Baserunning, Speed and the Rest of the Picture
A lot went wrong for the Mets a season ago, but they did manage to come out of nowhere to be the best baserunning team in baseball. I say that because, the season before, the team overall was below average. They took it upon themselves to be much more aggressive with their feet, and it paid off — even if, in the end, it didn’t pay off. A huge reason the Mets were so good was because of Eric Young, who is fast. Another huge reason the Mets were so good was Daniel Murphy, who is not fast. A month ago I wrote about Murphy and the curious case of a base-stealing threat without base-stealing legs. Murphy was also good at the other baserunning aspects, and he stands as a clear example of how baserunning is more than pure footspeed.
The opposite of a slow runner who’s better than expected on the basepaths is a fast runner who’s worse than expected on the basepaths. That player type exists, and I remember that last year Gerardo Parra went 10-for-20 stealing bases even though he was a plus-UZR outfielder. Watching a game Tuesday afternoon, Parra was thrown out trying for second by almost a literal mile. I’ll grant that it appeared to be a busted hit-and-run, but that doesn’t fully excuse Parra for being so far away from the base when the baseball arrived.
Busted hit-and-run, sure. It was also a pitch low-and-away, and the catcher had to deal with a batter right in front of him. Parra didn’t so much as attempt a slide. Jose Molina is 17-for-24 stealing bases in his career and this is what happened to one of the better-running outfielders in the league.
Murphy’s a guy who runs the bases better than you’d think. Parra’s a guy who runs the bases worse than you’d think. Are there any other examples, though? How strong is the relationship between speed and baserunning, and what players are furthest from the best-fit line? This can be analyzed, and it can be analyzed thanks to you. Literally, you, or at least those of you who have participated in the Fan Scouting Report.
I looked at the players who’ve batted at least 500 times since the start of 2012, and for each player I calculated a baserunning value (BsR) per 600 plate appearances. The sample pool numbers 299. Then, for each player, I imported the Fan Scouting Report’s speed rating. Obviously, this relies on the evaluation abilities of the fans, and fans aren’t always good evaluators. Still, it’s a hell of a lot easier to grade speed than to grade instincts. Speed’s pretty apparent. As the last step, using only the speed rating, I calculated an “expected” BsR/600. And the relationship between speed and baserunning is really really strong:
Or maybe just one “really.” It stands to reason the Fan Scouting Report might be in some way influenced by a player’s baserunning, anyway. But that’s a trend, and it indicates speed is presumably the primary factor when it comes to determining how good a baserunner a player is. But speed isn’t the only factor. There are also smarts and all the things that fall under that term. So why don’t we take a look at the deviations? I calculated the difference between actual baserunning and expected baserunning, and then plotted that over speed:
Viewed simply, you get four quadrants. There are slower runners and faster runners and there are guys who fall short of expectations and guys who exceed expectations. Of course, there are also guys who’re just about average. Perhaps the most average player in the sample is Joe Mauer, who gets a 49 speed rating and has been worth -0.2 BsR/600, with an expected figure of 0.0. In this regard, Joe Mauer is average.
Daniel Murphy shows up as a guy who’s been better than expected. Ryan Howard, somehow, has been worse than expected. But at the very extremes, we have Jarrod Dyson and Jose Tabata. Dyson’s actual BsR/600 has been better than his expected BsR/600 by nearly 10 runs. Tabata’s actual BsR/600 has been worse than his expected BsR/600 by more than five runs. Dyson’s speed is graded at elite, and still he’s been even better on the bases. Tabata’s speed is graded at about average, but on the bases he’s been a slight liability.
So it could well be that Dyson has an almost perfect blend of speed and smarts, while Tabata has more trouble getting good reads. Dyson’s stolen 64 bases, with 85% success. Tabata’s stolen 11 bases, with 46% success. Dyson, also, has made just two other outs on the bases, according to Baseball-Reference, while Tabata has made 11. Out of 28 opportunities to score from second on a single, Dyson’s scored 24 times. Out of 27 opportunities to score from second on a single, Tabata’s scored 14 times. I don’t need to keep running down the numbers. Wherever you look, Dyson’s been great, and Tabata’s been at best OK. You expect Dyson to outrun Tabata, but not to such a degree.
I should note Dyson has pinch-run on some occasions, so that could inflate his BsR/600 since those are times to accumulate BsR without accumulating plate appearances. Also, baserunning numbers ought to be regressed, since the samples are limited and a lot of the plays are pretty close. Most players above the best-fit line should regress closer to it, and most players below the best-fit line should do the same. But the best-fit line is based only on speed, and speed isn’t the only factor, so there’ll always be deviations. Even a regressed Daniel Murphy is still probably a better baserunner than he should be just based on how quickly he moves around.
The guys behind Dyson on the leaderboard? Eric Young, Rajai Davis, Emilio Bonifacio and Murphy. The guys in front of Tabata? A.J. Pollock, Brendan Ryan, Adeiny Hechavarria and Gerardo Parra. This post was inspired by Murphy and Parra, so it’s comforting to see them both near the expected extremes.
In one sense, baserunning is a relatively minor part of the game. In another sense, baserunning is one place to try to squeeze out extra runs — and runs are always at a premium. Some guys are better at it than others, and some guys are better or worse than they should be just based on how they sprint. We’re always looking for new reasons to praise players, and we’re also always looking for new reasons to be critical of them. Good news: There are always more reasons.
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.
Jeff, I had a question about the Fan Scouting Report numbers after reading your article in Grantland. Are the ratings taken as is, or is there any attempt to filter out any obvious outliers (fans giving a rock bottom rating to someone they hate, for example)?
As is. Nothing I can do. The overall results seem to be pretty all right, but yeah, there’s certainly some risk here.
I think I would be more worried about endogeneity than people making joke ratings or whatnot. I know, were I to provide my scouting report of a given player, it would be based a lot on the resulting statistics, consciously or otherwise. So like yeah, the player who I give a high power grade hits a lot of home runs, but I gave him a high power grade because he hit home runs. I’d rate Jacoby Ellsbury highly for speed, in part because of defense, but in large part because of baserunning skill. Obviously that’s not the only thing being used, since there is some variation in the numbers, but I think it makes the Fan Scouting Report significantly less predictive or helpful than it may seem.
Of all the things in there, I trust the speed rating to be the most accurate. It’s pretty immediately evident who’s quick and who isn’t. And I think people do have a sense of who’s quick and who isn’t somewhat independent of baserunning skill. Which we observe some of, above! If your point is that the analysis isn’t perfect, yeah, you’re right. But, there’s no doing better than this. Not realistically. And it’s not bad.
And people (that is, the kind of fans we’re talking about, who read fangraphs and contribute to the scouting report) are a lot more reflexively aware of a player’s HR total than his BsR. Off the top of my head I know the HR totals for most of the players on the team I follow, but I have no idea what their BsR ratings are without looking them up; and were I to fill out a Fan Scouting form, I’d be using those HR totals for “power” but I’d be using eyeballs and memory for “speed.”
There may be folks who contribute to the Fan Scouting Report by running down the stats on the team leaderboard, but I suspect for more subjective things like speed and defense a lot of them are using nothing more than impressions.
Oh absolutely, I wasn’t trying to suggest that this was useless or that there was a better option, just that there are reasons not to put all your confidence in these. I hadn’t thought of Speed being the most trustworthy of the categories, but that does make intuitive sense.