The One Thing Holding Back Yasiel Puig
There’s so much to like about Yasiel Puig, and the season he’s having in 2017. He has a career-high rate of walks, and a career-low rate of strikeouts. He has a career-high isolated power, and he also has a career-high 12 stolen bases. He’s remained, for the most part, totally healthy, even on a team that makes liberal use of the disabled list, and Puig’s even got easy career-best marks in both defensive runs saved and UZR. In so many different ways, Puig’s game is looking more polished. Yet his WAR is simply a hair over 2.
There’s only one thing that has held back that number. I mean, all right, sure, Puig could stand to have a higher BABIP. He’d look better if he gathered some missing singles. But there’s just one area where Puig doesn’t look good. If you know how WAR is calculated, you’re probably one step ahead, but I should issue some quick background context. Some numbers around these parts have changed.
A few of you might’ve noticed some funny-looking baserunning numbers over the course of the summer. It’s nothing we tend to pay a whole lot of attention to, but especially on the team level, the numbers were looking atypically extreme. For that reason, I refrained from writing about said stats, but, quietly, the other day, a correction was put in place. The baserunning numbers should be normal now. The baserunning numbers should be trusted now. Yasiel Puig rates as one of the worst baserunners in baseball.
Our baserunning stat is BsR. You see it on the leaderboards. The BsR leader is Billy Hamilton. Of course it’s Billy Hamilton. The player in last is Asdrubal Cabrera. He’s around some of the usual names, like Miguel Cabrera, Victor Martinez, and Kendrys Morales. But Puig is down there, tied for second-worst. The metric doesn’t like what Puig has done, and although it’s never claimed that Puig was excellent, this year looks like the career outlier. I don’t know how else to interpret this.
Let’s say you’re not totally convinced. Let’s say you’re hesitant to trust any formula that was just updated this week. Baseball Prospectus has its own baserunning stat. It thinks Puig has been way worse this season, too. And Baseball Reference has its own baserunning stat. Same page. So you could say there’s universal agreement: 2017 Puig has been an ineffective baserunner. What, then, is the story?
To make it all the more interesting, you might know that our baserunning stat is a combination of three sub-stats. Stolen bases make up one pillar, and this year, Puig has been a better basestealer than ever. He’s actually improved by some increment every year, since being a rookie. Despite that, Puig has still been a negative. He’s still been a negative, even though the Statcast sprint speeds don’t show him having significantly slowed down. The problems are elsewhere. The problems involve double plays, and baserunner advances.
The double plays are easy enough to analyze. You can look at Puig’s career rates of double plays, over double-play opportunities. Puig’s rate this year has skyrocketed. His rate is double what it was a season ago.
Puig, this season, has hit into a double play 22% of the time. The league average is 11%; Puig’s own average last year was 11%. The year before that, Puig hit into but one single double play. Now he’s tied for baseball’s fourth-highest total, one behind Tommy Joseph and Albert Pujols. Out of everyone with at least 50 double-play opportunities, Puig’s rate is the seventh-highest, and he’s got the fifth-biggest increase from a season ago. In that respect, he’s not, say, Matt Kemp, whose double plays are out of control, but you can have a problem without it being the biggest one around.
Part of this, I’m sure, is a fluke. Puig’s natural double-play rate couldn’t possibly be this high. But this season, with men on base, Puig has a career-low strikeout rate, a career-high pull rate, and a career-high grounder rate. Being a righty already gives him a slower start out of the box. The double plays have hurt, the Dodgers more broadly and Puig more specifically. This is part of the explanation.
Then you have the matter of advances. Baseball Reference makes this simple to analyze. It doesn’t show everything, but it does show how often a given player takes the extra base when he’s aboard for a hit. Like, say, going from first to third on a single. This year, the league-average rate is 39%. It’s held pretty steadily there for a while. Here is Puig’s career.
Puig this year is at 27%, which is half of where he was last summer. Historically, he’s always been around 50%, and the only player with a bigger drop from 2016 is Kevin Kiermaier, who’s fought lower-body issues. Puig has been a more conservative baserunner, yet at the same time, he’s run into a career-high rate of outs when trying to move up on a hit. Sometimes it’s just because a strong throw is made, and you’ve got a bang-bang play.
Other times, it’s not a bang-bang play.
One of the complicating factors is that baserunners don’t make all of their own decisions. They can’t control where the ball is hit, they can’t control who’s on base in front of them, and they can’t control the signal from the third-base coach. In that White Sox clip above, you might blame the coach for waving Puig around in the first place. Technically, Puig could’ve elected to stop if he wanted to. The runner is never entirely blameless. And it’s not like Puig hasn’t had to own his own mistakes. Puig ran into eight baserunning outs, combined, in 2015 – 2016. He’s at eight already in 2017, and this doesn’t count failed steals. Here’s a failed contact play.
And here’s a failed read of a different grounder.
The consolation, for me, is there’s almost no way this is a reflection of Puig’s actual new baserunning true-talent level. I can’t buy him as being as bad as Cabrera or Morales or Martinez. He’s so much faster, and so much more athletic, that he ought to end up with a greater rate of successes. Baserunning numbers, ideally, require a large sample of data to find their level, and Puig’s level is higher than this. He’ll take some more extra bases. He’ll come up with men on first and bounce into fewer double plays. Too many things to this point have gone wrong, too many to fully believe.
But, at least, that should explain why Puig’s WAR isn’t higher than the one that you see. Nearly everywhere, he’s taken a step forward. His baserunning performance, though, has ranked among the very worst. He ought to be able to shake that off, but it’s not going to happen overnight.
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.



I went to go look at Puig’s sprint speed score after Jeff’s suggestion that he hasn’t tailed off, since the people who are at the bottom of the baserunning leaderboards are often the same ones as the guys at the bottom of the sprint score. Puig sits right there with Mookie Betts and Jose Ramirez, and about 5 ft/sec higher than the guy at the bottom (Pujols, who to be fair is way lower than Brian McCann below him).
So either Puig has had extraordinarily bad luck or he’s making a lot of bad decisions. Given who we’re talking about we can’t rule out the latter, but Jeff does make a convincing argument for the former.
Could be a bit of both!
You have to wonder if it’s in his head a bit with people wanting him to be less aggressive
I think you are missing a piece here. Puig’s xBABIP is *horrible*. If he deserves a low BABIP because of quality of contact issues, that could easily drive the double play problem too
Puig’s batting average with runners on base is nearly 100 points lower this year than last year. Perhaps a defensive shift is to blame…?
On a broader note, perhaps “grounding into a double play” shouldn’t really count as a “baserunning” issue in the first place…
Eh, I think GIDPs should be factored in somehow. They’re certainly not context neutral, but not doing it has huge value. A guy who strikes out in that situation is somewhat more valuable than a guy that causes two outs, IMO.
While I understand the logic for it, I’ve never been a huge fan of including double – play avoidance in baserunning metrics rather than slightly increasing the batting penalty for hitting into a double play. Obviously, it’s an insignificant change, but it would seem to more adequately reflect a player’s skills in each area
Yes, it’s not as if speed isn’t already factored into batting statistics in the first place. BsR feels like it should be a measure of what players do once they’re a runner, not a batter-runner.
One that does cause some trouble is players being thrown out stretching hits into doubles or triples – if they succeed it goes on their batting line; if they fail it goes on their BsR.
To calculate UBR you need to know the average run expectancy for a given batting event given the base-out state and then the actual run expectancy change that occurred during the play.
I was right (source:http://www.fangraphs.com/library/offense/ubr/).
I guess this might raise the question if they’re double counting XBhs.
I misunderstood your last point, but as you point out with BsR already takes into account the result of a batter’s hits, not just what he does once on base, so I don’t think there’s much of a stretch to include GIDPs in it. Especially when you consider how UBR itself is calculated (even though it’s separated in wGDP and UBR):
“To calculate UBR you need to know the average run expectancy for a given batting event given the base-out state and then the actual run expectancy change that occurred during the play. ”
A double play is essentially always going to be the worst possible result for RE.
This is something I have been wondering about all season…thanks for the article…very informative.
It seems like the double play rate could be at least partly explained by his low babip (i.e., soft contact or hard contact directly at fielders)…If I’m understanding correctly, the WAR calculation considers the three components of baserunning in aggregate rather than as individual factors. Since his WAR is already being knocked for low BABIP via his hitting results, does the calculation factor out the negative hitting outcomes (i.e., fewer hits, lower on base from the BABIP Gods) when accounting for his baserunning value?
Cant most of Puigs DP and babip problems be attributed to his struggles vs LHP (44 wRC+ & 63 GB%)? His babip is normal vs RHP & he’s hit into 10 DP’s vs RHP and 6 of those came in April, seems like if he can stop pounding the ball into the ground vs LHP he fixes the DP and babip problems.
When I first saw that second graph, I couldn’t help but think of that Price is Right game (Cliff Hangers) with the little yodeling dude that climbs the mountain.
Now that damn yodel song is going to be stuck in my head.
Along with Bob Barker reminding me to help control the pet population.
I was expecting this article to be about hitting lefties: http://www.fangraphs.com/graphs.aspx?playerid=14225&legend=1&statArr=50&split=lr&time=season&ymin=&ymax=
Everybody knows the sample size is very small for RHB VS LHP splits. You should use the last 3 seasons for that split. More importantly, almost nobody in history has had reverse platoon splits over a large enough sample size.
I’d be fascinated to read a blame-free retrospective of how the baserunning issue was introduced, and what process changes you’d be considering to prevent similar future issues.
Earlier in his career, Puig batted higher in the order. Most of his ABs from 2014 were in the #2 hole. Now, he’s mostly batting 8th.
Especially in an NL lineup, I can see how this might have an adverse effect on some of the stuff that’s measured under the baserunning umbrella. Most notably, a baserunner is probably going to take more risks with Rich Hill batting vs. Justin Turner. Taking more risks may have a negative effect on one’s baserunning stats.
Of course, while a #8 batter in the NL who gets on base will have a pitcher follow him in the batting order, he’ll also have the top of order coming up next — provided the pitcher (or pinch hitter) doesn’t end the inning.