Where did Evan Gattis' sudden spike in triples come from?
Evan Gattis was probably never fast. That is, he wasn't fast in comparison to many baseball players he was around while coming up through the minors. He certainly isn't fast now, and at this point, we can confidently say that he probably never will be. And that's fine, because speed isn't really his game: coming into the 2015 season, he had zero stolen bases and one triple in his career. We're all familiar with how Gattis contributes in other ways, like hitting baseballs 450 feet. Yet despite his lackluster speed, by the end of the 2015 season, he updated his career statistics to read zero stolen bases -- and 12 triples.
In the span of one season, Gattis increased his triples total by a factor of 12. Because of that fact, this was a big story for most of the season; a simple Internet search yields many articles ranking and commenting on his ever-increasing number of triples during 2015. Today, instead of viewing and ranking each one, we're going to go deep on how strange and rare it is for someone as slow as Gattis to do this.
To begin with, we're going to use a statistic called Speed Score. Very simply put, it's a way of measuring a player's speed and baserunning ability. Speed Score is on a scale from zero (walks around the bases) to 10 (fastest/best baserunning human who has ever lived), so it's fairly easy to grasp, and it tends to make some intuitive sense. There are a few main factors that go into a player's score: stolen base rate, number of stolen base attempts, triple rate, and double play rate being a few of the main ones. To give you some examples, the best qualified hitter by Speed Score since 1920 is Jarrod Dyson. If you've watched the playoffs during the past two seasons, you know how fast Dyson is. The worst players by Speed Score are usually catchers, with Chris Snyder (who mainly played for the Diamondbacks in the early-to-mid 2000's) at the very bottom.
So let's see where Gattis fits into this Speed Score spectrum. First let's look at the first two seasons of his career combined -- 2013 and 2014. Here are the worst position players by Speed Score in those two years (minimum 750 plate appearances), with the number of triples each player hit:
We have three designated hitters here, a catcher, and a first baseman. Then we have Evan Gattis. We said before that he wasn't fast, but let's put it another way: Gattis was one of the very slowest -- and worst -- baserunners by Speed Score during 2013 and 2014. He was around the first or second percentile in the majors. These 10 players above hit a combined three triples during these two seasons -- basically, these types of players very rarely hit triples (and that helps bring their score down as well). Gattis actually improved remarkably by Speed Score in 2015, getting all the way up to 3.9 (46th percentile!) -- but we can be quite confident that most of that increase was due to the number of triples he hit, and not other factors, like his aptitude at stealing bases (which did not improve).
Let's visualize that last point by looking at the relationship between Speed Score and a player's rate of triples hit. I've pulled data for every qualified hitter from 2000 to 2015, with each dot representing a distinct season of a player's career. Take a look at the number of triples per 100 plate appearances and each player's Speed Score, with Gattis's 2015 season highlighted (feel free to mouse over each dot to see which player/season each represents):
We can see a fairly clear relationship: as a player's Speed Score goes up, their average number of triples per 100 plate appearances goes up as well. This makes sense not only because triple rate is a part of Speed Score, but also because players who have higher scores tend to hit more triples. The trend is well-defined enough to the point at which you could draw a diagonal line on the left side of the blob of dots if it weren't for, well, Gattis. He is an outlier among everyone, with the slowest Speed Score to highest triple rate out of any qualified hitter in the past 15 years. It's not particularly close, either: Nick Castellanos, who hit six triples with a score of 2.7 last season, was the closest relatively slow hitter to have an above average rate of triples.
How rare is Gattis' accomplishment this season if we look further back in time? If we look at qualified batters going all the way back to 1950, only three other players have hit at least 10 triples while posting a Speed Score under 4.0: Dale Long (1955, 13 triples), Jerry Lumpe (1962, 10), and Harold Baines (1984, 10). Outside of incredibly rare occasions, this many triples is not something that is accomplished by players who run the bases like these men do.
While Gattis' improvement in this category came out of nowhere, there could be a few concrete reasons behind it: he's no longer catching, taking a lot of strain off of his legs (he said in early October "when you're catching, you don't have the same spring or bounce. It's a little different."), and he also spoke of simply being more aggressive right out of the box than he used to. That goes a long way. Batted balls taking friendly caroms off of outfield walls does as well, which can't be discounted: triples are inherently flukey at times, and this is probably a little bit due to his adjustments -- and a lot due to luck. An examination of video of all 11 triples from this past season reveal poor relay throws, strange bounces, missed catches that might have been judged errors, and fielders falling down.
All of this evidence makes the argument that a man as slow as Gattis should never be able to hit 11 triples in a season. And, truthfully, next season he might only get half that many. Yet that glosses over the reality that we're here, praising his name alongside Long, Lumpe, and Baines: the men who just kept running because the bounces were friendly, the fielders had holes in their gloves, and the throw to third -- it came late, or not at all.
