The Anatomy of Jedd Gyorko’s streaky 2017

Jedd Gyorko has had a weird season so far. If you read this blog often, you’re probably a Cardinals fan, which means you’re well aware of that. If not, this graph says it all:

This is a rolling 15-game average of Jedd’s wRC+ throughout his career, provided by the good folks at Fangraphs.com. wRC+ aggregates all of a hitter’s contributions at the plate into and spits out a park and league adjusted single number to describe how good or bad a hitter was overall.

Jedd started out the year red hot, and just kept hitting better and better. This eventually culminated in holding a 184 wRC+ for the season on May 6th, 86 plate appearances into the season. That means he was 84% better than league average at the plate over that time frame. If he could have somehow kept that up the whole year, it likely would have resulted in an MVP award. Since then it’s been all downhill, recording a 76 wRC+ in the next 132 plate appearances. That means he’s been 24% worse than league average since that point.

That’s averages out to a 118 wRC+ going into Friday’s contest, which is solidly above-average. If he can hold that the rest of the year, it would represent a career high for him, so that sounds great. There is worry among fans though, that he’s in a bit of a free fall. Some think he’s getting exposed in a full-time role. Let’s see what the numbers say.

First off, let’s look at Gyorko’s strike out and walk numbers throughout the year.

Pitchers mostly have control over strikeouts and walks. They can influence contact quality, but to nowhere near the same level, and it’s hard to be sure without a really large sample. If pitchers figured something out, you’d expect to see it in Gyorko’s strikeouts or walks. That doesn’t seem to be the case though. Continuing with May 6th as the arbitrary dividing line, here’s his K% and BB%:

Nearly identical in walks and strikeouts. Non-contact wOBA is also shown, and that did change. Non-contact wOBA is the non-contact portion of wOBA, which is strikeouts, unintentional walks, and hit-by-pitches. For context, league average is .200. Gyorko is below average over both stretches, but he was .017 points worse over his slump period. However, that’s actually only because Gyorko was hit by a pitch during his hot streak, but wasn’t during his slump.

So no, Gyorko’s struggles haven’t come in non-contact situations. His slump is due entirely to worse results on-contact. Let’s look at a 15-game rolling average of his BABIP (Batting Average on Balls In Play) and ISO (isolated slugging, or slugging percentage minus batting average, to get a measure of extra base hit ability):

Both peaked around the same time that his wRC+ did, and have plummeted since. The BABIP was always going to regress. No one in the history of the game has a true BABIP ability of .400, let alone .500. And if they could, it wouldn’t be someone like Jedd, who is slower than average and tries to hit as many fly balls as he can, which have the worst chance of being an in-play hit.

That’s OK, because hitting for power can make up for that, and that’s the game Gyorko plays. The ISO has plummeted as well though. On it’s own, that doesn’t spell doom. While it’s not as noisy as BABIP, ISO varies a lot over the course of a season. Baseball is a game of inches, and the difference between a homer and a fly out to the warning track is just fractions of an inch in terms of where the ball makes contact with the bat.

So let’s dig into Gyorko’s contact quality. Here’s his radial chart for the 2017 season, provided by the statcast data hosted at BaseballSavant.com.

If you haven’t seen this image before, the protractor-shaped image above is used to represent any batted ball by Exit Velocity and Launch Angle. The six shaded regions represent the six qualities of contact, and each dot represents one of Gyorko’s batted balls in 2017. If that’s confusing to you, don’t worry, I was too at first. Check out this post for a full breakdown of the six types of contact quality.

While this is a great visual, it lacks context. We want to know how Gyorko’s quality compares to league average, and it would also be nice to know how productive each category is. For this situation, it would also be helpful to see a split between what his hot streak looked like compared to his current slump. Here’s what all that looks like:

When Gyorko was hot, he was hitting a well above average amount of “barrels”, the best type of batted ball. He was getting very little “solid contact” though, which is basically a border surrounding barrels. You have to wonder if that was the first part of the regression. When one player has a lot of barrels or solid contact, but not that much of the other, it seems likely that the two would be more even going forward. Despite barrels representing much more area on the radial chart than solid contact, they’re very comparable in terms of the average likelihood of one occurring.

Also notice the “weak” category, which is any batted ball under 60 MPH, regardless of angle. Those batted balls actually perform at an above-average rate, likely because the defense usually isn’t positioned in preparation for batted balls under 60 MPH.

Regardless, it’s hard to believe it’s a good sign for a hitter. Like the connection between barrels and solid contact, it seems believable that if a hitter has a lot of weak contact, going forward some of that will turn into “under” or “topped” batted balls, which are the two below-average performing batted balls. That looks like the case with Gyorko, who saw a big reduction in weak contact, and a big increase in getting under the ball.

Next up, we’ll look at Gyorko’s xwOBA. xwOBA is also brought to you by BaseballSavant.com, and replaces the on-contact portion of wOBA with what it “should” have been, based on the average performance of each of the player’s batted balls. This doesn’t take into account foot speed or how often a player is shifted against, but nevertheless Craig found that xwOBA was more predictive of future wOBA than wOBA itself. Here’s how Gyorko grades out during both streaks:

During his hot streak, Gyorko out-performed his contact quality by .060 points of wOBA. For context, that’s the 32nd biggest negative difference of 317 players with at least 40 at bats over that time frame, just outside the top 10%. Still, .399 is an impressive xwOBA, 29th over that time frame, or among the top 9% of the league. Gyorko’s hot streak wasn’t just luck, he was also just plain tearing the cover off the ball.

On the flip side, his cold streak was entirely earned, with the wOBA matching the xwOBA. What does this mean? Again, it doesn’t have to mean anything. Baseball is a game of inches, and hitting is one of the most obvious examples of that fact. Let’s push forward though, and see if pitchers are attacking Gyorko any differently. First off, we’ll look at a heatmap of where Gyorko has been pitched in 2017, provided by Baseball Savant. On the left is the hot streak, on the right is the cold streak:

The differences are very minor. Generally, Gyorko is pitched a little bit to the outside portion of the plate, but with the highest density middle-middle. That was the case during both his hot and cold streaks. But that’s just part of the equation. What about pitch selection? Are pitchers throwing him a different mix of pitches? Here’s a pie chart by pitch type during the hot streak, again courtesy of Baseball Savant:

And here’s the same thing, but for his slump:

Again, strikingly familiar. Jedd’s Zone% is also almost the same over this stretch as well, so pitchers aren’t avoiding the zone more often.

Jedd is walking and striking out at the same pace. If he wasn’t, it would be a good sign that either something was wrong or pitchers were successfully adjusting to him. While there’s a .160 point difference between his wOBA in both time frames, xwOBA brings it down to a .100 difference. That’s still very significant, but at the same time, he’s getting pitched virtually the same way. All that appears to be going on here is a great example of the contrast between a hitter being as locked in as he can be (and have some things goes his way) vs. something being a little off.

It’s always fun to find something more concrete, but “the season is just full of ups and downs” is a valid answer in baseball. Hitters talk about this all the time. Sometimes the ball looks like a beach ball, other times a marble. It’s just part of the game. Going forward, he’ll almost certainly be somewhere in-between. Baseball is a game of streaks. Let’s hope Jedd is about done with this one.