Brad Peacock Might Finally Have An Answer For Lefthanded Hitters

Brad Peacock has been a lot of things in his career. A near-afterthought as a draft pick, a top-50 prospect, a starter with an unsightly walk rate, injured with bone spurs near his spine, a nifty reliever, a great swingman…they’re all apt descriptions of the various phases he’s experienced. This year, he’s back in the rotation for the Astros as they’re primed for another season of success. Whether in an upswing or a downturn, though, lefty hitters have always been a bugaboo for him.

He came into 2019 sporting a wOBA against lefthanded hitters that was 29% worse than the average righthanded pitcher since he came into the league. I wanted to round that to 30% for the sake of convenience, but then I checked how Peacock has performed against same-handed hitters compared to other righties in the same time frame. Get this — he’s been 29% better at limiting quality contact than his peers. Talk about wild!

But back to how he fares against lefties. He has a three-quarter arm slot and his best pitches — his fastball and slider — aren’t necessarily ideal to go to battle with against them. With that in mind it’s understandable that he could have his limitations; that his ultimate effectiveness might be limited by the same things that got him to the Majors. Just consider what he’s thrown when facing a lefthanded hitter through his career:

peacock v lefties career
There are a couple clerical-type things to note here. One is that we could merge Peacock’s two-seamer (FT) and sinker (SI) into 229 total pitches. Even if we want to split hairs and distinguish them from each other, odds are infinitesimal that Peacock was trying to throw two different pitches. It’s just a pitch classification quirk. The other is that anything thrown by a pitcher less than about 10% is something we don’t need to sweat. It probably isn’t offered enough to really buy into the data it produces. So we can eliminate the two-seamer as a whole to get a closer look at how Peacock really approached lefties from the time he debuted in MLB through last year.

That leaves us with a four-seamer, a knuckle curveball, a slider, and possibly a changeup. Based on usage, let’s draw our focus tighter to just the breaking balls. Here’s where he located them:

peacock HHEEAATT

That’s from the catcher’s perspective. He alternated location with each, sometimes trying to get chases low and in and sometimes trying to backdoor them over the plate. They’re distinct pitches and that’s usually good. But in this case, maybe not so much. Peacock’s curveball has always dropped between six and 11 inches more than his slider. Meanwhile, its side-to-side movement shrunk from a four inch difference with the slider to less than an inch in the last two-plus seasons. Locating the pitches in different spots could have given lefty hitters the chance to distinguish the break on each pitch, line it up and drive it, or spit on it and wait for a fastball. With all that deep yellow outside of the zone above, there was probably a good chance for a walk, too.

In his first start of 2019, he changed things against lefties. To the pitch mix!

peacock peeitch meeix
Ok, so let’s ditch the two-seamer/sinker amalgamation. Let’s also ditch the single cutter (FC), because it was probably just a bad slider. We can take or leave the five changeups, and since we already left them earlier, we’re going to do it again. We’re going to continue focusing on only the four-seamer, the knuckle curveball, and the slider.

The first thing that jumps out is the fastball is down below 50% usage. The next thing that sticks out is the breaking ball preference has flipped. By flipping the script, Peacock got four called strikes on sliders to lefties in his 2019 debut, plus one whiff. The lefties let the curveball go and it became a ball half the time. 

There’s one more thing about how Peacock employed his breaking balls against lefties to start the season. Here they are in heatmap form from the catcher’s perspective again:

peacock heat

Both pitches fell to the same lower, inside corner. With the side-to-side break less than an inch apart from each other, and the curveball dropping more than an additional 10 inches, Peacock used them in tandem to cause fits for lefties.

Think of this with the heatmap as an aid. The pitches most frequently fell to the black dot below the zone, but the pitches that created the gradation of maroon out to orange up and into the zone are the ones that made it so hard for the hitter to decide whether to swing.

You might be thinking, “Big whoop. The Astros played the Rangers in Peacock’s first start. They stink.” You wouldn’t be wrong. They’re projected to win about 75 games, give or take a few. But Texas offers some dangerous lefties. Nomar Mazara, Joey Gallo, Rougned Odor, Shin-Soo Choo, Asdrubal Cabrera, and Ronald Guzman combined to go 2-for-14 against Peacock. The lineup was stacked with established lefty big leaguers (save for Guzman) whose wOBA was between average and 50% above average against righties as recently as last year, and they were collectively helpless. 

The season’s about a week old. What’s happened so far is by no means an indicator of what will continue to happen. But Brad Peacock might have changed his approach to lefties and that could be a big deal. It might not solve his times-through-the-order problem, where he’s 75% worse than his peers the third time seeing an offense. He’s also 25% better than them when seeing hitters the first and second time, though, so an uptick in efficiency could produce some scary-good results. The Astros have pitchers in the system like Forrest Whitley and Josh James, plus others, who could eventually take Peacock’s rotation spot. For now, he’s locked in, and might have a new trick up his sleeve to help keep it that way. 

All data from Statcast unless otherwise noted. Feature photo Ron Schwane/Associated Press.



A Peak Into The Astros’ Secret Sauce for Pitching

The Franklin Institute is a science and research museum located in Philadelphia, PA. Among its many draws are a giant heart you can walk through, the SportsZone where you can sprint the 40-yard dash and compare your time to professional athletes, and a Changing Earth exhibit made entirely of sustainable materials that focuses on the ways the planet has transformed over time. Through all of that, plus rotating feature exhibits, it’s easy to lose sight of a tried and true experiment: The Ruler Drop Test.

If you never performed the experiment in middle school, the Ruler Drop Test is exactly as it sounds. Take a ruler — or, in the case of the Franklin Institute, a yardstick — and hold it vertically between your index finger and thumb on your dominant hand, about one-fourth from the bottom. Then release it, and see where you can catch it. The shorter the distance between where you let go and where you catch it, the faster your reactions are. Science!

It’s a simple experiment, but illustrative. And, with how it’s centered on vertical drop and expectations, it could help us understand how the Houston Astros have used advanced technology and data to tweak pitchers’ repertories to reach new levels of success.

The league average pitching staff has put up a shade over 14 fWAR per season since 2016. Meanwhile, the Astros have averaged five wins more than that over the same time span, capping it off last year with a bat-guano crazy 30.6 fWAR. The organization clearly knows something about pitching that the rest of the league hasn’t figured out yet. Justin Verlander and Gerrit Cole are second and fifth in pitching fWAR for the team since then, despite only combining for two seasons (and a month) as Astros. Here’s a big part of why:

Animated GIF-downsized_large (2)

Both Verlander and Cole came to Houston and changed. On the left in this gif is each pitcher’s breaking ball release point in the season before they got to Houston. On the right is those pitch release points after becoming an Astro. The change is two-fold: the release point on each has tightened up, and curveballs have become more distinguished and over the top.

One possible explanation could be how the team uses Edgertronic cameras to fight a pitcher’s intuitive notion of knowing exactly how the ball is leaving their hands. The high-speed cameras can capture thousands of frames per second. A pitcher using one means they can see exactly when the ball leaves their hands, where they last put pressure on it and with which finger(s), and how their wrist either pronates (meaning the palm moves inward, like with a curveball) or supinates (meaning the palm ends up facing outward, like with a changeup). It’s the kind of tech that teams are racing to incorporate just so they don’t get left behind at this point, that the Astros have been using longer than anyone.’center

There may not be another way of getting so granular with technique that makes it equally easy to understand in a matter of seconds. It’s almost a paradox. It’s not that Justin Verlander and Gerrit Cole weren’t good, but their abilities clearly weren’t optimized, and this is something we can only truly appreciate after seeing them take a step up we couldn’t even tell was there.

By cleaning up their release points and deliberately going over the top with their curveballs, each pitch became more devastating because it started falling off the table more. Their wrists may have started staying locked instead of leaking out toward third base. Verlander’s slider has gained more than two inches of drop, while his curve has gained more than an additional inch. Cole’s slider has actually lost about half an inch of drop, but his curveball has tumbled two-and-a-half inches more.

Those numbers sound well and good, but the results are plain eerie.

Eerie whiffs

The league-average whiff rate for sliders is 17%. For curveballs, it’s 13%. Houston acquired two really good power pitchers, made the same tweak to each, raised their whiff rates on one pitch from average to above, and their whiff rates on another from below average to average. These are the kinds of incremental improvements a team would be happy to take one at a time, from one average player at a time. They basically made both tweaks at once in back-to-back seasons with two guys who have elite talent. It seems there are times when we run out of superlatives to describe the Astros, but this is plain eye-popping.

It may also serve as grounds for how the team could tweak Wade Miley, who added a cutter last season that helped him tally 1.5 fWAR in just half a season’s-worth of starts with the Brewers. Below are his release points for the pitch, compared to his primary breaking ball, which was a curve.

Miley MIL

There are differences when you compare Miley with Verlander and Cole, and they’re obvious. He throws about seven mph slower than either of them. He doesn’t have two breaking balls like they do. These are no small things. But a cutter can be a bit in between a fastball and a slider, and bringing his curveball release over the top may help both pitches play up.

Even if last year’s numbers were flukey — his cutter, while useful, generated whiffs well below average — a rate of improvement from Miley comparable to Verlander’s and Cole’s could see him become a league-average pitcher who helps steady a staff that’s going to have a lot of turnover in 2019. The aces are still there, but Lance McCullers will miss the year with Tommy John, Dallas Kuechel’s on the open market, and Charlie Morton already left for Tampa Bay.

Let’s bring it all back to the Ruler Drop Test. When you perform it, you know the weight of the item you’re dropping. You’re almost completely still, save for a couple fingers releasing and pinching. It’s moving extremely slow. The environment is almost completely controlled.

Now, as a hitter, speed up the moving object to 80+ mph. Throw in a variable, downward path and less predictable, more severe vertical movement. And you’re not just catching the object with two fingers. You have to use your whole body in a smooth kinetic sequence to “catch” it with your bat. The Astros have taken what we know about the game environment and found a way to exaggerate its difficulties for the opposition by way of the breaking ball’s drop.

Maybe it’s not the full recipe to Houston’s secret sauce for pitching. But this basic test combined with the specific improvements from Justin Verlander and Gerrit Cole helps us get a grip on what the team does so much better than the rest of baseball.

Pitch movement data from Brooks Baseball. League-average whiff rates from Alex Chamberlain. Pitcher-specific whiff rates and release points from Baseball Savant. Feature photo Karen Warren/Houston Chronicle

Debut Review: Josh James Is The Next Ascendant Astro

Consider this: there is a prospect who starts the season in Double-A. He shoves, and is promoted to Triple-A. He shoves there, too. Naturally. He totals 171 strikeouts across both levels in just 114.1 innings with easy mid-to-high 90s heat and a sharp, horizontal slider. He’s bound to catch some eyes around the game, right?

Except if your Big League club happens to be the Astros, who have spent the majority of the season in first place, and largely because of an historic rotation. In that case, you might not be catching many eyes at all, because the team that could use you most hasn’t had to even really think about it. And so it went for Josh James through August of this season.

But as tends to happen, the grind of the schedule eventually knicked up a couple Astros starters. Between that and rosters expanding, James has gotten his shot. And he’s cashing in.

He debuted on September 1 against Mike Trout and the Angels. He delivered nine strikeouts over just five innings and 91 pitches. His velocity peaked at 101.1 mph. It wasn’t all roses, as he also gave up three free passes and three earned runs. If you were to only scout the statline, you might walk away neither underwhelmed or overwhelmed, but merely…whelmed.

That wouldn’t tell the whole story, though. For one thing, he threw 52 of his 91 pitches in the first two innings alone. Getting through the final three frames with just 39 pitches shows, at the very least, that James settled in. If we look at where he was spotting his Ks, it could tell us more.

James Ks

The red dots are fastballs. The yellow ones are sliders. Six of his nine punch out pitches hit the edges of the strike zone. One coaxed a chase out of the zone. James managed five swinging strikeouts and four called strikeouts. In his first game he threw 101, painted edges, got batters to whiff, and fooled them into taking pitches they should’ve cut at.

If we look at the low points in that start — the three walks and three runs — there are a couple dots to connect. All three runs came on a pitch laced middle-in to the left-handed Kole Calhoun, who rocketed it over the fence and into space for a monster dinger. One of the three walks came in Calhoun’s next plate appearance in a full count. It appeared as though James didn’t even want to throw anything only Vlad Guerrero could hit at that point. That’s not exactly brilliant or anything, and probably shows that James has steps to take in order to get better. But stacking it all up as it happened makes it a pretty darn good debut.

After his electric first outing, James was moved to the bullpen because of Houston’s remarkable depth. He’s appeared in two games since then. In the first, he went two and two-thirds innings in Fenway against the Red Sox, gave up only one hit, walked none, and struck out four. Then in his next outing, against the Tigers in Detroit, he went another three innings. He struck out four, walked one, and hung a slider that Nicholas Castellanos took yard. What we’re seeing here is a lot of talent and a relative amount of volatility that might need something to give it a boost.

Enter James’s changeup. In his first appearance it didn’t have much shape. Maybe he was gripping it too tight, or releasing it from a funky spot, but it was a pitch that daggered to its location and wasn’t particularly useful. In each of his two relief appearances, he’s gotten a strikeout with it, and, per Pitch Info on FanGraphs, it’s had at least another half inch of drop and three inches of fade. One of those strikeouts was against fellow righty Mookie Betts in a 2-2 count, showing confidence in the offering.

Now, yes, we’ve gotten particularly granular. We’ve analyzed each of the three pitch types James throws, against specific hitters, over only 10.2 innings and three appearances. We’ve considered results that aren’t necessarily sticky, and certainly aren’t over such a small sample size. But we’re also seeing data that we couldn’t publicly see on James from the minor leagues. These three appearances have proven useful in whether to buy into his breakout or not. We can see that he’s got a fastball he can pound the zone with, a breaker he can throw against righties, and an off-speed offering he can throw to lefties. He’s shown a willingness to throw all three in different counts and contexts, as well an ability to locate them.

If he stays what he is, Josh James is already another incredibly useful asset for the Astros. He can go multiple innings out of the bullpen getting whiffs just like Chris Devenski and Brad Peacock and Collin McHugh have for the team in the last two years. If he continues to develop some nuance in his repertoire, he’s the kind of starter built for modern baseball. Either way, he provides depth for Houston and can help them sustain their success moving forward. He still might not be drawing the amount of eyes his talent should, but he probably will soon. Keep yours peeled.

Strikeout chart, velocity, and pitch data from Baseball Savant. All other data from FanGraphs. Feature image from Karen Warren/Houston Chronicle

George Springer Isn’t Quite Seeing What He Wants

Look up and down the Houston Astros roster and it’s difficult to imagine them getting any better. They’re on pace to win 105 games, which is four more than even last year. But it’s possible. Though their pitching staff is the best in the league and maybe one of the best ever, their offense has been more middling. And it may start at the top with George Springer.

So far Springer has registered 1.6 fWAR and a 113 wRC+ in 97 games. While being 13% better than league average is pretty good, it’s not quite what you’d expect from him. Last year he surged to a 140 wRC+ mark. Even if you account for regression, you don’t account for him being almost 30% less than the batter he was just a season ago.

Right now the projections love him. He’s pegged to account for at least 1.5 wins for the rest of the year, in less than 60 games, and a wRC+ of at least 128. And remember that as projection systems evaluate a player’s true underlying talent level at a given point in time, they’re also conservative in nature. You could somewhat reasonably argue, then, that Springer could possibly manage an even bigger rebound here as the season resumes.

But there’s a catch with projection systems. They might capture a player’s true talent level, but by nature they can’t capture all that goes into preparing for that player. Maybe based on George Springer’s past body of work, compared to players of his ilk and age, he really is a hitter who is at least 30% better than his peers right now. But based on how pitchers have attacked him this year, he hasn’t been, and there’s one reason that sticks out as to why.


Pitchers are locating their four-seamers to Springer differently this year. On the left, you see where they spotted the pitch to him in 2017, mostly outside. Springer is 6’3 and looks every bit of it in the box. He has an upright stance. When he gets ready to swing he becomes relatively compact. His arms move down while his hands load and he has a moderately  pronounced leg kick. Given how he condenses himself, it’s possible pitchers felt there was an opportunity to attack away because of how it would take him more time to expend the energy to get there on their fastest pitch.

But if you look on the right side of the heat maps above, you’ll see the fastballs Springer swung at. He didn’t have difficulty getting to those pitches and you can see why for yourself if you get out of your seat and pretend to take a swing as you read this next part. (That’s what I did. It was fun!) Go ahead. Stare down Luis Severino or Jacob deGrom fearlessly as you ready yourself for what’s about to come. Make sure you’re in an upright stance. Slowly coil up as you get ready to take your cut. Follow through.

Notice where your arms and legs go? They explode out. They pretty much have to, right? Now imagine you’re a top tier athlete on a top team in the world, like George Springer is, and you can see how he’d shred fastballs on the outer half. He accumulated a 17.4 pitch value against four-seamers last season, which was good for 15th in all of the Majors.


So, the solution? Try to take advantage of the way Springer coils up. Bust him inside some more to keep his body and bat from exploding through the pitch. And so far this season, it’s working. He’s managed only a 6.9 pitch value against four-seamers so far. That’s still relatively nice, and top 40 in the Majors. His wOBA against four-seamers this year is still .381, but that’s down nearly 50 points from last year. In many ways he’s been perfectly cromulent, even if a far cry from the top six outfielder and top 20 hitter in all of baseball he was in 2017.  

Pitch values come with caveats. They can be deceptive because on the surface they look like they report only on one specific pitch, but the value of each pitch is often heavily tied to how it’s sequenced with others. We don’t immediately know what set up the performance of the pitch we’re evaluating, and that’s a big deal. However, Springer has seen and offered at pretty much the same volume of four-seamers as just a season ago. Pitchers have merely changed where he’s seeing it.


If he’s still favoring swinging at the fastballs on the outer half, it’s probably because he knows he can crush them. Just last night, on June 21, poor Noe Ramirez served him a pitch right in his happy zone and the Angels paid dearly. The Astros are a smart club. Maybe they’ve tried or are brainstorming possible solutions to Springer getting crowded with heat. Maybe they’ haven’t and they’re just telling him to keep on keeping on because it’s not like he’s turned into a liability. But good gravy, imagine if he figures it out before October.

Pitch percentages and heat maps from Statcast. All other data from FanGraphs. Feature image from Melissa Phillip/Houston Chronicle


Relievers Who Will Matter in the Second Half

A slump-proof, lockdown bullpen doesn’t just win games. It can effectively end them before they’re over. But relievers are weird. Even when they’re not pooping their pants, they’re probably the most volatile players in all of baseball. They seem to represent only the foremost moment in any given season, making trying to project which ones will be good largely a fool’s errand.

But there is a tool that can help, maybe: SIERA. That’s Skill-Interactive ERA. It’s an ERA estimator like FIP or xFIP, but it’s better because it accounts for more of the noise that can result from batted balls. It also has a stronger correlation to predicting a pitcher’s future ERA.

It’s important to acknowledge that it isn’t an ERA projector, but can inform us of the quality of the skills a pitcher has demonstrated most recently. And now, as the season heats up, and as potential playoff teams show more urgency, and we’re in the foremost moment the season has to offer, we can use SIERA to see which of baseball’s oddest bunch could offer big benefits in the second half. Let’s dig in.

Juan Nicasio currently has a SIERA of 2.49. His ERA is a flat 6.00 through 34 appearances. Because SIERA is best used as a starting point for evaluating a player, the disparity between his results versus how he’s actually pitched pushes us to look further. One thing that jumps out is his strand rate, which stands at a homely 53.3%. That’s 20% worse than league average for relievers. It’s probably fueled by a .396 BABIP which is a whole hundred points worse than league average, and this is all happening while he’s striking out more and walking less batters than he ever has.

The thing about Nicasio isn’t any of those wonky stats, though. It’s that it’s hard to see him not getting better while playing on a team that’s been thriving in one-run games all season. The Mariners may effectively gain a lockdown arm for their bullpen as the ledger balances for him, and they’ve already had a top ten group by fWAR. What they’re doing is unprecedented and Nicasio is another reason it could keep happening.

Photo: Karen Warren/Houston Chronicle

Will Harris would likely elicit a shrug from anyone who peered at his 4.15 ERA. His FIP and xFIP are both sub-3.00, though, and his SIERA is an even tinier 2.40. Including him here might be considered cheating in two ways: he’s appeared in ten games over the last month with an ERA of 2.70, and he’s an Astro.

He was victimized by home runs earlier in the year and has been better at keeping the ball in the park, having allowed only one dinger over the last 30 days. It helps that he’s striking out a career high, too, with a reworked curveball that’s tighter and sharper than ever. Remarkably, he might only be the third-best Astros reliever behind Collin McHugh and Brad Peacock. Maybe it’s not cheating so much as it is just unfair that the Astros could get even better with a guy they’re already trotting out there.

And then there’s a pair of Phillies, Hector Neris and Tommy Hunter. Neris has become much maligned and was even sent to the minors to figure himself out. He’s given up a homer on nearly every third flyball allowed, which is bonkers. His fastballs have flattened out, which probably plays into his splitter playing down, too. While his 6.90 ERA is woof-worthy, his 2.95 SIERA is pretty nice and tells us his fastballs being worse shouldn’t make him this bad.

Phillies general manager Matt Klentak caught some flak on talk radio for recently saying that Tommy Hunter’s 2018 has actually been one of his best. His ERA is approaching 5.00 but his SIERA sits at 2.87, so maybe Klentak’s statement gives us a glimpse into the team’s beefed up sabermetric approach. Hunter has fallen victim to similar issues as the others above — high BABIP causing a lower strand rate.

Photo: Chris Young/CP

The thing about Hunter (25.1) and Neris (30) is they’ve accounted for 55.1 innings out of the Philadelphia bullpen. Positive regression for them could be critical for the team, as others like Edubray Ramos and Victor Arano are slightly outperforming their peripherals so far. They’re on pace for 88 wins, and every inning is going to be important for them in the second half as the team pushes for the playoffs for the first time since 2011.

Looking at a pitcher’s SIERA gives us a stronger sense of their most recent performance. It can also give us a sound starting point for where else to look to understand how the moment has treated them. Beyond that, it can also help us zoom out and examine a pitcher’s potential impact on their team while we move onward to October, no matter how weird they are or have been.

Data from FanGraphs. Feature image from Elaine Thompson/AP.