The secret behind the descent of Danny Salazar

Jason Miller

Small sample sizes are fun. Occasionally, they can be revealing, as well.

Danny Salazar set the world on fire after an unexpected call up last season, allowing just 18 runs in 52 innings and striking out 65. This season, it seems more like he's been setting himself on fire, allowing 28 runs in 40.2 innings at the major league level before being demoted. Was last season's performance a flash in the pan? Did his candle simply burn out? Will I run out of flame metaphors?

The answer to those questions is: probably not. Salazar has thrown with less heat this season, with an average fourseam velocity that dipped almost 1.5 mph. But that doesn't seem like it could explain such a big swing in results. As in many baseball cases, when a player has outlier performances at opposite ends of a spectrum, the truth is almost definitely somewhere in the middle.

-- IP K% BB% HR/FB ERA FIP xFIP
Salazar 2013 52.0 30.8% 7.1% 13.7% 3.12 3.16 2.75
Salazar 2014 40.2 25.5% 9.2% 14.8% 5.53 4.63 3.79
LgAvg 2013 -- 19.9% 7.9% 10.5% 3.87 3.87 3.87

Per FIP, it doesn't look like Salazar was particularly lucky last season. It does look, however, like Salazar was particularly unlucky this season, with fly balls turning into home runs, given the difference between his 2014 FIP and his 2014 xFIP. The fact that Salazar's xFIP is so close to a (2013) league average rate is also a red flag that Salazar has been burned by luck this season. So what gives? With most of the other numbers more or less the same, it seems that a higher walk rate is largely responsible for Salazar's increase in xFIP from 2.75 to 3.79. And here's where it starts to get weird.

Batted balls rates and home runs

Check out Salazar's batted ball rates for the last two seasons:

Salazar_batted_ball_rates

Salazar's ground ball rate stayed more or less the same, dropping from 34.4% of batted balls, but only to 33.3%. Fly balls, on the other hand, increased from 39.8% to 46.2%, a notable difference even in small samples (3 more fly balls in 11.1 fewer innings). Oddly, the additional fly balls seem to have come at the expense of Salazar's line drive rate, which dropped from 25.8% to 20.5%.

The increase in fly balls meant that although his 2014 HR/FB rate was essentially the same as his 2013 rate, he let up eight bombs in 40.2 innings, his HR/9 rose from 1.21 to 1.77. Although a lower line drive rate might suggest that hitters were having more trouble squaring him up, the increase in fly ball rate may be here to stay.

Batting average on balls in play

Like with fly balls and home run rates, it also seems from BABIP that Salazar has been unlucky this season. League-average BABIP has been .294 in both seasons, and while his .298 rate in 2013 was fairly normal, his .369 rate this season absolutely is not. There's some selection bias in that terrible pitchers are unlikely to qualify for the ERA title, but last year's worst qualifed pitchers per BABIP (Bud Norris, Joe Saunders, Doug Fister) had rates of just .333 or .332.

Thanks in large part to the fact that home runs aren't "in play," the extremely weak "correlation" between GB/FB and BABIP (0.147, based on a 2007-2013 sample of 30-1,000 IP pitcher seasons) works against ground ball pitchers, not for them. Last year, just 13 of 81 qualified pitchers had GB/FB ratios under 1. Only three of the 13 ranked in the top 30 for highest BABIP, and two of those three just barely made the list of 13 (Justin Verlander, 0.99; Jeremy Hellickson, 0.98). Salazar's 2014 GB/FB rate (0.72) would have been the second-lowest among 2013 qualifiers, so it seems on the surface, at least, like his 2014 BABIP is not just an extreme outlier, it's an unexpected one.

Add to that the fact that Salazar substantially lowered his line drive rate in 2014 and increased his fly ball rate, and the increase in BABIP becomes even more curious. If we expected a BABIP of 70% for line drives, 20% for ground balls, and 12% for fly balls, we'd expect a .265 BABIP for Salazar in 2014. That's somewhat lower than league average, not extremely higher. A difference this extreme is almost definitely a combination of something real and something luck. The luck is unlikely to continue, but it appears from the small 2014 sample that hitters were able to square up Salazar's pitches more frequently. Another reason to suspect that hitters actually were able to square Salazar's pitches up more frequently in 2014: it's easier to square up in-zone pitches.

Swing rates and walks

Danny Salazar hit the strike zone at an average rate in 2013, with only 46.0% of his pitches in the zone per Baseball Info Solutions (Pitchf/x has it at 52.3%). Whether that changed in 2014 depends on who you believe, as BIS has his Zone% staying pretty consistent in 2014 (45.3%), while Pitchf/x has it dropping from 52.3% to 46.0%.

Combine this average zone rate with a relatively high swing rate at pitches outside the zone (38.2% per BIS, 35.8% per Pitchf/x), and it should come as no surprise that a high percentage of Salazar's strikes that season came from swings and misses. His 14.6% SwStr% would have blown away the field in 2013 if he had qualified for the ERA title; Yu Darvish (12.6%) and Matt Harvey (12.5%) had little on Salazar.

The problem is that batters wised up. Salazar kept throwing balls at a high rate, but batters stopped swinging at them as frequently. From 2013 to 2014, Salazar's O-Swing% dropped precipitously. Here are Salazar's plate discipline numbers from BIS:

-- O-Swing% Z-Swing% Swing% O-Contact% Z-Contact% Contact% Zone% SwStr%
Salazar 2013 38.2 69.8 52.8 58.8 79.0 71.1 46.0 14.6
Salazar 2014 28.8 67.5 46.3 62.8 82.7 76.0 45.3 10.6
LgAvg 2013 31.0 65.5 46.4 66.6 87.0 79.5 44.7 9.3

Hitters weren't getting fooled the same way. They were still getting fooled (25.5% K%), but on their terms, not Salazar's. Check out his swing rates, from Brooks Baseball:

Salazar_swing_rates_2013_and_2014

Off the plate horizontally, not much appears to have changed; hitters never really swung at those pitches in 2013, and they continued to lay off in 2014. But look at above and below the zone, in particular. In the three blocks directly below the strike zone, swing percentage dropped from 53% in 2013 to 44% in 2014. Swing rates for the three blocks above the zone dropped even more precipitously, from 48% in 2013 to just 27% in 2014.

Note that while Salazar's SwStr% rate dropped quite a bit from 2013 to 2014, it was still good in 2014: his 10.6% rate would have tied for 14th among last season's 81 qualified pitchers. Still, the higher-than-average SwStr% may be the product of pitching outside the zone at a clip at which he just can't be successful, and it's clear that hitters stopped swinging outside the zone in general. What happens when you combine a high percentage of pitches outside the zone with a low O-Swing%? A high walk rate, that's what (9.2 BB% in 2014, up from 7.1% in 2013).

Pitch selection and sequencing

I don't have too much to add about sequencing, exactly, other than to note that Salazar's first strike percentage took a big dive in 2014 (from 67.3% to 57.1%, per BIS). That's at least partly a function of hitters swinging less frequently at pitches out of the zone, but not completely. As one of the few things that changed from one season to the next in a significant way, one wonders if the Cleveland Indians should stress first pitch strikes while Salazar is in Triple A.

Selection is another matter, however. In his eight starts this season, Salazar kept his fourseam fastball percentage more or less the same, but threw more sinkers, at the expense of his slider and splitter. Again from Brooks Baseball:

Year Fourseam Sinker Curve Slider Split
2013 64.14 4.02 0.00 12.53 19.31
2014 65.37 10.29 0.27 8.82 14.44

Part of why this is troubling is that it was Salazar's sliders and splits that produced the swings and misses at the highest rates:

Year Fourseam Sinker Slider Split
2013 13.98 5.71 16.51 22.02
2014 10.22 5.19 13.64 26.85

In fact, although batters were hitless against the sinker in 2013, they hit it at a blistering pace in 2014. Meanwhile, batting average against for his fourseam went way up:

Year Fourseam Sinker Slider Split
2013 .241 .000 .231 .231
2014 .312 .348 .235 .267

Throwing the sinker may have been meant as a solution before it itself became a bigger problem; he didn't throw any in his first two games this season, but relied on them heavily thereafter. In terms of results, however, the sinker has done extremely poorly. It seems to provide a very limited contrast to his fourseamer. Average velocity for the two pitches this season have been very similar (95.41 mph for the fourseam, 94.62 for the sinker), and the difference in vertical movement has been fairly small (9.92 inches for the fourseam, 7.5 for the sinker). Taking into account that the less-used split had been Salazar's main change-of-pace pitch, I wonder if increased use of the sinker is also at least partly responsible for hitters' increased success with the fourseam.

But as we've seen, Salazar's 2013 success and 2014 struggles may have much to do with pitches above or below the zone. Relying less on the split, in particular, could have been part of what changed that -- especially if the sinker just isn't working.

Maybe the slider and split have been effective for Salazar precisely because he doesn't throw them very frequently. But he did well with them last year, and in piecing together the puzzle of how to get Salazar back on track, one wonders if the Indians could have him try to throw both pitches a bit more frequently. At any rate, it looks like the sinker could be shelved.

Where to go from here

Something extreme is going on with the dumpster fire that has been Danny Salazar's 2014 season. It looks like the story of Salazar's day-then-night transformation is partly explainable by declining stuff, partly by pitch selection, and partly by luck. Salazar can't (or shouldn't) will an extra mph or two on his fourseam fastball, and it may be that his fourseamer won't ever be as effective as it was last season. But pitch selection is something that Salazar and the Indians can control, and to the extent that his struggles are luck, all the team can do is keep putting him on the mound.

Maybe Salazar has been tipping his pitches, or maybe it's just that by increasing his total percentage of fastballs (fourseam and sinker) all the way to 75.6%, he's become too predictable. Maybe, given the very small size of the samples, we can blame Salazar's struggles on an obscene and unexpected surge in BABIP. But it seems like there are only two things that Salazar can control.

The first is pitch selection. A return to using sliders and splits over 30% of the time could help restore some of the outside-the-zone swing-and-miss to his game. The second is location. If Salazar simply can't throw sliders and splits that much, or if he does and hitters continue to lay off, he's going to have to try pitching in the zone a bit more often.

. . .

All statistics courtesy of FanGraphs and Brooks Baseball.

Ryan P. Morrison is a featured writer at Beyond the Box Score, and co-author of Inside the 'Zona, a site on the Arizona Diamondbacks with a sabermetrics slant. You can follow him on Twitter: @InsidetheZona.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker