More on Streaky Hitters

In my introductory article on streaky hitters, I used Anthony Rendon as a narrative hook. I wish I had waited a couple of weeks so I could have used Scooter Gennett instead. Gennett hadn’t hit a home run since April 11 before he smashed four on Tuesday, and based on my criteria, Gennett has now officially snapped the cold streak he had been on since May 12 which featured just one hit over his most recent seven games. More to the point, now that I’ve defined what I consider a streaky hitter to be, I wanted to look at some multi-year trends to see if I could identify certain hitters as consistently streaky or not streaky and hopefully pinpoint some specific skills that lead to those tendencies.

Before I throw some leaderboards at you, I’ll mention that I restricted my pool of players in a few different ways. As I mentioned in the previous article, I selected 300 plate appearances as the minimum to qualify a player in an individual season. That creates some awkwardness since I ended up calculating streaks per season rather than streaks per X plate appearances or something more logical, but since my code is a bear to run, I’ll just go with it. I doubt an improvement there would result in a tangible difference in my conclusions.

Meanwhile, I opted to include players only if they had four or more qualifying seasons since 2008. And finally, I restricted the players for the following to leaderboards to include only players with at least .340 wOBAs from 2008-16. That is the glossary threshold for an above-average hitter, which I’m using as a proxy for fantasy-relevant.

With that said, let’s start with the hitters who had the most streaks per season since 2008.

The names on that list like Miguel Cabrera and Mike Trout are all you need to disabuse the notion that streakiness might be a reframing of “bad” or “flawed” in some manner. But before you jump to the conclusion that all good hitters are streaky, peruse the streak per season trailers.

That list may not be quite as visually impactful, but Troy Tulowitzki, Jose Bautista, and David Ortiz were hardly fantasy slouches at their peaks, which definitely overlapped with the 2008-16 window I selected.

At a glance, I did not notice anything that stood out as a likely underlying contributor to streakiness. Both the leaders and trailers include some lefties, some righties, some power hitters, and some walkers. The trailer list features several players whose performance level changed dramatically during the window, in particular with Bautista and J.D. Martinez breaking out and Yasiel Puig breaking bad. However, the leaderboard includes Josh Donaldson, who is the poster boy for unexpected breakouts.

As such, I felt I needed to look beyond the extremes and do some comprehensive testing. For that, I ran correlations of various statistics against streaks per season for the 111 hitters who qualified. Batter handedness, wOBA, strikeout rate, and walk rate all produced low correlations, but home run rate definitely did not.

That 0.71 r-value is pretty high—positive correlations scale from a low of 0.00 to a high of 1.00. This is something that did come up in my previous research on the topic since Kris Bryant had the second-most streaks in 2016. Of course, it makes no sense to pass over Bryant or the long-term streak leaders like Cabrera and Trout because they hit home runs and may be streaky because of it. They are obviously really good at baseball, and home runs are one of the categories in standard roto. But even if streakiness is heavily tied to power and therefore not a reasonable criteria to use to evaluate a player’s draft-worthiness, I do still wonder whether it is possible to realize streaks are happening in real-time and make start-sit decisions based on them. I’ll look more into that next week.