Where Did the Nationals Go Wrong?
The Nationals’ season isn’t quite on life support, but there’s been a bad accident, and the family’s been notified. The situation appears to be dire, and while Tuesday brought its own fresh horrors, that wasn’t the day that slaughtered the dream. It’s the most recent event, the currently most upsetting event, but the Nationals weren’t even supposed to be in this situation in the first place. They’ve had problems for weeks, for months, and now they’re about out of time to save themselves and move on to the tournament. The Mets have simply done too much. The Nationals have simply done too little.
The story would be interesting if it were any division race. What makes this one extra interesting is that it’s this division race. The Nationals were expected to run away with the NL East. The Phillies, as assumed, have been bad. The Braves, as assumed, have been bad. The Marlins had the look of being mediocre. The Mets had the look of a fringe contender. The Nationals had the look of a champion. These are our preseason odds. The Nationals were given an 86% chance to win the division. They were projected to clear the Mets by 13 games. If the Nationals won out, and the Mets lost out, the Nationals would clear the Mets by 18 games. This was supposed to be a relative cakewalk. It’s been more of a…nailwalk? I don’t know. It’s been bad, is the point.
So: why? How has this happened? Obviously, the Mets have quite a bit to do with it, but this post is to focus on the Nationals. Where did this go wrong?
You’ll observe that this post has a lot in common with this other post, about the also-disappointing Mariners from the end of July. I like the format well enough, as I think it works to capture the good and bad surprises. Getting right to it, the Nationals are 71-67. Plugging in the preseason projection, we thought at this point the Nationals would be 81-57. There’s a 10-game difference right there, which is clearly enormous. How could we explain half of that? It’s interesting to look at the BaseRuns standings. Based on those estimates, both the Nationals and the Mets should be 76-62. They definitely are not, with the Mets one game better, and the Nationals five games worse. That’s the whole difference between first and second place, coming just out of sequencing. Some of that, I’m sure, people are more than happy to pin on Matt Williams. He’s presumably not all of it. Whether it’s just bad luck or bad timing, the Nationals have sort of gotten it in the shorts, and that’s a reason why their hopes are fading.
But even 76-62 would be short of the preseason expectation. Remember, the Nationals weren’t supposed to be about as good as the Mets; they were supposed to be a lot better than the Mets. Where can we find those other five missing wins? The answers are in the graph below.
What we have right now are individual player performances for 2015. What we had months ago were projected individual player performances for 2015. So this makes for an easy comparison, taking the projections and putting them over 138 games instead of 162. Then we can compare actual and projected performances, and see who’s over-achieved, and who’s under-achieved. My statistic of choice is WAR, and while, yeah, I know, flaws and everything, WAR’s the best we’ve got, and nothing else is close. This graph contains most of what’s important. It’s built from preseason depth charts, preseason projections, and actual performances.
Data going into that graph: 39.3 projected WAR, and 32.8 actual WAR, for a difference of -6.5 wins. Unseen are some surprise contributors, like Joe Ross. Players who didn’t appear in the preseason depth chart. Those players have combined for 1.6 WAR, so then that leaves us at -4.9, or, better, 5. Here are your five missing wins, as we’ve been looking for. I don’t think anything should really come as a shock, but this lists it out player by player.
The thing that really stands out to me is Bryce Harper. I think we’ve all talked about this, but one of the craziest things about the Nationals’ disappointing season is it’s also a season where Harper has had one of the better performances in recent history. He’s broken through, achieving the Mike Trout tier, resembling one of the most complete players in the game, and still the team has struggled around him. It would’ve made sense to say that, if the Nationals were going to disappoint, Harper would probably be hurt. Instead, he’s beat his projection by 4.5 wins. Put another way, he’s more than doubled it. Think about what that means for the supporting cast.
Danny Espinosa‘s done more than expected. Yunel Escobar, too. Michael Taylor has awesome defensive numbers. But this graph quickly turns sour. Or, bitter, if you like sour. There are the littler disappointments, like the injured Denard Span. There are the quieter disappointments, like the underwhelming Tyler Moore. Then there are the star-level disappointments. The Nationals were supposed to have an excellent core. Much of that core has under-achieved. Obviously, some of that is directly due to injury. Some more of that is due to lingering effects of injuries long since healed. But, Jayson Werth has been either hurt or bad. Anthony Rendon has been either hurt or average. Stephen Strasburg’s been a little bipolar. Doug Fister came apart. And on, and on.
For the sake of example: Harper has beat his projection, so far, by 4.5 wins. That’s a massive, massive gift. Yet Harper’s breakthrough has been more than negated by Werth and Rendon. Combined, they’re at -5.5 wins, and though some of the time they missed went to occasionally productive fill-ins, those fill-ins don’t have the skills to match what a healthy Werth or Rendon could do. Fister has completely negated the Espinosa surprise. Even just Tyler Moore cancels out Joe Ross emerging from the shadows.
Nothing is really that much of a surprise, but the embedded image shows how a team can blow one of the most impressive individual breakthroughs we’ve ever seen. It’s hard to blame the Nationals for the bad injury luck they’ve experienced, and in that sense it’s been a little out of their hands, but it’s not all injuries. And it’s not like the Mets haven’t had injuries too, not that theirs have been quite as damaging. The injury explanation is valid, but it’s only partial, and the Nationals have the rest to answer for. I’m sure they’ve been searching for those answers for weeks.
It’s kind of funny — right now, the narrative is that the Nationals are a colossal disappointment. If it weren’t for sequencing, the Nationals and the Mets would be tied, and maybe the story would be more about how that Nationals have overcome adversity to remain competitive. Even in a season this disappointing, it’s arguably just cluster-luck keeping the Nationals out of a share of first. But then you get the Matt Williams effect, and that’s not just luck. The Nationals were supposed to be sufficiently good that no one would pay too close attention to their manager. That is not the way that it’s gone, I’d say.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
Cluster-luck is very close to a good description of the Nationals’ season so far.
nats fans crucify med staff and cry about minor injuries while rooting for a team that has yet to actually lose an important player for substantial portion of a season when anything mattered to their fanbase, based on DC SPORTZ bogeyman and need to paint themselves as poor, long-suffering fans = ok
having concern over a pitcher’s injury/regression based on previous performance/injury history and franchise track record with injuries = totally unreasonable and akin to rooting for bad things to happen to perpetuate a shtick or just a symptom of obvious lunacy
Do you speak English?
Fan Nat nailed medical personnel and lightly wounded moaned while I cheer for the team that was in fact losing an important player for a significant part of the season when something important for our fans, based on the DC Sportz Imp and I had to the poor fans draw UG = OK
I have a concern about a violation of the pitcher / regression based on performance history, previous injuries and franchise history with injury = totally unreasonable and as the root of bad things happen, deepen outline or just a symptom seems crazy
wow — this post was improved in clarity and quality by going through the SVF meatgrinder…
also, can we invite the DC Sportz Imp to the Jeff Sullivan Salmon Feast?
Sounds great, he’d be great to have around!
Just don’t get him comparing Bulldog and Sooner fans….
“lightly wounded moaned” is sheer poetry. Finally, we get a worthy sequel to “Swan Swan H.”
We in the mental health profession refer to this garbled statement as a “word salad,” a symptom of schizophrenia, schizotypal personality disorder, drug-induced psychosis, or just plain ol’ idiocy.
*thpthpthpthpthpthp*
Denard Span and Anthony Rendon have both missed significant time to not insignificant injuries. But you know that.
Whatever, Nats suck this year for a lot of reasons. Injuries are just one of them.
dc fan answer to anyone taking shots at them, ignore and/or mock because it can’t possibly be true
what a fanbase!
such fanbase
much wow
Is there a medication that you may need a lot of and have taken none of or maybe too much of today?
small sample size much?
Williams is a really bad coach.
“And it’s not like the Mets haven’t had injuries too, not that theirs have been quite as damaging. ”
An unbiased source says the Mets have had the worst injury damage other than Texas in all of MLB.
http://www.mangameslost.com/category/mlb/
Granted, the Nats are close, but the Mets lost David Wright and Travis d’Arnaud for most of the season, and long stretches of other key players, and their #3 pitcher, Zack Wheeler, never threw a pitch. Jerry Blevins, Vic Black, Bobby Parnell in the bullpen (and when Parnell did come back from TJ he cost them games). That graph also does not include, as it shouldn’t, the PED suspension of Jenrry Mejia, who was their closer pre-season.
Injuries, though part of this narrative, have nothing to do with why the Mets are ahead of Washington.
Obviously interesting but I don’t love the impact methodology.
http://www.mangameslost.com/about/frequently-asked-questions/
Of course not. It doesn’t match the narrative.
Read the methodology. It is not very good! I don’t have an agenda here. I don’t really care about the Nationals.
I agree, it’s not great. Probably a better way would be to take the preseason expected WAR for a player, and then remove the % of those for the time a player missed. That’s obviously also a very unrefined way (it would ignore players like Lagares or Zimmerman who have also struggled while playing because of their injuries), but I’d bet the Mets still come pretty close to the Nats in terms of % of WAR lost. For example, Wright and d’Arnaud were expected to produce > 33% of the position player wins. Be interesting to see something like that.
Perhaps the methodology is flawed, but whether the Mets or the Nats were more impacted by injuries, it’s close. The Mets lost their 3b, #3 hitter and captain for very nearly the entire season until recently, in addition to their cleanup hitter (Duda) for a bit, their middle of the order catcher for a long stretch, their mid-rotation starter for the entire season, their left handed specialist in the ‘pen, and their gold-glove CF is playing with a clearly injured arm all season. I cry no tears for the Nationals.
And you didn’t even mention that Murphy, who was a consistent 3 WAR player from 2012-2014, was out for all of June, crushing the Mets offense.
wow. espn made a note of this, while fangraphs missed it!
http://espn.go.com/mlb/story/_/id/13610059/how-new-york-mets-stole-nl-east-washington-nationals
the world is coming to an end!
I guess the argument is that because the Nats have more productive players, the injuries are a bigger deal. There’s not a single Mets hitter (oh wait never mind we have Cespedes) who could outpace a healthy Anthony Rendon. Granderson has been our best hitter this year, and he’d probably be the third or fourth most productive guy on the Nats.
Of course, I don’t think that’s a particularly good argument, given the fact that every team does have to deal with some injuries. We’ve seen teams before that had the BEST GUYS but then had crappy season because their backups were terrible. The weird thing is that for the Nats their backups have been decent but their best guys except for Harper have all been awful.
And then you have the Mets, who were playing their third string of guys in June. I mean, who honestly thought Robles, Morris, or Cecillani would see playing time this year?
Yeah, to some extent Rizzo has been building fragile rosters. That’s a roster construction choice, not just bad luck.
This. And by itself it wouldn’t be an insurmountable problem, except that then you have to be willing to patch those holes midseason, which Rizzo has been entirely unwilling to do for the duration of his tenure as GM. Just saying, “Oh when we get our guys back it’ll be as good as a trade” is inadequate given the uncertainties surrounding the performance of players coming back from injuries. Relying on guys coming back from injuries to replicate their projected performance, and to stay healthy, is Rizzo’s prime flaw as a GM. Also his total inability to construct a decent bullpen. Not sure why we didn’t just do what the Blue Jays did this year or the Angels did last year and turn the bullpen from a liability into a strength by picking guys up at the trade deadline.
Yep, complaining about Stras, Rendon, Zimmerman, or Werth being hurt is pretty silly based on track record.
That’s assuming that “a healthy Anthony Rendon” is the 2014 version when that season is dramatically different from any than Rendon has ever produced as a professional. Even if he had stayed healthy, he was due for major regression.
I think that’s a very fair point. Rendon was excellent last year but expecting a repeat of that performance may have been a tad ambitious/optimistic.
I don’t know how to quote someone as well as you, so I’ll piggy back on your comment.
Rendon, in last year’s breakout:
.287/.351/.473, .361 wOBA, 131 wRC+. 21 HR in 683 PA.
Travis d’Arnaud, ever since he was demoted last June and changed his approach:
.280/.339/.515, .368 wOBA, 139 wRC+. 21 HR in 478 PA.
d’Arnaud has been every bit as good of a hitter as Rendon’s breakout, and that’s while catching. Of course, the big drawback to d’Arnaud is the injury history – but if we’re giving Rendon a pass for injuries, we should do the same to d’Arnaud.
Are you sure there aren’t any Mets hitters who could outpace Anthony Rendon? You might want to look a little bit closer.
How about David Wright, owner of 52.6 career WAR with a career 134 wRC+. And that’s not ancient history; he’s 32 years old.
If you ignore 2014, where he admitted to playing through a serious shoulder injury, Wright combined for 13.6 WAR in his two most recent seasons, eclipsing a 140 wRC+ in both years.
Wright was probably the best player on either team to miss significant time.
David Wright had two 6 fWAR seasons from 2012-2014. Anthony Rendon had one. They were both projected for 4 fWAR this season.
I fell four hundred and sixty towers for the reasoning that I don’t explain. Even delivering and not being just hostile for the reasons I just explained can’t find me something that churns the table here. Best to be left overture handling happy prospective clientele whenever starting buttercup and not taking any prisoners.
How about the fact that the Nats defense sucks by every measure? Forgetting a few things, Jeff?
Defense is a part of WAR
Yes, but in this day and age of shifting and positioning taking such a prominent role, can we not attribute bad team defense to management more than we can other aspects of the game?
Talk about shifting the goalposts. That is not the point you originally made.
Thanks for beating Williams up on the BaseRuns +/- thing and generally giving me a free pass, guys. Really appreciate this. I mean, I know that I’ve had some pretty favorable history on this one, but -14 really leaves some brown stains in the sheets! KnowwhatImean!
Hold on a sec guys, Beane needs me at a press conference. Hope this thing doesn’t bite me in the butt! LOLZ!!
I’ve also gotten it in the shorts when my Fister came apart.
Hey guys, just checking in.
Um, Jeff forgot the Curse of Strasburg: the Nats’ presumption in 2012 that they’d have all sorts of future playoff series for Strasburg to pitch in, and thus could sit him for the 2012 playoffs, angered the Baseball Gods.
When the Mets faltered to a 7-1 deficit last night, I thought that the Curse of Harvey had replaced the Curse of Strasburg. But then I realized the Baseball Gods were punishing only Harvey himself, not the entire Mets organization.
I know it’s hard to sabermetrically quantify the anger of the Baseball Gods, but that doesn’t mean it isn’t there. (WAR = Wins Almighty Removed?)
Much like other gods, the baseball gods too appear to have been co-opted by the interests of the powered class.
Do you think the Cubs are playoff-bound this year because the Baseball Gods are “Back to the Future” fans?
But too bad they can’t play Miami in the WS. Perhaps the baseball Gods could have also put Tampa in the playoffs too, and then it would have at least been close.
The real problem with pulling Strasburg from the ’12 stretch run aside from taking out your best pitcher, is that it says to your team “we’re not really trying to win this year”. I think teams say similar things to their players when they buy, hold or sell at the trade deadline. The ’14 and ’15 Blue Jays are a great contrast for that. 2014, they made no moves and fell apart, based, the GM said, on the notion that many key players were hurt. The ’15 club made major moves…
Those odds are…amusing.
I was really hoping that during last night’s game Harvey would go 8 perfect innings and then get pulled in the 9th. I was hoping that when asked about the decision after the game, Collins would say, “Well, Harvey is on an innings limit”
Afterwards, I was hoping Boras would then send out a tweet saying the Mets unfairly restricted Harvey’s IP in an effort to artificially suppress arbitration pay-outs.
Then I was hoping Alderson would bite back and say that the Mets were only heeding Dr. Andrews’ mandated innings limit, and that Boras needs to be more concerned with the future health of his client.
Then I was hoping Harvey, too, would get pissed, and Alderson would try to assuage him by saying, “I couldn’t risk you getting hurt just to take a shot at history; I have the rest of my career to worry about.”
In conclusion, the Nats have fallen short of expectations.
That would have been especially good if he threw those 8 perfect innings with about 75-80 pitches, showing how foolish it is to worry more about innings than about number of pitches thrown.
Where’s my ring?
Dan Uggla got it already, bro.
Love your writing, Jeff, and this article is great, but may I suggest not referring to erratic performance as “bipolar,” which is a legit medical condition and at times horribly misunderstood and treated? I just think there are probably better uses than that. You wouldn’t refer to something as Hepatitis-like, or Multiple-Sclerosis-like, and bipolar, while a radically different illness, is still an illness all the same. Thanks dude, keep on writing. 🙂
As someone with bipolar ii, I find attempts to restrict the language in this way counterproductive and silly. It’s like you’re saying my condition is so terrible as to be unspeakable, which is actually a much worse feeling than seeing the word used casually in this context.
So don’t be that guy anymore, please.
I couldn’t disagree with you more, but I cringe whenever I hear someone use the term bipolar to refer to something like a baseball season or someone driving. Flippant use of the word really makes it seem like it shouldn’t be a bother to those who have it. I try to avoid saying things like retarded or schizophrenic as well. I am heartened whenever someone has the courage to say something like I-hate-to-be-THAT-guy did. I appreciate it. But I leave it to each person to figure out what they feel is appropriate and what goes too far. Am I offended? NO, but would I like to hear something else? Very much so YES
Please note that if you look up the definition of Bipolar, it has multiple meanings including the perfectly acceptable way Jeff used it here.
Please stop getting offended by every little thing. Please.
The season has been so gay it’s retarded.
[sits at end of bench hoping THAT Guy forgets to notice he’s been there for like a century]
Stop trying to be me. Or try harder. Whatever.
look up bipolar in a dictionary. “having or relating to two poles or extremities.” bi-polar literally means two poles. The disease is just one use of the word.
those other words have no accepted definitions outside of their medical ones, bipolar does.
Dictionaries respond to how words are being used by people at large. So I don’t think it should be used like some sort of bible when it goes along with the point we want to make. It’s a reflection of how words are being used, which is fluid over time. I don’t think it’s a rule book that defines our lives.
There are probably some offensive words that have non-offensive Dictionary Definitions. That doesn’t mean you and I should use them, or that people can’t be upset by their usage simply because it’s written in a book somewhere.
On the plus side, at least football season’s here?
Oh…
HTTWFT
Now that Bryce is firmly in the Mike Trout tier, how long until he is in his own tier?
Staying healthy is one thing, but I dont think this is peak level Bryce. Both Trout and Harper could have 100 WAR by the age of 30, but who will end up with more?
also, imagine what Harpers numbers would look like in Yankee Stadium.
Will look like. Will.
“Both Trout and Harper could have 100 WAR by the age of 30”
Naw. They are amazing, but naw.
Rounding up this season’s WAR a bit, Harper would need 80 WAR over the next 8 seasons (age-30 seasons included) to get to 100 WAR. From 1920-27, Babe Ruth compiled ~84 WAR–including the top 3 highest single-seasons for position players by WAR, 5 of the top 10, and 6 of the top 11. Somehow, I don’t think Bryce quite has it in him.
Is the relative quality of the league around him a factor in WAR? Because there’s about a 0% chance of the Bambino putting up those numbers in a league where black guys are allowed to play, and there’s also baseball factories in the Dominican, the best Japanese and Cuban players, etc.
As amazing as Ruth was, he had it pretty easy, in baseball terms. How many people were competing for every spot on a team then? (dividing total population by open spots).
I think what’s lost in talking about where the Nationals went “wrong” is the fact that most pundits vastly underestimated this Mets team.
None of the FanGraphs writers expected this to be a close division. FanGraphs expected the Nationals to run away with the division. Even if you normalize for “cluster” luck, the Mets and Nationals are still neck and neck in the division, with the Mets slightly ahead. Cluster luck doesn’t explain the Mets being as good as the Nationals.
Neither does injuries. Sure, the Nats have been hit extremely hard with injuries. Rendon, Span, Werth, and Strasburg have all missed significant time.
But the Mets have had more than their share of injuries. David Wright, who had been a 6 WAR player in 2 of the last 3 years, was out of the lineup from early April all the way through mid-August. Travis d’Arnaud, a former top prospect coming off an impressive 2nd half last year, has been injured for roughly 90 games. Daniel Murphy has missed time. Juan Lagares has played through a bad elbow and rib-cage, causing his defensive performance to deteriorate.
And that’s just the position players. Zack Wheeler, expected to improve on a 3 WAR season in 2014, missed the entire year due to TJ surgery. Jennry Mejia, the Mets presumptive closer, has been suspended for most of the season. Josh Edgin, the Mets LOOGY, missed the year due to TJ surgery. Jerry Blevins, the LOOGY the Mets traded for after Edgin got hurt, has been out since April with a broken arm.
So if the Nats haven’t had worse injury luck, then why have they played similarly according to BaseRuns?
I’ll answer that question in a reply to this comment.
Raise your hand if you thought the Mets would be 17 games over .500 going into mid-September.
Didn’t think so.
Uh, you can say that, but go back and read what I wrote before the season.
Here’s what I wrote in a comment to a FanGraphs article about the Mets rotation:
“Now that I’ve got my mechanical critique out of the way, I’ll say this: isn’t anyone skeptical of the projection systems?
The Mets finished last year at 79-83 with a pythag record of 82-80. They did that while a) missing their best player in Matt Harvey b) getting less than 2 WAR from their best hitter, David Wright c) getting no production from the LF and SS positions and the C position until July d) Granderson having an unusually low BABIP and ARM rating component of UZR and e) having a bullpen filled with retreads Kyle Farnsworth, Jose Valverde, and John Lannan until June.
These factors, combined with the fact that most of the Mets team is young and still improving, a weak NL East, and a strong NL Central (hurting NL Central contenders), gives the Mets a ton of upside and very limited downside. I wouldn’t be surprised if you’re hitting yourself in a few months for writing this article, as many other FanGraphs writers will be as well.”
Looks pretty spot on to me. I’ll raise my hand.
Don’t break your arm patting yourself on the back.
That doesn’t come close to you saying you thought the Mets would be THIS good. It says you thought they would be a better than they were projected.
Side Note: Why can’t can’t fans just enjoy what’s happening? What is out culture’s obsession with feeling as if they are being persecuted in some way at all times?
I’m not obsessed with feeling persecuted. We’re talking about what went wrong with the Nationals, and I disagree with you about the root cause of the Nationals being 7 games back.
The real reason? The projections (and the pundits) had recency bias when projecting future performance.
Take Jayson Werth and Curtis Granderson, for example. The projections (and FG writers) expected Werth to be a 3-4 WAR player, while Granderson was projected for 1-2 WAR in most circles.
If you look at the two players’ careers, however, this makes little sense; Werth and Granderson were similarly valuable over the past 4-5 years and Granderson was two years younger than Werth. Not only that, Granderson had played all of those years in the AL East, meaning you could expect his performance to elevate as he played worse competition. The only reason that Granderson’s WAR was so low in 2014 was because of a fluky BABIP and a very questionably low ARM rating component of UZR (his range was fine).
There are countless examples like this. The projections liked Rendon way more than Wright despite Rendon because of Rendon’s career year and Wright’s down year playing through a shoulder injury.
The Nationals were filled with players due for regression. Desmond had a career high K rate last year. Roark, Fister, and Zimmerman all over-performed their peripherals significantly, with Fister’s peripherals dropping alarmingly.
Meanwhile, the Mets had lots of players due for positive regression. d’Arnaud, Flores, and Granderson all were extremely unlucky in 2014 BABIP-wise, with d’Arnaud and Flores due for youthful improvement anyway.
The bottom line is that both the Nats and Mets were really well constructed teams heading into 2015. Both teams suffered far more injuries than normal, and have hence played worse.
But the thing that people missed was just how much talent the Mets had coming into the season; they weren’t actually behind the Nationals, if anything they were neck and neck. With Cespedes and some poor cluster-luck for the Nationals, that’s been enough to put them over the top.
You won’t win here with a differing opinion, even if it’s backed by facts. The national narrative was that the Nats were going to be the best in the NL. Now that they aren’t, it’s one excuse after another as to why they collapsed. You won’t get much support for the idea that the Mets were in solid position for a playoff berth at the beginning of the season.
There is no “narrative adjustment” in the projections. That’s the whole point.
L. Ron Hoyabembe: But projections aren’t infallible. Part of the reason people didn’t pick the Mets was because of a MAJOR error in the way FanGraphs calculated park factors for pitchers.
After FanGraphs fixed the error around Opening Day, the Mets went from around 25th in projected WAR to 10th. And that’s with very pessimistic projections for Granderson, Wright, deGrom, Syndergaard, and others.
Not infallible.
But more reliable than fans of team X using hindsight.
Actually, if I recall correctly, this isn’t hindsight by Noah. He’s the one who brought the FIP park adjustment issue up in a community research post.
Not recalling correctly. Anyone challenging the status quo, proven right or otherwise, must be pilloried. It’s been an interesting turn by the once-curious sabermetric community.
“Not infallible”. You’ve seen the projections at the top of the page right, where they got one guy right out of the whole team, and were massively out on several players? Why do we waste so much time on projections again?
Werth was actually projected for 2.7 over 125 games per depth charts (on Steamer 2.8). Depthcharts were a mix of steamer and zips with playing time projected. And generally regression and injuries were expected. Werth averaged a lot more than 2.8 the last 2 years, even with some down time, and Granderson hadn’t done much the last 2 years (acknowledging some flukiness to his arm rating a babip last year). He’s been great. But Werth was projected for 125 and he’s played 68. Rendon projected for 117 but he’s played 59. Span 125 games projected, 61 played. Zimmermann projected for 133 and he’s played 95. And they’ve been worse than projected by WAR, a lot likely regression and a lot just playing hurt it seems.
Not making excuses for them. I am a Mets fan. But I don’t think the projections were that unreasonable. They projected injuries and some slippage, just not the acrosss the board almost less time and degree of the slippage. And Ramos and Desmond have played or will end up playing about the same number of games or more than projected. But they’ve also been much worse than projected. Werth, Rendon, Desmond, Zimmermann, Ramos and Span all underperformed the depth chart projections by a combined 14.1 WAR.
Oh no, I don’t think the projections for the Nationals were unreasonable at all. I think they were slightly optimistic, but nothing crazy.
Rather, I think the projections were significantly underselling the Mets, who were a good team all along.
I would argue that injuries hurt the Mets and the Nationals much harder than most teams, and as a result both teams underperformed (Nats still underperforming, Mets until the trade deadline acquisitions).
But I think the projections undersold the Mets significantly for the very explicit reasons I outlined above. And that’s not even getting into the FIP Park Factor mistake, which projected a very good Mets rotation to be 26th in pitching WAR due to a very fixable error.
True enough. I recall in June or so the Mets hitting was horrible but I looked at the ZiPS projections and the whole line-up was hurt or under performing projections. Duda was in a massive slump, Murphy hadn’t come around yet, Wright down, d’Arnaud down, Cuddyer under performing. Only Granderson was doing ok (ot sure if he had broken out yet or not). Even the guys like Campbell were below projections. It was almost across the board and you could see that Alderson had assembled what looked to be at least an average hitting attack (going by projections almost all spots other than cf were thought to be average or better by rest of season projections) that in fact was performing woefully.
The deadline acquisitions and Conforto call up helped, but so did d’Arnaud return, Wright return, Duda got red hot for a bit then hurt, Grandy bounced, Murphy got hot (then hurt again), even Cuddyer and Lagares, now spotted against lefties. A perfect storm leading to the Mets leading the majors in post-break WAR, both position players and overall.
“So if the Nats haven’t had worse injury luck, then why have they played similarly according to BaseRuns?”
Yoenis.
The nats went wrong when they kept Matt Williams after he showed he had no clue how to manage in the playoffs last year (Let me let this LOOGY face Posey who was up third in the inning, then post game talk about how the plan was to get a righty ready for posey if he was gonna make it up…)
Can’t blame him for everything, but he screws up so much and doesn’t even have someone trying to stop him. Just look at his double switch a few days ago where he actually made the pitchers spot up sooner.
Sigh.
I agree with the idea that managers seldom have a major impact on a team’s record, at least directly, but in the case of Williams, we have an outlyer. He’s appalling. When he removed Strasburg last night, you could hear the Mets dugout cheer. Stras was not about to pitch a complete game, but an early hook gave the Mets new life, which was exactly what the Nats did not want to have happen. His moves have been consistently awful, not just in hindsight but as they were happening. In addition, while Terry Collins is no brilliant tactician, he’s had his teams play hard – very hard – day in and day out. There’s never been a sign of surrender there. This Nats team has collapsed like a house of cards at the first sign of misfortune. That’s leadership; that’s Williams.
The preseason odds you linked at the top are actually quite hilarious. Compared to how bad you guys did on many of the other teams, your Nationals prediction actually stands out for being somewhat close.
You’ve got the Red Sox winning the East, the Indians winning the central, and the Mariners winning the West. You’ve got the Royals with a 9% chance of winning the Central, the Astros and Rangers with 4.8%, and 0.8% chances of winning the West respectively.
You’ve taken the angle that the projections were fine and the players have unexpectedly underperformed. You could reasonably take the opposite angle that the players have done nothing unusual and the projections were bad. In reality it is a bit of both.
Yea – the problem with these projection systems is that their tails are way too small… I am not conversant in the maths, but even late in the season they appear way too overconfident in both directions. In reality, we should probably adjust them all closer to 50%…
Preseason projections are not predictions of the future. They are meant to tell you the statistically most likely scenario, and you can do with that information whatever you want. They tell you, “Flip 100 coins, you’ll get 50 heads.” Of course, it’s actually rare that you would get exactly 50 heads.
There is also an additional wrench thrown in, the trade deadline. Teams that maybe hit a run of “heads” or wins have the opportunity to alter the coin to get more heads. Likewise for the losers.
This is not correct for the playoff odds. To use your coin flip analogy, the playoff odds FanGraphs gives are more like flipping a coin 100 times and giving the probability of getting heads at least X number of times. FanGraphs said if you ran the 2015 season over and over again, the Astros would only win the west 4.8% of the time, and that most of those trials would have the Mariners winning. They were either extraordinarily unlucky, or they were just plain wrong.
The analogy isn’t perfect because it is obviously not that simple. The projections said that, simulating the season many times, the Astros won the division 4.8% of the time. We can’t really know the true odds because we can’t run multiple trials of the actual season, we are stuck with this one data point where it looks like the Astros could win the division. Bear in mind that the simulation probably included seasons where the Astros played well, perhaps even as well as they have this season, and still didn’t win the division or even make the playoffs. There are a lot of moving and changing parts to account for, and I think the projections can and will be improved. I just want someone to come up with a stronger indictment of the current projection system than “the results of our n=1 data set were unlikely given the predicted odds.”
I am not sure they can reasonably be expected to do much better, However, It is not n=1.
They did poorly with:
Red Sox
Blue Jays
Yankees
Indians
Tigers
Royals
Twins
Mariners
Angels
Athletics
Astros
Rangers
Nationals
Mets
Cubs
Brewers
They did well with:
Rays
Orioles
White Sox
Braves
Phillies
Cardinals
Pirates
Reds
Brewers
Dodgers
Rockies
Diamondbacks
Giants
You are never going to look very good predicting baseball.
Sure. But you can use statistics to prove whether or not the projections themselves are actually accurate.
For instance, if you flip a coin 1000 times and get 650 heads, you can statistically infer that it’s not a fair coin, because the probability of getting that many heads over that large a sample is so small.
Likewise, if the projections continue to be largely incorrect year after year, you can statistically infer that the projections may have to be altered. I’d argue that until the projections themselves become more accurate they should reduce their confidence in the projected win totals.
This is a good point. Others have argued that many of the public systems are too slow in integrating new performance data. To some extent i agree, particularly with older players who appear to be declining in production or are oft injured. The Nats of this season are the perfect example of the type of team that public projections are bad at identifying.
Of course, folks are also claiming that paying too much attention to recent performance (Rendon’s good 2014, and Wirght’s bad 2014) is the REAL reason the projections don’t work.
I think people are just grasping at straws.
Well that’s your opinion, but I was saying that the Mets were being overlooked before the season started and was ridiculed.
I don’t even think the projections were wrong regarding the Nationals, outside of a handful of examples. The Nationals had a ton of injuries and bad cluster luck, which led to underperformance.
What nobody had addressed is how the Mets have managed to vastly exceed the projections while also having just as many injuries to key players (if not more) than the Nationals.
And this isn’t a Mets fan thing either. People on Mets websites also expected the team to finish with a low to mid 80s win total (which might have something to do with the cynical nature of Mets fans, but that’s a whole different discussion).
The Mets were completely overlooked heading into this year, and I can think of a few reasons why (that I pointed out at the time):
The consensus was that Granderson was done, despite a very unlucky 2014 season.
The consensus was that Syndergaard would have a minimal impact in 2015, despite elite peripherals in an extremely hitter-friendly environment.
The consensus was that d’Arnaud would have a 2-ish WAR season, despite being a former top prospect who had a massive 2nd half in 2014.
The consensus was that deGrom was a fluke, despite peripherals that backed up his 2014 performance.
The consensus was that Duda would regress, despite strong exit velocity numbers and generally sound peripherals.
There were plenty of things that didn’t check out that people simply went with because two rather basic projection systems told them to.
Papelbon is a team cancer, he upset the entire chemistry of the bullpen which spread to the rest of the team.
and Matt Williams shouldnt be managing in Little League, much less MLB
As good as Mike Trout has been doesn’t equal success for the Angels either.
You obviously can’t do it with one player, Harper injured would have put this team 6-7 games under .500. Just one large collapse in general.
Holy crap – the Nationals’ playoff odds are lower than the Indians’.
Pretty simple explanation. This is a team that has been injured and unlike other clubs this year that have struggled with injuries to important pieces (Cardinals, Dodgers, Jays), the Nationals are not constructed with sufficient depth to withstand all of those injuries. Whether it was reasonable to think the Nationals would get the projected production out of oft-injured players like Ryan Zimmerman, Jayson Werth, Anthony Rendon, and others is an important discussion. Somehow I think the projections undersold the wide variance possible with these Nationals due to the injury history.
I haven’t finished the whole article yet, because I had to immediately comment how great that first sentence is. Spot on, Jeff!
Look, Matt Williams is not the best mgr in baseball, that is so, but the NATS cannot cry “it’s the manager” or “injuries” when the Mets had as much if not worse.
The real key has been head to head where the Nats killed the Mets last year, and this year the Mets have won the last 6 straight, kicked them in the groin to catch them in NY, and now in DC, to bury them.
Nats split those last 6 and this is a damned good race right now. Next year will be quite a race. The Phils,Braves and Marlins will still suck, and the Mets and Nats, hopefully without key injuries, will be a great watch. Nats may lose P Zimmerman, Mets will pick up P.Wheeler. Nats will have a new Mgr. If Mets keep Cespedes,I think their lineup is better and give nod to them. Maybe Nats Wild Card, but the Central Division is a B!%#H.