Every fanbase suffers through cursory glances from talking heads and fans of other teams that don't know what's going on. The Nationals primary cursory glance is the expected "Bryce Harper is gone! Oh no! Panic!" But the Nationals also have a secondary cursory glance, probably in part because the first one is so prevalent, that goes "They've improved their defense and rotation so much!" This is because they lost Bryce and Murphy and added Corbin.
But are these things true? Today we'll talk about defense. Defense stats are funny things. No matter how often we are told that it takes at least a couple years of defensive data to really fairly portray the skill, we are still given annual data. Because of that we can't help but try to interpret data in that manner and if something happens that is out of sorts with the general knowledge we try to fill that in. Sometimes it makes sense - an injury or a rapidly aging player showing much worse defense. But sometimes there isn't rhyme or reason as we try to fill in the gaps. In the Nats case you had Bryce become one of the worst outfielders in baseball while Trea Turner became an elite defender. While I didn't see any obvious signs of such drastic change, I have heard from other fans "oh yes, Bryce was terrible. Trea really worked on things". Maybe that's the case but from here on out in this analysis we're working with both the 3 year averages and last year's numbers to present what is probably a fairer picture of what's going on.
The Nats did do poorly on defense but not terribly. They were a below average team, heading up the bottom third of defense. Fangraphs' D stat, UZR, Total Zone all think this. Their defensive efficiency was pretty good though an artifact of having the fewest errors in the National League. In short - they fielded well when they got to the ball, but they didn't get to the ball often enough.
The infield should be pretty similar to last year. Rendon, at third, is a perennial gold glove candidate, (elite 3yr ave, top 5 last year). Turner, at short (blah 3yr, Top 10 last year) still is a question. Is he average or good for the position? Regardless he's not a defensive liability. Zimmerman, at first (terrible, terrible), with his injury limited mobility and dead arm is possibly the worst regular first baseman in the game. What's changing is 2nd base. Murphy (terrible, terrible) is admittedly a terrible defender but most of the innings last year were actually played by Difo (Top 10, Top 10) who is a very good defender. All in all 2nd wasn't a hole for most of the season and would grade out pretty average. Dozier (below average, terrible), was a pretty solid defender in his youth but has taken a sharp downturn as he has aged. Kendrick is a hard read. He's been good in the past but has spent most of the last 3 years in the OF. Last year he did play a little more 2B early on and the stats aren't good but now we're talking a piece of a piece. It's likely he's still ok, but after three years of aging and lack of regular practice at the position, I have a hard time believing he's a plus. The end results in the IF is likely no improvement taking the season as a whole.
The outfield is a little more complicated because it was a mix that is now going to have a new mix with some of the same pieces.
Bryce (below average, terrible) after showing the bare minimum at corner OF, played a third of last season in CF and proved what we all thought - he can't play CF. Most of the year at center was taken up by MAT (Top 5, Top 10) who is an elite fielder though not in the conversation for best in the game in a competitive position. Juan Soto (?, terrible) in the corner was legit awful but one season of stats and his young age leaves us with a big question mark for actual skill. Adam Eaton (above average, bad) started his career as a mediocre CF but when he started to falter he got pushed to the corner and shone for a year. However the last couple of years have not been kind and show Eaton as not a good fielder. This year the Nats will lose Bryce (good!) and MAT (bad!) to get more Eaton (probably bad!) and more Soto (also probably bad!). They also lost more of those other guys, who except for Stevenson weren't any good.
There's one name I haven't mentioned yet and he's frankly the key here because everything I said above? It's not good! The question is what is Robles as a fielder. If he is elite, the OF defense likely improves. If he is MAT level Top 5/10 ish, the defense likely stays around the same. If he is not that good, the OF defense actually likely gets worse. The good news is the preliminary stats agree with the scouting reports that suggest Robles is the MAT or better fielder we've been thinking he is. Still it's questionable data so hold off getting excited but it's better than the alternative.
We haven't talked catcher yet. The Nats split time between Wieters (Top 10, top half), Pedro Severino (?, Top 10) and Kieboom (?, Top 10), accepting those Severino and Kieboom numbers are very tentative (though Severino seems to consistently grade well in limited exposure). Next year it'll be between Yan Gomes (Top 10, Top 10) and Kurt Suzuki (average, top half). We also have to at least look at framing. Wieters is a pretty lousy framer, Severino a pretty good one, Kieboom we lack enough data for to say. Gomes has been pretty average but had a good season last year. Suzuki has been pretty bad and was one of the worst last year. If the catcher situation ends up being Gomes, with Suzuki as a back-up I'd say you can expect slighty better catcher defense. If we see more of a 50/50 split the Nats are probably actually taking a little step back.
So while the gut says - "No lackadaisical, out of position Bryce, No flat footed, injured Murphy - Nats defense must get better!" the analysis doesn't agree. Second base had a lot of Difo last year who was really good, and neither Dozier or Kendrick will be any better than average. Corner OF will still likely be an issue. Catcher, despite the memories of NLDS Wieters, wasn't actually bad and won't likely improve.
I suppose you can figure out a way this cursory thought ends up being true. Turner is that good again. Soto gets better. Robles is one of the best CF of recent times. That would do it. But are we really going to say that's the likely scenario?
This doesn't mean the Nats aren't better but are the Nats better because their defense is much improved? No.
In the first part of your blog you spoke about injuries as one of the reasons for a sudden drop off. If 2-3 years of above average data hits a major injury (Eaton knee/ankle, or Dozier playing through a knee injury and playing out of position and platooning in LA) for someone that is still under 32 could you expect some increase once they are back to 100% not saying these two players are but i think you can definitely expect some improvement from Eaton playing RF above non diving bryce with a knee and ankle that went through a full winter of strength training rathe than rehab
ReplyDeleteAs always, a good analysis that brings my expectations down to Earth (reality)! I found myself really glomming on to the "improved defense" narrative, but you're right, with further scrutiny it doesn't really hold up.
ReplyDeleteThe question I have for you, Harper, is:
In terms of Dozier replacing Difo at 2nd - I think we can safely say that Difo is a superior fielder, but will Dozier's bat be enough of an improvement over Difo to overcome their differences in fielding, or will the swap between these two players be negligible regarding their overall contributions to the team?
I am confused about how top 5 isn't elite? MAT can't hit but he sure can field. Robles (also elite by all reports) replacing MAT should be the same fielding with an average bat instead of a dreadful bat.
ReplyDeleteRoark - You may be right but I couldn't find an example of someone becoming BAD and getting back to good/average because of injury. I think this jives with what I know of the game - you don't come back and play the field unless you are ready to or absolutely must and neither Eaton or Dozier is an absolute have to guy. Rendon, who is closer, killed his knee and had an off-year on D coming back but was still very good, just not top of his game. I think Dozier and Eaton could be better then they were but I think at best they'll struggle to be close to average.
ReplyDeleteAnon - I would think the latter. I'm far more optimistic about offensive return after injury and even though I don't think we'll get peak Minnesota Dozier, I think his bat will be good enough to make a noticeable difference in comparison with no-hit Difo
cass - The top couple guys at most positions tend to be above beyond even the guys at 4-5, that's why I make a distinction. I think the Nats will improve on the non-pitching side but because of bats and depth mainly.
I think folks are way off assuming the cake is baked on Soto's defense.
ReplyDeleteYes, most players' defense monotonically declines from their first MLB game, but most debuts aren't at 19, and at a position where they've previously played 7(!) professional games. Hell, he only played 117 games in the minors total, which pretty much makes him the rawest rookie in recent memory.
And most players aren't so unpolished as amateurs that their pro coaches are able to meaningfully increase their sprint speed. Soto was scouted with below average speed and last year he was slightly above average (27.2 vs 27.0 mph). (I'm not sure if this was a mechanical change or a series of exercises - old reports seemed to talk about it as mechanical, but Soto answered a question on it this spring and said something like "Yes, I keep working on it. If I don't, I'll be slow again." Which seems to argue it's more a core / fast twitch workout type thing. But it's real and acknowledged by both the team and the player.)
I mean, obviously, we won't know for sure until he actually does improve. And it's perfectly possible that he and/or the coaches decide that the defense is just fine and that the focus should be squeezing all possible value out of the bat. And it's definitely possible that he's never a good defender; he has terrible instincts and maybe he can't improve.
But I do think that he is in rather unique situation where it's reasonable to assign a decent probability to him improving his defense. Especially because his offensive approach argues that he's a cerebral player who can self-critique and adjust accordingly.
I would expect Soto to be better but still bad, Eaton to back to average to slightly above. Dozier worries me as I don't completely buy the injury excuse I view it the same way as I did the Harper "injury" that was his substandard 2016. Zim is a black hole. Our catching situation is worlds better. I'm convinced Trea is good, and your unfeeling cyborg analysis won't sway my human feelings on that.
ReplyDeleteAlso boo you Harper for not addressing my Kimbrel question in the last post. I'll never forgive you.
ReplyDelete@Harper. I think you’re missing something here that I believe is happening and important, which is the possible beginning of Rendon’s defensive decline from “as good as anybody not named Arenado/Machado” to “slightly above average.” Last year Rendon for the first time did straight up poorly in multiple advanced metrics, particularly DRS, where he was negative. Single years can be iffy, due to positioning and various issues, but from the eye test, he did seem like he had less phenomenal range last year, albeit with his usual great hands. We’ll see, but this year will tell us whether the era of “Anthony Rendon, elite defender” is coming to a close. (Some other metrics last year rated him as still above average, but none had him as still elite).
ReplyDeleteRe Turner, I’m fairly positive he’s a great defender, not a good one...he’s just the type who is incredibly quick and gets to spots but without the flashy plays you associate with being a great defender (many infielders make tons of flashy plays and it’s because they barely get to balls they should easily get to....this was Brandon Phillips every year basically).
Re Robles and Soto...yeah, their sample sizes are small, but when the sample size results exactly line up with the universally agreed upon scouting report, I tend to give them credence. I think Robles is an absolute beast out there and Soto is a bad defender—-whether he’s utterly terrible or just bad we’ll see. But I think with Robles, I’m comfortable with saying we’re talking a floor of above average and a ceiling of best defensive CF in baseball. Scouts lose their mind talking about his range, and it makes sense given that he is (barely) faster even than Turner (by statcast.)
Re Zim. He’s just a disaster. The guy could have another 2017 where he hits 35 homers and .300 and he’d still be worth like 2.5 WAR.
But roster wise....we look like we will have an above average offense, above average rotation, maybe average-ish defense, above average bench, and our very obvious team weakness at the moment is the bullpen. Sign Kimbrel.
Wieters wasn't THAT bad after all. Apparently, it's his "framing" that's terrible.
ReplyDeleteWhat I'd like to know is whether "framing" is really a quantifiable statistic. (Sounds more like carpentry, like marquetry or architrave.)
Ha
DeleteWieters was bad. The advance metrics can go to hell on that if they rate him good. He couldn't block balls in the dirt.
ReplyDelete@SM Framing is a quantifiable stat. Here is a relevant link:
ReplyDeletehttp://www.statcorner.com/CatcherReport.php
Thanks, G Cracka X. I've seen it, and Mike Fast's piece in Baseball Prospectus, too.
ReplyDeleteI don't disagree that there is a numeric value attached to framing, but still . . . .
@SM How they arrive at the framing numbers isn’t that complex. There is an established strike zone, and we are able to know when a ball outside the zone is called a strike (call it plus 1 for the catcher) or a ball in the zone is called a ball (call it minus 1). And there are some catchers who consistently have high scores and others who consistently have poor scores, I.e. the latter not only don’t “steal strikes,” but they also “lose” strikes. (And it’s not like the umpires are the ones who make a difference for those scores...mostly umpires are either good umpires (they’re accurate) or they’re bad (inaccurate, that is, they miss ball AND strike calls)....plus every catcher in the league deals with same umps. Anyway, it’s fairly straightforward.
DeleteFirst, defensive metrics are tricky and not nearly as reliable as offensive metrics, but I don't like UZR and while DRS used to be the best, Inside Edge, I believe, is the best, but even then it's limited in that it's skewed to number of chances. So, kind of curious which stat or stats you were using to make your assessment? For example, UZR and DRS rely too heavily on the game situation. So a good play made with 2 outs and runners on second and third is more valuable than a good play made to start an inning. If you value "clutch" then it's a great stat you can use to calculate overall grit and gamerness.
ReplyDeleteSecond, I agree with @Gabe Roark, I wouldn't put much stock in either Eaton or Dozier's defensive stats from last year because both were hurt. Now, that could be a sign of things to come, but it also could not.
Finally, there's a rhythm to defense and constantly having a new double play partner or outfield configuration would seem that it would hurt the overall defense, but how much is hard to quantify. Lot of words to say that I think this defense is going to be much improved, especially with Wieters gone.
Looks like you are probably right about Howie defensively and injury-wise, he left the ST game today with a left hamstring strain...
ReplyDeleteJW - I used UZR, fangraphs DEF measure (which itself uses UZR in part), and Total Zone FRAA to get a general consensus point. I agree with the whole "mushiness" of defense statistics so I'm not going to argue with you if you want to shift a player's assessment around a bit, as long as it's not crazy. Like MAT could be best in league or close to average, but he's not bad.
ReplyDeleteMike Rizzo told reporters today that the Nats will not be going over the luxury tax threshold this year. That is their "goal and strategy." I'm not sure that is actually a strategy but no Kimbrel for sure. You can argue that there are some great free agents next year, so go over the luxury tax next year. Anyway, it's safe to say there will be no additions to the Nats roster.
ReplyDelete@Jay well, not necessarily, if they can make a move and counter move to stay under. Or make a small move. For example maybe they decide to sign Gio for 6m on one year deal.
ReplyDeleteTrue, they would have to extend Zim somehow and lower his AAV but I am not sure they can do that enough to sign Kimbrel. I'm not sure there is anyone else they could move - maybe trade Barraclough but can they even do that since they just traded for him? I'm guessing the short answer is no.
ReplyDelete@Jay yeah I’m guessing what happened is they really were talking to kimbrel and were trying to convince ownership to go over the cap and they definitively said no this week or something.
ReplyDeleteSeverino was a nightmare behind the plate
ReplyDeleteHarper - have an idea for a blog post in the future. Which stat would deem the most important in determining the success of a team. If too broad, perhaps weighing offensive/defensive/pitching/baserunning value and which stats are the most indicative of each facet of the game in terms of team success. For example, a lot of people don't like batting average, so which would be the best measure for team success at the plate? Could be cool to explore as we've seen the Nats put an premium on various traits/skills since 2012.
ReplyDeleteThey've always had (or at least seem to value) a strong front-end of the rotation and have historically been negligent on the bullpen. They seem to be moving away from power bats and putting an emphasis on OBP guys. I'm aware that there is no singular answer to this question, but it might be useful to see how these strategies could be/have been employed for our team.
Here's another idea for the suggestion box.
ReplyDeleteRather than trade Bryce, as last season wound down/fell apart and get someone in return -- say, a reliever -- Rizzo decided to wait for the Phillies to send them a compensation first rounder. Does that signify that draft picks are going up in value relative to middling veterans?
2019 Soto > 2018 Soto/Other (He was 19 with 7 games experience... geez)
ReplyDelete2019 Robles/MAT => 2018 MAT/HARPER/Other
2019 Eaton >> 2018 Bryce/Eaton/Other
2019 Rendon = 2018 Rendon
2019 Turner = 2018 Turner
2019 Dozier => 2018 Defo/Murphy
2019 Zim/Adams = 2018 Zim/Adams
2019 Gomes/Suziki => 2018 Weiters/Severino
There was some smoke and mirrors in that analysis... put the 2018 player(s) head to head with the 2019(s) player. I'm convinced Eaton and Dozier will be better this year than last because of injury recovery.
As always, 2019 injuries will be a huge factor
Mark, maybe you're right. But how did you arrive at your prediction of 2019 Eaton > 2018 Eaton + Harper. That is not a prediction I've seen anyone else make.
ReplyDelete