On July 9, a Marine in blue dress D's picked up a baseball that had landed in the lounge area beyond the left-field wall at Marlins Park. He dropped the ball into a cup of beer and then chugged the beer, an amusing faux pas that got the attention of the internet. That attention dramatically increased the viewership of Starlin Castro's seventh home run of the season, which had occurred before MLB's 10th-smallest crowd of the year.
So what most of the world saw was only the final pitch of the at-bat. It had come on the 11th pitch from Josh Hader, one of 2018's breakout stars -- this was, in fact, his first batter after being named an All-Star. At that point in the season, there was a plausible case no pitcher had ever been as difficult to face as Hader, used as Hader had been used. Batters were hitting .104 against him, which would have been the lowest in history for a season of at least 20 innings. More than half the batters he had faced had struck out -- the second highest K rate in history. That's how this plate appearance started, and then, for Castro, it got exponentially more hopeless.
And yet you know how it ended. We spent a lot of time looking at Hader this year. Let's take a moment to look at Castro, and how a battle is won.
Pitches 1 & 2: Danger has a bracing effect
Batters in 2018 had two routes to success against Hader. One was forcing Hader to a three-ball count. The other was winning the count early, either by putting an early strike in play -- he had one of the highest first-pitch swing rates in baseball, and batters slugged .762 when they put that pitch in play -- or by getting ahead in the count with a first-pitch ball. Batters who got ahead 1-0 managed to hit .198/.364/.333 against Hader in those at-bats, which is almost a normal hitter's batting line. Batters who fell behind 0-1 hit .058/.093/.130, which is almost a high school locker combination.
Castro was aggressive. Hader threw two fastballs -- one a coin-flip strike, the other certainly one -- and Castro swung at both. He fouled one. He whiffed on the second.
This is where baseball becomes the most lopsided matchup in all of sports. Castro's situation seemed nearly hopeless. Ninety-one other batters fell behind 0-2 against Hader this year, and three-quarters struck out. Those 91 would go on to hit .034/.044/.068 against Hader. Stare at that number and allow yourself to admire each third on its own pathetic terms. A lineup of .034/.044/.068 hitters would, according to this tool, score minus-3.823 runs per game. I buy it.
Pitch 3: If you fight with all your might, there is a chance of life, whereas death is certain if you cling to your corner
Hader threw another fastball, over the plate and up in the zone -- almost identical to the 0-1 fastball Castro had whiffed on a pitch earlier. This time, Castro just got a piece of it and fouled it back.
Two-strike fouls, of course, change nothing. The count stays the same. We say merely that the batter "stayed alive," and the assumption is a foul ball just preserves the unbalanced status quo. But sabermetrician Russell Carleton has found that foul balls on 0-2 counts actually move the needle closer to the hitter's favor. Batters who fell behind 0-2, at the time Carleton researched this, had a .209 on-base percentage if they didn't foul any balls off. If they fouled one or more, their OBP jumped about 40 points. Carleton controlled for the quality of pitchers and batters, because it could be that better hitters or worse pitchers were likely to be involved in foul balls. But the effect persisted: Staying alive isn't just staying alive -- calling time and stepping out so the pitcher never even throws that pitch would be staying alive, too -- but, in a subtle and invisible way, it moves the hitter toward fuller health.
Pitches 4, 5 & 6: Therefore, just as water retains no constant shape, so in warfare there are no constant conditions
Hader, reaching back to add a tick to his fastball, threw a fourth fastball, up and out of the zone. The pitch didn't end up being very close, but from Hader's perspective, it was a smart 0-2 pitch. Castro had seen three high fastballs and swung at all three, so this pitch was tempting him to go a little bit higher. If Castro had been expecting to see his first slider -- he'd never faced Hader before, by the way -- then he could have misidentified this pitch out of Hader's hand as a slider in the zone. And the high fastball could set up Castro for a slider down on the next pitch.
Castro's hands flinched a little, but he took it for ball one. The new count would do almost nothing to change Castro's chances -- batters had been almost as badly overmatched after 1-2 counts against Hader as they had been after 0-2 counts -- but the walk back to a favorable count takes multiple steps, and Castro had taken the first one.
Hader then threw two more fastballs:
Neither pitch was likely to be called a strike, according to ESPN Stats & Information: The first, just off the outer edge, is in a location that is called a strike only 21 percent of the time, and the second -- clearly high -- is a strike only 2 percent of the time. Hader was hitting his targets; he was using the count to get Castro to expand the zone, and it was working.
The high fastball Castro fouled off, the sixth pitch of the at-bat, was Hader's first to hit 96. Castro just missed it, lining it straight back before taking a walk out of the batter's box. It was objectively a much more difficult pitch to hit than the 0-1 fastball Castro had whiffed on: 2 mph faster and eight inches higher. But Castro had become accustomed to Hader's fastball. Instead of swinging right through it, Castro nearly caught up to it. It's the first pitch in this at-bat where you could envision Castro actually beating Hader's fastball. The balance shifted slightly.
Pitches 7 & 8: If your enemy is secure at all points, be prepared for him
Hader was still ahead in the count. But now he had seen Castro on his fastball, so he had to work a little bit. He threw a slider. It was a terrible slider, way outside, truly non-competitive -- Hader only threw four sliders farther outside to righties all year. It never came close to the strike zone, didn't cross through it, didn't tumble out of it, didn't at any point resemble a pitch that might come into the strike zone.
But Castro swung at it. That's the power for the pitcher of a 1-2 count, and it's the benefit for the pitcher of having thrown six pitches without showing the breaking ball. Castro had reached a point where he had to be looking for fastballs but, somewhere in the back of his mind, cognizant of the slider, to which he could have no approach other than "protect." He protected!
I should mention that to this point, Castro had been taking what you might call big swings at every fastball. (This might seem like bad two-strike hitting, not befitting the tradition of Dave Magadan or Carney Lansford or whichever .288 hitter you grew up admiring, but keep in mind that you know how this plate appearance ends.)
But within that broad description, there had been adjustment. If you were to look very closely at each swing on the previous six fastballs -- which I'm not recommending so much as letting you know that I did -- you'd see he'd been scaling his swing back a little bit each time. He was keeping his hips just a little bit more closed. At full speed, this isn't apparent. But in slow motion, it's clearer Castro was adapting, which is probably how, in this desperate swing against an unhittable distant slider, he was able to stay back, reach out and foul the pitch away.
It was another victory. He'd now seen both of his primary pitches. Further, Hader might have acquired a tinge of doubt about his slider, as the first one he threw after entering the game was a mess. His next pitch was a fastball, and Castro once again took a very comfortable swing but just missed it. Castro was still way behind in the count, but he'd narrowed the gap.
Pitch 9: Those skilled at making the enemy move do so by creating a situation to which he must conform
On his fifth 1-2 pitch, Hader threw a changeup.
Hader has a changeup in roughly the same way every position player claims to have a knuckleball. He knows how to throw it. He can show you his grip. He probably toys with it a few times a year. But he almost never actually throws it. Hader will end the season having thrown four changeups all year, none of them for a strike.
In fact, going further: Three of the changeups he threw came in a single outing in early April. Which means that from April 7 through the end of the season, Josh Hader would throw only one changeup -- this one, to Starlin Castro.
We're making a lot out of this one at-bat, so you know we're going to make a lot out of this one pitch: Castro, facing the most dominant pitcher in baseball, and still down in the count 1-2, a situation Hader "wins" more than 90 percent of the time, had so frustrated Hader's attempts to strike him out, Hader turned to a pitch he otherwise never throws to anybody under any circumstances. It's enough to make you think Castro was winning this at-bat.
The pitch was almost a foot and a half outside. Castro didn't even need to flinch. While Hader wiped his hand on his leg and retreated to rub up the baseball, Castro stayed perfectly still, watching Hader. This is different from Castro's routine throughout the at-bat, when he would retreat to the corner of the batter's box to reset between pitches. He was now in this plate appearance. He was the first batter in three months to take Hader to a 10-pitch at-bat, and the first in just as long to force Hader to throw all three of his pitches. And, with the count at 2-2, he had increased his chances of a positive outcome -- most likely a walk -- to almost 1 in 5:
Batters vs. Hader, after 0-2: .034/.044/.068
Batters vs. Hader, after 1-2: .035/.090/.079
Batters vs. Hader, after 2-2: .052/.191/.103
Pitch 10: If it is to your advantage, make a forward move; if not, stay where you are
Hader threw one more fastball, but for the first time in 10 pitches, he went not outside but inside; and for the first time with a fastball, he went not high but low.
The result is probably in the rulebook strike zone:
But that's not an automatic strike -- 69 percent, according to ESPN Stats & Info -- and, because Castro (and perhaps the home-plate umpire, and perhaps we at home) had become so accustomed to Hader working the outer half of the plate, the pitch felt more inside than it was. Castro jerked his legs out of the way (though the pitch wouldn't have hit him), and his body language didn't seem too concerned about the pitch being called strike three.
Castro wouldn't have known it was 69 percent likely to be called a strike. But if he merely intuited the call could reasonably go either way, he made a rational decision. After Hader's steady stream of fastballs up and away, it's hard to think Castro would have put a very good swing on the low-and-in pitch. But getting another ball on the count changed everything, because 3-2 is finally when Hader's two-strike edge wears off:
Batters vs. Hader, on 3-2: .185/.500/.407
A huge part of that is the walk rate. Hader's weakness, if he has one, is that some of his pitches aren't strikes, which is costly in three-ball counts. But a fair part is he becomes more hittable. He throws a much higher percentage of pitches in the zone, instead of just outside of it or right on the edge of it:
0-2: 36 percent of Hader's pitches are in the zone
1-2: 52 percent
2-2: 59 percent
3-2: 63 percent
Batters make more contact:
0-2: 37 percent whiff/swing rate
1-2: 41
2-2: 50
3-2: 19
And they hit the ball harder:
0-2: 82.5 mph average exit velocity
1-2: 84.0
2-2: 83.6
3-2: 88.5
Starlin Castro started out an underdog -- an average hitter against an extremely good pitcher. Then he became prohibitively disadvantaged. It was almost unreasonable to expect him to do anything good. But through the steady erosion of Hader's edge, he worked himself into slightly more favorable situations until, at 3-2, he finally had the advantage. None of what happened was extraordinary. None of it even really reflects Castro's gifts as a hitter: Fouling off balls on two strikes is a skill, but Castro ranks near the bottom of the league. But what Castro usually does didn't really matter. What Castro actually did was fight for an inch at a time until he'd changed the entire plate appearance. And finally, on 3-2, he got the pitch he could hit.
Pitch 11: In war, then, let your great object be victory, not lengthy campaigns
Pitch 1 (to the next batter): Every battle is won before it's ever fought
You'll never be able to prove Castro's at-bat contributed to Brian Anderson's home run on Hader's very next pitch. But as Anderson's home run landed, the camera cut not to Anderson but to Castro, who had stepped out of the dugout and was almost posing. Anderson's home run certainly felt like an extension of Castro's at-bat. On Milwaukee radio, the broadcasters bemoaned Anderson's home run by bringing up the 10th pitch of Castro's at-bat, the fastball that wasn't called a strike. They were tying Anderson's swing to Castro's take.
"He put up a good fight," Hader said afterward. "I tried to switch it up and get him with my inside fastball and then elevated after that. He got the pitch."
The league is filled with incredible athletes who can do incredible things. We talked a lot about Hader this year. We rarely talked about Castro, an average player on a terrible team that nobody cared about or watched unless their favorite team was playing against them or a fan had done something funny. But for a minute that night, Hader did: "These guys are here for a reason, too. He did his job."
All header quotes have been attributed to Sun Tzu. Here's another: "There are not more than five primary colors (blue, yellow, red, white and black), yet in combination they produce more hues than can ever be seen."