Felix Hernandez’s Other Weapon

Wednesday afternoon in Tampa Bay, Felix Hernandez wrapped up a complete-game shutout. It was just his second in the past three seasons, but you shouldn’t let that fool you. He’s one of the best pitchers in baseball — pitching at his peak — and he’s one of those guys where a single underwhelming outing is enough to make people wonder if something might be wrong. People don’t wonder very much these days, though. The two greatest challenges are hitting Felix and not taking Felix for granted.

For the most part, Felix is well understood. Among all of his strengths, the changeup stands out for its movement, for its velocity and for its location. It’s one of the best changeups in the world, and Felix is comfortable throwing it in any count to any hitter. It can be difficult to distinguish from his fastball, which is faster by only a handful of ticks. Everyone, therefore, is familiar with Felix’s main weapon. Fewer people think about another weapon. Felix’s slowest weapon. But he’s going to it with increasing frequency.

When he was a prospect, and then a rookie, Felix dealt what was referred to as the “Royal Curve.” He went to the curve once every five pitches, in part because it was good, in part because he’d been discouraged from throwing his slider and in part because he didn’t yet have an effective changeup. But, over time, the curve started to fade. It didn’t fade completely, but by 2008, it was Felix’s fourth pitch. First, the slider took over. Then, the changeup took over. Even as recently as 2012, Felix threw one curve for every seven non-curves. He was doing just fine.

Skip ahead to the present. Against the Rays, a quarter of Felix’s pitches were curveballs. His season rate is about 20%, following last year’s bump to 16%. These days, Felix has almost eliminated the slider. He’s even throwing fewer changeups. The change is still the thing that makes Felix Felix, but the curveball is back, and here you can see it in action. These .gifs were chosen intentionally! There’s a fun fact coming, and these .gifs are clues.

A swing and a miss:

A swing and a miss:

A take and a strike:

A take and another strike:

Felix threw 24 curves Wednesday. Seventeen were strikes. One was put into play. The pitch was giving the Rays’ hitters fits. They’re hardly alone.

It’s time for a simple thought experiment: We know, generally, whether a pitch ends up in the strike zone. If a pitch is in the zone, a pitcher would like for it to not be swung at — therefore, the pitch goes for a strike, almost all of the time. There’s no negative cost. On the other hand, if a pitch is outside of the zone, the pitcher wants a swing. Without a swing, that pitch is a ball. With a swing, you get some sort of strike, or perhaps weak contact. You know the Z-Swing% and O-Swing% stats we have available. The smaller the gap between Z-Swing% and O-Swing%, the more deceptive a pitcher or pitch probably is. A small gap means hitters are left with effectively terrible discipline.

So let’s look not at Felix overall, but at just Felix’s curveball. Let’s look at Felix since the start of last season, when he increased his curveball usage. What follows uses the FanGraphs plate-discipline stats you can find on a pitcher’s player page.

When Felix has thrown a curveball in the strike zone, how often has it been swung at?

• 33%

That seems low. What about the opposite? When Felix has thrown a curveball outside of the strike zone, how often has it been swung at?

• 40%

Read that again. Since the start of 2014, Felix’s curveball has generated a higher swing rate out of the zone than in it. Which makes it a tremendously effective pitch, because the would-be strikes are strikes, and the would-be balls are strikes. With the curve, Felix is stealing strikes when he wants, and he’s also getting guys to chase when he wants. It’s exactly what you want out of a pitch. Which helps explain why, on a rate basis, Felix’s curve has been every bit as valuable as his changeup. The changeup is still the better pitch, but because people are thinking about that changeup, the curve has been outstanding.

Naturally, different places will give you slightly different numbers. They look a little different on Brooks Baseball, and they look a little different on Baseball Savant. But the point remains no less valid: The swing-rate gap is impressively small, if not reversed. I don’t know how many other pitches are like this, but it can’t be very many.

Felix seems to have increased confidence in the curve. Above, he shook to the pitch a couple times. From Brooks Baseball, you can see a change in pattern. This is a plot of average vertical location, relative to the middle of the zone. This year, Felix has been better than ever at keeping the curveball down at or below the lower edge:

Whether that’s a sign of better command, I don’t know, but the probability is high. The curve has become Felix’s go-to breaking ball. He’ll use it in any circumstance, against righties and lefties, and this is one reason why Felix’s ground-ball rate is trending north. His curve is a tough pitch to drive when he doesn’t hang it, and he decreasingly hangs it.

I wouldn’t say this has unlocked a new level of performance or anything. Felix is excellent now, but he’s been excellent for a while, and his numbers are similar to where they were in 2012. He was and is an ace, and though his ERAs have dropped, that’s hardly conclusive. I’d say this has allowed him to sustain, to stay around the same level despite more opponents getting more looks in more at-bats. Even the best players need to stay ahead of the game as best they can, making little adjustments so that opponents can’t so easily prepare. Last year, Felix threw a lot more changeups, and he sharpened his curve a little. This year, he’s throwing more curves in changeup counts, and he’s keeping the curveball farther down. It’s all for the cause of staying Felix Hernandez. On the surface, he’s a familiar workhorse. Underneath, he’s getting better to stay the same.