).While Arenado’s overall contact rate decline has been less drastic, that’s far from comforting. After all, contact rate is a predictor of strikeout rate, and a player with a below 80% contact rate is generally expected to strike out over 20% of the time. I’d be much less concerned if this decline in overall contact rate was coming from whiffing more on pitches off the plate, but he’s missing more on strikes, where it’s easier to make high-quality contact. Perhaps he could compensate for this if he had improved his chase rate, but that actually got worse, making his career-high walk rate also seem unsustainable.
One may be tempted to save him with a hail mary pass by bringing up Nick Castellanos, a rather extreme example of a player who has maintained a high batting average despite declining plate skills. Usually invoking his name brings up the argument that some players just have that X-factor; that knack for getting hits that these other stats can’t see. Fortunately, we have BACON (Batting Average on CONtact). In 2018, Castellanos posted an elite BACON of .392 supported by an xBACON of .393. Arenado’s 2018 BACON was a solid .369, but his xBACON was actually slightly below league average at .342, and also right in line with his career xBACON of .345. From 2016 to 2018, he’s maintained a steady Value Hit rate (VH%) of 10%, which is above average, but with a poor hit rate (PH%) of 30%, which is below average. For comparison’s sake, Matt Chapman’s 2018 VH% was 12% PH% was 24%. So as far as Arenado’s contact is concerned, the quality was unchanged while the quantity went down. Sound familiar?
What to Expect Going Forward
Perhaps I’m being baseball’s Chicken Little and saying the sky is falling when it’s really just a few concrete crumbles from the roof of the Superdome. Still, if you noticed the fallen chunks, you might feel a bit more uneasy about the dome’s structural integrity when the crowd reaches a fever pitch, and maybe you wouldn’t add to the noise. Nobody else has seemed to notice the fallen chunks in the Nolan Arenadome. How about that: his name even contains the word “arena.”
If I were the Rockies GM, I’d keep these concerns in the back of my mind as his camp confidently explains to the arbitrator why he should be a $30 million man in 2019 before netting a lucrative extension or signing a mega-contract in free agency. With his NFBC Min Pick of 4 and a Max Pick of 15, I acknowledge that I am one vote for Nay in an ocean of Ayes, but I will be avoiding Arenado in the first round and probably even the second round, and opting for a player with less downside. After all, I can wait several rounds later and get Matt Chapman.
I’m not saying that Nolan Arenado is bad at baseball, or even that he’s not a star and future Hall of Fame candidate. I project him to hit .280-.285 and hit 30-33 Home Runs and about 200 combined Runs + RBI. That would still quite valuable, just not for what he will likely cost. But for all I know, being the superlative talent that he is, maybe Arenado is already well aware of the good fortune masking his recent shortcomings. Maybe he finds a way to bring back his old fly ball rate or his old contact rate, or both. Especially if his struggles were due to playing through a hidden injury that he’s now recovered from, that’s an entirely plausible outcome. But what I am saying is that if Arenado continues his recent fly ball and contact rate trends into 2019, I think his batted-ball luck bubble may burst, and he will burn teams that are counting on first-round production. And right now, nobody is talking about there being a bubble at all.
Really good stuff.. I’m quite certain his legion of fantasy owners will hop to his defense because you factually lowered his trade / draft value.
Thanks! Reddit has been happy to confirm your hypothesis.
Now I’m all scared of Nolan!!! Thanks a lot!!!! Haha! Good stuff Ben.
Haha thanks Harley!
This is hilarious.
Thank you, I do always try my best to add some humor to balance out the nerdy statistical arguments.
Dear Ben, great article, love your work at PL.
My league’s batting cats are 8: R, HR, TB, RBIs, SBs, AVG, BBs and Ks. I have the 3rd overall pick and was thinking Arenado. After this, i might take JoRam, but there is also this new info on JoRam struggling against breaking pitches. Are you concerned as well on JoRam or would you easily take him 3rd?
Thanks a lot man!
Thanks so much for the kind words! I’d say that especially with your league settings, I would not go Arenado at #3, and not even close (considering my argument that he’s not even a first rounder, this probably isn’t surprising). But especially in a league that weighs BB and Ks as two separate categories, note that Arenado is worse than most other first rounders in both.
White the hot thing now seems to trash JoRam, I still think he’s solid for #3, especially in your league with his 15.2% BB% and 11.5% K%. Pitchers did throw him fewer fastballs, but they still threw them 49% even knowing he clobbers them, because it’s hard for most pitchers to stay away from their fastball. He’ll regress either way, but worst case he’s still probably going 27/24 with .270 AVG and great BB and K numbers.
But if you want a power bat, I’d say J.D. Martinez is a much better pick than Nolan. Unlike Arenado, all of JD’s power (and associates’) numbers are backed by Statcast/xStats, better lineup, and he even gets a handful of SBs.
After riding Nolan to the playoffs for a 4th consecutive season in my dynasty league in 2018, I traded him this offseason for 3 first round picks (with which I drafted Peter Alonso, Alex Kirilloff and Garrett Hampson — I’m pretty happy about it).
I made the trade because I too am concerned about Nolan — but not as much for the reasons you’ve given above. You briefly mentioned Nolan’s shoulder in the article, and I think that could explain more about his (relative) struggles than you give credit for. There are a number of articles from last August and September that mention Nolan was playing through pain, which is not surprising, as he is a such a fierce competitor. (I don’t have all of the fancy advanced sabermetrics at my finger tips just now, but a quick look at basic splits shows a .312/.395/.586 first half vs. a .275/.343/.526 second half).
My bigger worry about Nolan is his contract situation. Between his glove, his bat, his age, and his competitiveness and leadership, Nolan is going to command $300m next offseason.
Are the Rockies going to pay that? Between Trevor Story, Brendan Rodgers, Garrett Hampson, Ryan McMahon (he’ll come around), Josh Fuentes, Colton Welker, Tyler Nevin and Grant Lavigne (oh and Daniel Murphy and other potential FA on the market) the Rockies infield is looking pretty loaded in the future. (Any of those first SEVEN guys could play 3B defensively).
Are the Rockies going to give Nolan $300m with all that young infield depth? If they don’t — I couldn’t bear the thought of holding the bag on a Coors-less Nolan Arenado. Throughout his career, Nolan slashes .320/.374/.609 at home (1st ballot HoF numbers) — but on the road, Nolan only manages .269/.318/.469 (not even close to top 100 fantasy player, especially without any SB).
Then again, the Rockies may just pay the man. Or, given those road splits, the market might not be what Nolan expects it to be, which would give the Rockies a significant leg up in resigning him. That said, as I mentioned in my very first sentence, in a dynasty context, I’m not sticking around long enough to find out. I’d rather take the cash and run.
any of those first SIX* guys could play 3B.
I’ve heard a lot about this injury hypothesis, so I investigated. He sustained his injury August 10th. Of course, players play through minor injuries all the time, and not all impact performance. Signs of being injury negatively affecting his game play would be significant changes in GB%, K%, or Hard Contact%. In August and September, he had a GB% of 42%, 18% K%, and 40.5% Hard%. These are all right in line with his full season rates of 40% GB, 18% K%, and 42.9% Hard%, So even though he had 3 HR in August, that seems to be just bad luck, as his 11% HR/FB that month was far below his season HR/FB of 21%.
Also, his September did not in any way contribute to the poor expected stats. In September he hit .265/.330/.569 with 6 HR, compared to his xSlash of .283/.346/.530 with 5 xHR. So in September he actually seemed completely healthy, since the xStats actually thought he deserved better offensive numbers in September than his overall rate stats for the 2018 season.
Therefore it seems that the 2nd half dropoff in production wasn’t due to the nagging injuries, but an exceptionally lucky in the first half that fell back to earth. Unless he dealt with a nagging injury all season (I haven’t seen evidence to support that), it just seems like a player who peaked with an elite 2016 who has been declining from a skills perspective since.
He’ll still be a very good player in Colorado, especially with his elite defense and team leadership skills, but those don’t count in fantasy and offensively he may just be very good but not elite. Since other teams value him as a 1st rounder since they don’t see the decline potential, it’s smart of you to cash out now.
Oh man, his skills really are just declining. Glad I got out when I did.