Dumpster Diving for Elite Relievers

Written by: Will Scharnagl (@WillScharnagl)

Follow us on Twitter! @Prospects365

In the 1989 film Major League, Rick Vaughn, played by Carlos Estevez (better known as Charlie Sheen), is an ex-convict who becomes a reliever for the Cleveland Indians. He’s known for his plus velocity on his fastball, but has an overall lack of command, hence the nickname “The Wild Thing” and he has no secondary pitches. The real life Carlos Estevez (the Rockies pitcher not Charlie Sheen), also sports premium velocity on his fastball, but unlike The Wild Thing, Estevez has excellent control, and a plus secondary pitch.

Given the short description I’ve just given of Estevez, he seems like he’d be, at the very least, a solid setup man, and possibly a high level closer, but let’s take a look at his StatCast profile.

estevez statcast

Yeah so scratch what I just said about being at least a solid setup man. The K% in the 79th percentile is obviously great, and FB velocity in the 98th percentile is elite, but the exit velocity (4th percentile), and hard hit% (13th percentile) are both among the worst in the league, and based on his xwOBA, he’s essentially an average pitcher. This is supported by his 3.96 ERA and 3.93 xFIP, which point to his current performance being mostly unaffected by luck. At this point you’re probably asking yourself why I’m wasting my time talking about a league average reliever, but as you dig deeper into Carlos Estevez, you find the shining star in an epidemic plaguing the MLB.

With any Rockies pitcher you have to address the splits, as Estevez’s 2.70 ERA and 1.20 WHIP on the road are much better than his 5.40 ERA and 1.59 WHIP at home, but he’s still the 4th option in a bad bullpen, and has only put up 0.4 fWAR this year, so he’s definitely considered an average pitcher.

In order to fully understand how Carlos Estevez became the pitcher he is today, it’s important to look back to his days as a prospect. In 2015, MLB Pipeline rated Carlos Estevez as the Rockies #30 prospect, and threw out “future closer” as a possibility for him. They also gave him a 70-grade fastball, but 40-grade on both his slider (which was a new pitch at the time), and his changeup. I want you to remember those grades as we move forward.

At 6’6” 275 pounds, Estevez is one of the biggest players in the MLB, and uses that size to generate elite velocity, yet for Fangraphs’ Pitch Value with fastballs among 171 qualified relievers, he ranks dead last at -9.7, tied with Touki Toussaint of the Braves. That former “40 grade” slider on the other hand ranks 25th, with a value of 4.7. It’s important to note that Pitch Value is pretty much like ERA for a pitch, and is not predictive like xFIP would be, so let’s take a look at his xwOBA as well. Hitters facing his FB have an xwOBA of .369, vs .194 against his SL. Obviously given the nature of these two pitches, such as the counts they’re thrown in and how often they’re thrown in the zone, fastballs will almost always have a worse xwOBA, but a difference this large cannot be supported by just that fact.

There’s a few reasons why Estevez struggles, and it’s not as simple as his fastball being bad. For one, aside from velocity, Estevez’s fastball really isn’t a special pitch. He has average spin rate, below average for his velocity, and both the horizontal and vertical movement are below league average. What this means is that while it comes in harder than most, his fastball is fairly easy to read out of the hand. While this would’ve been fine 20+ years ago, hitters now have no problem hitting 100 MPH fastballs, as Max Muncy shows here, demolishing Estevez’s fastball.

estevez gif

Another issue for Estevez is his control over command style, as he throws a lot of strikes, but doesn’t necessarily have the best command. Believe it or not, Estevez actually throws too many strikes, making him too hittable. Even relievers with elite walk rates, such as Sam Dyson, throw less pitches in the strike zone than Estevez does. This plot shows the location of all of his fastballs this season, and the lack of a trend is concerning.

estevez spray chart

When Estevez gets the ball up and elevates his fastball, it’s actually an effective pitch, and his overall 31.6 Whiff% is very good, but he leaves the ball out over the plate too much, and he gets punished like Josh Bell does here on this pitch which was right down the middle.

josh bell gif

As mentioned before, on his fastball, hitter’s have an xwOBA of .369, which for reference is the same as Alex Bregman’s xwOBA. He’s given up a total of 38 hits off his fastball, with 57.9% of them going for extra bases, and 9 of them leaving the ballpark.

Now let’s compare that to his slider, which hitters have a .194 xwOBA against. While hitters are hitting .233 against the slider, all 14 of those hits are singles. He hasn’t allowed a single XBH this season off his slider, and hitter’s have a putrid 83.8 MPH average exit velocity. This is a huge improvement from where he was in 2017, where his xwOBA on the slider was .293. Part of the reason he spent 2018 in AAA was to work on his slider, and it’s clear that he made the necessary changes to become more than a one pitch pitcher. His slider isn’t anything special in terms of spin rate or movement, with both being slightly above average, but when you factor in the velocity it’s a very difficult pitch to hit. He’s also getting whiffs at a rate of 34.9%, like this one here from Carlos Gomez.

Where this all gets confusing is when you look at his pitch usage. Despite the numbers pointing to his slider being a plus pitch, Estevez only throws his slider 27% of the time. On the contrary, while his fastball is statistically one of the worst pitches in baseball, he throws it 69.5% of the time. Of all the relievers with negative pitch value fastballs, only Victor Alcantara (76.0%) and Kenley Jansen (89.6%), has a higher usage rate. Both are far better than Estevez (-2.0 for Alcantara and -0.4 for Jansen), and both are also easily explained. Unlike Estevez, Alcantara doesn’t have any other good pitches, as his slider and changeup are both negative in value as well, and Jansen is a guy who’s always had a great fastball, but is struggling to adjust with his slight loss of velocity and spin. He’s getting much less movement on his cutter than he did in previous years, but he’s also experienced some bad luck as well, as his wOBA on that pitch is 26 points higher than his xwOBA.

There’s a few points that I’m trying to illustrate here. The first one is that fastball velocity isn’t everything, which is something P365’s Trevor Powers wrote a great article on this topic earlier this year. You should definitely check it out, as it goes way more in depth on this idea that the fastball grading system is all wrong, and Estevez is proof of that. While Estevez has premier velocity, everything else about his fastball is average or worse, so when he throws it as often as he does, and where he does, it’s bound to get hit.

The other point I’m trying to highlight is how horribly the Rockies have misused Estevez, which seems to be a common theme with the Rockies. Just because a guy can hit 101, and consistently sits 97+, doesn’t mean he has a good fastball, and it especially doesn’t mean he should be throwing the pitch 70% of the time. As is right now, Carlos Estevez’s best pitch is his slider, and no matter how hard he can throw his fastball, he should be throwing his slider most of the time. There’s absolutely no reason that the pitcher who currently has the worst fastball should be throwing that pitch for almost every pitch of an at bat.

I honestly believe that Carlos Estevez is an extremely talented pitcher, but not for the reasons he was once considered to have future closer potential. I think he’s a great strike thrower, with a nasty slider, and I think with the right direction he could actually reach that potential that scouts saw in him. Obviously baseball is a complicated sport, but I think a change as simple as adjusting his pitch usage would make a massive difference for Estevez.

I think Estevez has very similar stuff to Diego Castillo, who throws his slider 52.6% of the time, and even that seems low to me. If Estevez got his slider usage up to 55%+ (it currently sits at 27.0%), and mixed in his changeup a bit more (currently 3.6%), while keeping his fastball elevated when he uses it, I genuinely think he could be one of the best relievers in baseball. If I was the GM of any MLB team I would be looking to trade for him and make those changes myself.

As I mentioned earlier on, Estevez isn’t the only pitcher like this, who throws his fastball way too much, he’s just the most polarizing case. He’s not even the only one with premium velocity, so let’s talk about another guy who’s very similar, his former teammate Miguel Castro of the Baltimore Orioles.

castro statcast

Castro’s statcast profile is actually more inspiring than Estevez’s, but the rest of his stats leave little to be desired. In 236.2 career innings he has a career ERA of 4.37 and a career xFIP of 5.09, and career K% is only 17.0, which certainly doesn’t compliment an 11.8 BB% very well. He’s improved his K% to 21.9, but he’s also raised his BB% to 12.3. His ERA and xFIP are an unexciting 5.05 and 4.86 respectively, but as with Estevez, I think Castro is actually a very talented pitcher, whose skill set has not been correctly put to use.

If you remember Miguel Castro’s name vaguely, it’s likely because he was a part of that blockbuster Troy Tulowitzki trade, which sent him and Jeff Hoffman to Colorado. Castro actually showed promise as a very young starter, but the Blue Jays chose to convert him to a reliever and rushed him to the big leagues, as he made his debut at just 20 before he was shipped off to the Rockies. Castro was actually featured on that same 2015 MLB Pipeline Rockies Top 30 with Estevez, but Castro was ranked much higher. Sandwiched right between 2018 4th place Cy Young finisher Kyle Freeland and 2018 8th place MVP finisher Trevor Story at #10, Miguel Castro clearly hasn’t had the success that those other two have. MLB pipeline also gave him a 70 grade fastball, with Fangraphs giving him one as well, with an 80 future value on top of it. In terms of his secondary pitches, MLB Pipeline gave him a 45 on his slider, and a 55 on his changeup, while Fangraphs gave him a 50 on his slider with a 60 future value and a 40 on his changeup with a 45 future value.

Now let’s fast forward to 2019, where Castro has one of the worst fastballs in the league (despite velocity in the 96th percentile), and not one but TWO plus secondary pitches. Again, Fangraphs’ pitch value stat isn’t the end all be all, but Castro’s fastball ranks 7th worst among relievers, at a measly -7.6. His slider, on the other hand, ranks 36th at 3.5. His changeup ranks 17th at 3.2. His xwOBA for all three pitches lines up with these stats, as his fastball has given up a .381 xwOBA while his slider and changeup have given up a .168 and .212 xwOBA respectively.

One of the biggest improvements that Castro has made this season is adding velocity across the board. While it hasn’t helped his fastball, it’s certainly improved the other two offerings. Additionally, he’s added almost 150 RPM to his slider, which now sits at 3,027 RPM on average. That average is 5th among all sliders in the MLB, and one of only 11 pitches (5 sliders and 6 curveballs) that averages over 3,000 RPM. This leads to a slider with plus velocity, as well as above average movement both vertically and horizontally. He usually throws it 87-89 MPH, but sometimes takes a bit off it, and throws it in the 82-84 MPH range with slurve-like action. These changes have allowed him to upgrade his slider from a decent pitch, to an extremely good pitch, and allows him to make hitters look silly, like he does with Mike Trout here.

miguel castro gif 2

Despite the levels of success he’s had with his slider this year, Castro’s changeup might actually be his best pitch, as he commands it very well and is able to throw it against both lefties and righties. He also ranks fairly highly in spin rate, at 24th among changeups. He also ranks 6th in horizontal movement and 2nd in velocity at 90.5 MPH behind only Noah Syndergaard. He can sometimes crank it up even higher, as he does here, hitting 92 to get Eric Hosmer swinging.

miguel castro gif

While he mostly throws the pitch against lefties, as I mentioned, he can use it against righties as well. He’s been extremely successful in the small sample size, allowing a .136 xwOBA on the 34 changeups he’s thrown against righties this year, including the pitch above that struck out Trout.

His fastball, on the other hand, just really isn’t a great pitch. He throws a sinker with average spin rate and slightly above average movement, but the only thing that really stands out is his velocity, which ranks 15th among sinkers. While he generates a lot of ground balls, which is the goal of a sinker, he also gets hit hard a lot. Really hard. A lot. Hitters have an average exit velocity of 90.6 mph versus the pitch this season. Even when he misses his target by mere inches, Castro’s fastball often gets punished. Rowdy Tellez wasn’t fooled here.

miguel castro gif 3

Typically, a sinker has to be low in the zone and induce ground balls to be effective. When you hang it, they bang it. That’s what happens here, and Willy Adames hits an absolute bomb to walk off for the Rays.

miguel castro gif 4

Despite all of this, Castro is still throwing his sinker 55.1% of the time, and is getting crushed because of it. Like Estevez, I believe this is a case of a guy who’s been told that he’s had a good fastball because of how hard he throws, so he continues to throw it even though he gets hit. I’m really not quite sure how the Orioles and Rockies have the same stats available, yet continue to let him throw this pitch over and over again. I think Castro should be throwing his sinker no more than 25% of the time, and should be throwing his slider and changeup both more than 35%. I think these changes could help Castro reach the potential that evaluators once saw in him.

What makes me even more excited about Castro is the potential he possesses. This is a guy who was a former top prospect, in an athletic, wiry 6-foot-7 frame, who has two plus pitches, and is only 24 years old. He’s always struggled with control, and that’s no different this year, but with stuff as good as his is, you can be extremely successful with a walk rate in the 10-14% range. If you need examples of this, look at Aroldis Chapman, Matt Barnes, Reyes Moronta and fightin’ Amir Garrett for reference. I honestly wouldn’t even mind stretching Castro out and trying him as a starter, as long as the fastball usage stays low. The Orioles aren’t winning any time soon, what do they have to lose by trying out a guy like Castro in the rotation?

Miguel Castro and Carlos Estevez obviously aren’t the only two guys in baseball like this, they just happen to be the most interesting in my opinion. The last guy I’m going to talk about is certainly less interesting, but that fact could actually make him the most attractive option for any contending team. That guy is Luis Garcia of the Angels.

luis garcia statcast

Looking at Garcia’s stats, you’d think he’s one of the worst relievers in the league, and to be completely honest, he likely is. He’s currently in the midst season where he has a 5.37 xFIP, and has been worth -0.5 fWAR. He’s also 32, has never posted an xFIP below 3.50, and despite fastball velocity in the 94th percentile, his fastball might actually be the worst of anyone mentioned in this article, as he gives up a .513 xwOBA on fastballs. For reference on what a good xwOBA is, Trout leads all players at .465, and only eight players are above .400. If you’ve noticed the trend yet, you’ve probably guessed that Garcia isn’t fazed by this, as he still throws his fastball 50.4% of the time.

The reason as to why I’ve included Garcia, as there were other better and more talented options, is because of just how good one of his secondary options is. Garcia is one of only 42 pitchers in the MLB who throws a splitter. It just so happens that his splitter is very, very good. Since he started throwing it in 2017, Garcia’s splitter has allowed an xwOBA of .229 in 2017, .145 in 2018 and it’s all the way down to .095 in 2019. He’s also posted a whiff rate above 40% on his splitter in each of the last three seasons. After watching Steve Wilkerson trying to hit this 87 MPH splitter, it’s easy to see why it’s been so hard to hit.

luis garcia gif

His slider is also very good, as he’s never allowed higher than a .250 xwOBA against it, and has been below .200 3 of the past 5 years. I won’t go in depth as I did with the previous two pitchers, as I probably sound like a broken record at this point, but there’s absolutely no reason for a guy with an incredible splitter, a very good slider, and a really bad fastball to being throwing a fastball for 50.4% of his pitches, and a splitter for 15.8%. I think any contending team willing to take a risk and make these changes would be rewarded with a very cheap, solid addition to the back end of their bullpen.

These guys obviously aren’t the only three of their kind, so I put together a list of relievers (including the 3 mentioned already) who I believe would greatly benefit from throwing their fastball way less, and one or both of their secondary pitches way more. These guys range from closers who are already studs, to guys you may have never even heard of. I included their xwOBA, Whiff% and usage for the three main pitches, as well as their ERA and xFIP for a frame of reference.

pressly chart

Like I said before, I won’t go into too much depth on any of these guys to avoid beating a dead horse, but I’ll touch on a few that stick out. Osuna and Jansen still have very good fastballs and are obviously elite relievers, but their secondary pitches are a bit underused. Jansen rarely uses his slider even though it’s a great pitch, and Osuna’s slider and changeup are both borderline unhittable. He should definitely throw this pitches a bit more.

Kelvin Herrera is a guy who’s had a great fastball in the past but has seemingly failed to adjust to the changing game as the viability of the pitch has diminished. His secondary pitches are both solid; if he threw them both more, he could potentially get back to the pitcher he once was.

Chris Devenski really surprises me because this seems like something the Astros would be on top of, but despite his fastball getting rocked this year, they still have him throwing it 44.5% of the time. He’s utilizing his slider and changeup way less despite the fact they’re superior pitches.

Corbin Burnes is definitely a strange case, with lefties basically taking batting practice against him every time they face him. The difference between his ERA and xFIP might be the biggest I’ve ever seen, and while he definitely needs to develop his changeup to consistently get lefties out, I think throwing his fastball less is a good place to start in terms of improving his overall game.

I’m sure there are a few concerns about increasing the use of secondary pitches. The first has a lot to do with relievers often being unable to command any pitch in their arsenal outside of the fastball. The other would be the more you throw a pitch, the easier it becomes to hit, as hitters will adjust to it with repetition. That’s certainly a possibility, and if pitching better was as simple as throwing more sliders everyone would do it. But here’s a few of the relievers who had the highest jumps in slider usage, and how their performance varied.

pressly chart 2

Obviously these are slightly mixed results, but none of their walk rates got that much worse, and none of their sliders got significantly less effective, and some even became more effective. While a slider is thrown outside the zone more than a fastball would be, every one of these guys saw their O-Swing% increase with their slider usage. Every single one of these pitchers also saw their xFIP decrease, indicating a better overall performance. Matt Wisler is an especially interesting case, as he has increased his slider usage by a significant margin every year since he’s debuted, starting at 23.5% in 2015, all the way up to 70.2% in 2019. Despite that constant increase, his highest BB% and xwOBA against his slider came in 2017, when he threw about half as many of them that he has this year.

Am I saying that throwing their secondary pitches more will make every guy mentioned in the article an elite closer? Of course not. There’s a good chance it will make them better pitchers, but it’s also possible it will have little to no effect, or could even make them worse. What I do think is absolutely true is that the risk of trying to adjust their pitch usages is basically zero, so I see no reason that teams wouldn’t at least try it. So many pitchers are misused, and pitch usage is such an easy change to make that it baffles me that a guy like Carlos Estevez is still throwing 70% fastballs.

I also think that overall, fastballs as a whole are overrated. Don’t get me wrong, there are pitchers with great fastballs, but almost every MLB hitter can hit a fastball, no matter how hard you can throw it. In an article posted in January, Alex Chamberlain of FanGraphs put together this chart (in this article), attempting to quantify ERA and xFIP for individual pitches. This is what he came up with.

alex chamberlain fip table

As I mentioned earlier, a big part of this is the way a fastball is used, and the counts that it’s thrown in. Fastballs are thrown in the strike zone more than any other pitch, and are usually the go-to pitch in hitters counts, but that’s not enough to warrant a three run difference in FIP and xFIP. I would argue that a majority of MLB pitchers, both starters and relievers, should be throwing fastballs way less.

In the waning minutes leading up to the trade deadline, the Astros traded for Joe Biagini and Aaron Sanchez. Everyone loves to talk about how the Astros “fix” guys, and I expect them to do the same with these two, but it’s really as simple as increasing pitch usage. Sanchez and Biagini both have really good secondary pitches (.234 xwOBA w/CB for Sanchez. .261 xwOBA w/SL and .216 xwOBA w/CU for Biagini), and both have curveballs with elite spin rates (2875 RPM for Sanchez and 2793 RPM for Biagini). Don’t be surprised when these guys start throwing these pitches way more with the Astros, and they subsequently become much better pitchers. The Astros obviously do a lot more than just tell guys to throw certain pitches more, but it’s definitely a integral part of it. Ryan Pressly, for example, has seen his fastball usage drop from 55.3% in 2017 with the Twins, to 35.5% in 2019 with the Astros.

I only talked about relievers in this piece, as I think it’s much more likely that you’ll make a Ryan Pressly out of Miguel Castro than a Gerrit Cole out of Aaron Sanchez, but this sentiment stands true across the league amongst both starters and relievers. To put it as simply as possible, your best pitch should be the one you throw the most often. I don’t think that’s a very controversial statement, and I’m not sure why the Astros are seemingly the only organization in baseball that truly realizes it. None of the three guys I discussed are big names, but I think I’ve made a strong case that one small change could propel any of them straight into the backend of a bullpen. For some reason, that hasn’t happened yet.

Maybe I’m wrong, and all these guys just aren’t good pitchers. Maybe the sample sizes for their secondary pitches is too small, and maybe it’s not as easy as simply adjusting a pitcher’s pitch usage. But I believe unless those changes are made, pitchers like Carlos Estevez, Miguel Castro, Luis Garcia and many more will have their careers shortened because their organizations won’t correctly utilize data that is publicly available for all to see. That’s pretty disappointing in my eyes. Are these guys ever going to be as dominant as Ryan Pressly? Probably not, but they undoubtedly have a lot of potential that is currently being held back by something along the lines of willing ignorance. Hopefully organizations throughout baseball will figure this out before it’s too late for these relievers.

Follow P365 contributor Will Scharnagl on Twitter! @WillScharnagl

Follow us on Twitter! @Prospects365

Featured image courtesy of photographer Justin Edmonds and Getty Images

Leave a Reply