The secret to Atlanta's defensive turnaround
This browser does not support the video element.
The Braves, having dispatched with the Brewers, have some time before starting Game 1 of the National League Championship Series vs. the Dodgers on Saturday night at Truist Park. That gives us some time to ask if you¡¯ve noticed the substantial change they made in the middle of the season -- and no, we don¡¯t mean their rebuilt-on-the-fly outfield, impressive as that was.
Did you notice a team that spent years not shifting ¡ suddenly shifts now? A lot? And turned around their infield defense entirely from a problem to a strength?
The formerly shift-averse Braves are moving their infielders around out of nowhere, and largely without notice. If you watched the NLDS, you could see it. You saw it when Avisa¨ªl Garc¨ªa smoked a line drive directly to Austin Riley. You saw it when Omar Narv¨¢ez hit a grounder right at Ozzie Albies, who started a double play. You saw it when Willy Adames, who had been shifted against by Atlanta 19% of the time in the regular season, saw shifts more than 80% of the time in October.
Of the six runs Milwaukee scored in four games, only two came on non-homer hits. Credit to the pitching, obviously. Credit, also, to the infield that allowed just five ground-ball hits, especially when you look back to where they came from at the start of the season.
¡°I leave that up to Wash,¡± manager Brian Snitker said last summer when we asked him why the Braves almost never shifted, referring to third-base/infield coach Ron Washington. Fair enough, so let¡¯s go back to 2017, the first full season for both men on Atlanta¡¯s staff. See if you can note a trend here.
"They¡¯re shifting more than they used to, so what, plenty of teams have upped their shift rate over the last five seasons,¡± you might say, and that's fair. Let¡¯s put it a different way.
From 2017 through April 30, 2021, only two teams shifted less often than Atlanta did. In 2019-¡¯20, no team shifted less often. Even in April of this season, no team shifted less often. The Braves, as a rule, hardly ever shifted.
Until, of course, they did.
From May 1 through the end of the year, only two teams (the Dodgers and Mets) shifted more than the Braves. In the month of June, no team shifted more than the Braves. They¡¯re still doing it in the playoffs, too, more than twice as much as they did last year -- which is ironic, given that after their Game 7 NLCS loss to the Dodgers a year ago, the Ringer ran an article headlined ¡°The Shift May Have Cost the Braves a Pennant.¡±
The Braves didn¡¯t just shift a little more, they shifted a lot more, and they started doing it in the middle of a season. This is the same team that in early April once went nearly a week without shifting a lefty batter, despite facing Bryce Harper and Kyle Schwarber in that time.
What in the name of Mark Lemke and Jeff Blauser is going on here? And: Has it helped? Let¡¯s try to find out.
This browser does not support the video element.
Ask anyone about the Braves' infield defense, and they¡¯ll invariably start talking about Washington, a legendary infield coach dating back to his days with the A¡¯s in the 1990s. (Yes, the guy from the ¡°Moneyball¡± scene, if you want to distill decades of a baseball life into three possibly apocryphal words from Brent Jennings to a pre-stardom Chris Pratt.) He¡¯s still out there doing pre-game work with his infielders every single day, and Freddie Freeman and company swear by it.
Washington turns 70 next April and is perceived as old-school in many ways, though his Texas teams were shifting against right-handed power hitters like Vladimir Guerrero Sr. as far back as 2008, and it¡¯s hardly like he¡¯s refused to acknowledge the advances in the game over the last several years.
"We are playing that way because [it] is a reflection of what the spray charts say,¡± he told the Dallas Morning News in 2014, after being asked about a shift, ¡°and we're going to do what the spray charts say."
Makes sense enough, and he¡¯s hardly alone in that approach, then as now. What stands out to us, though, is how sudden this change was. Although we used monthly data above for high-level convenience, the Braves didn¡¯t exactly start shifting the exact moment that April became May. They very specifically began shifting in the middle of a mid-May homestand, and not gradually. Immediately.
On May 19, a 5-4 win over the Mets in Atlanta, they shifted on just 6% of pitches. They went home that night and returned the next day to face the Pirates. They shifted 10 times as often (64%) against Pittsburgh as they did against New York merely a day before. The next day, 60%. The day after that, 68%. After that, they never really looked back.
Only twice for the remainder of the year would they shift less than 20% of the time, each coming in September games against weak Miami and Washington lineups. In one notable game, a 4-3 win over the Mets on July 1, they shifted a full 90% of the time. It was the fourth-highest shift rate in a single game of any team all season long. Baseball¡¯s anti-shifters had suddenly gone full shift.
You have to split it, though, by handedness to get the full story. When the Braves shifted less than anyone in baseball in 2020, they were shifting right-handed batters a mere 6% of the time (third-lowest) and left-handed batters only 12% of the time. That wasn¡¯t just the lowest mark; it was less than half as much as even the 29th-most club, St. Louis.
Given the increasing evidence of what we know about shifting -- that the data doesn¡¯t exactly support doing it as much as teams do against right-handed batters, though it remains sound against lefties, which is what the Ringer article was describing in regards to Will Smith¡¯s game-tying grounder -- you¡¯d have expected that if the Braves would start doing it, they¡¯d mostly up their game against lefties.
They did so, going from last year¡¯s 12% against lefties to 61%, 6th-most. Yet they¡¯ve done it against righties, too, going from last year¡¯s 6% to 45%, second most behind the Mets, despite some seriously open questions about whether that¡¯s actually a good idea.
That discussion remains unfinished. It¡¯s fascinating to see what happened in Atlanta, however. On grounders and liners in play, lefty batters had almost no change in outcomes pre- and post-May 19, going from a .360 average and a .346 wOBA before the shift change to a nearly identical .359 and .345 following it.
Righty batters, however, saw a large decrease in production, dropping from a .403 average and a .388 wOBA through May 19 all the way down to .346 and .330 starting May 20. (If these numbers all seem high,remember that we¡¯ve cut out strikeouts, fly balls, and pop-ups here.)
Whatever the strategy, the Braves have shifted into high gear, literally.
But what really caused the change? It couldn¡¯t just be Washington on a whim, and it wasn¡¯t.
¡°It¡¯s about guys doing a lot of hard work upstairs,¡± bench coach Walt Weiss, who joined Snitker¡¯s staff following the 2017 season, told MLB.com¡¯s Mark Bowman before Tuesday¡¯s Game 4. ¡°The guys who are digging in on our defensive alignments have done a really good job.¡±
¡°Early on, there was some trial and error with some things. There were too many ground balls getting through. So we got together and we started to discuss how we could fix this. It's a credit to [general manager] Alex [Anthopoulos¡¯s] team up there. The adjustments they made on the fly. I mean, after the first four or five weeks of the season, [the infield defense] has been airtight.¡±
After the first four or five weeks of the season lines up pretty well with the mid-May change. Weiss¡¯s note about poor defensive performance lines up with what we observed, too. The early-season Braves weren¡¯t a great run prevention club. That changed.
Braves through May 19
- 59.8% of grounders/liners converted into outs (sixth-worst)
- 213 runs allowed (sixth-most)
Braves since May 20
- 64.4% of grounders/liners converted into outs (best!)
- 443 runs allowed (second-fewest)
At times, it¡¯s seemed like the only Atlanta infielder to miss a catch has been Chipper Jones.
That¡¯s a pretty stunning turnaround, and unlike in the outfield, which had to deal with multiple losses before several July trades remade the roster, the infield has been consistent all year long. Of the 644 regular season starts at the four infield spots, the quartet of Freeman, Albies, Dansby Swanson, and Riley made 628 of them. It's not about new players.
We won¡¯t attribute the improvement entirely to shifting -- more on that below -- because obviously an in-flux pitching staff had something to say about this, but in the same division, in the same year, the Mets had a massive shifting increase and also saw their defense perform far better.
¡°We ended up having a meeting,¡± said third baseman Riley on Tuesday, before Game 4. ¡°All the infielders, kind of coming together, trying to figure out, make sure we're all on the same game plan as the pitchers. We want them to be comfortable, feel comfortable, when they're on the mound.¡±
You can see how it¡¯s supposed to work, in practice anyway. On May 14 against these same Brewers, a week before the shift took over, Kolten Wong singled up the middle, to the right of second base.
This browser does not support the video element.
On the final day of the regular season, Dom Smith hit a ball to almost exactly the same spot, but into the shift. This time, a Braves infield full of backups turned a double play.
This browser does not support the video element.
But defensive positioning isn¡¯t only about full shifting -- defined as having three infielders to one side of second base -- either. Sometimes, it¡¯s a lot more subtle than that.
In Game 3 of the NLDS, a pivotal moment appeared in the eighth inning. Atlanta was up, 3-0, thanks to Joc Pederson¡¯s pinch-hit homer, but Milwaukee had put runners on the corners with one out against Luke Jackson, as Christian Yelich strode to the plate. Yelich is hardly the MVP-caliber hitter he appeared to be in 2018-¡¯19, but he still represented the tying run in a series that was tied at one game apiece.
Jackson¡¯s first pitch was a slider to the outer half of the plate, and Yelich hit it hard up the middle, nearly 105 MPH off the bat, though on the ground. If the ball gets through, the runner on third easily scores, Willy Adames probably goes first-to-third, and Garc¨ªa comes up as the potential go-ahead run.
The ball did not go through. Swanson, the shortstop, was not in a shift -- he was on the shortstop side of second base -- but he was positioned so far up the middle that he had to move all of 12 feet in two seconds to get the grounder, step on second, and throw to Freeman to end the rally and all but seal the game.
This browser does not support the video element.
Now: Here¡¯s how many times, in the entire 2020 season, the Braves positioned their shortstop in a similar spot (within five feet) in a similar situation (a lefty batter, with a runner on first and less than two outs): Once. One pitch. That¡¯s it. It was 14 times in the first two months of 2021. He was there 13 times in the first three games of the NLDS alone, and this is kind of it. Those weren¡¯t all shifts. It¡¯s not always about shifts.
Nor is it always perfect. Riley, too, noted the trial-and-error Weiss had mentioned, which might explain why he¡¯s viewed so differently by some different metrics, coming in at a positive in some, but a negative-6 in Statcast¡¯s Outs Above Average.
¡°Early on we were shifting, like I said, every lefty I was going over [to short right field],¡± Riley said. ¡°And we kind of were getting beat a little bit with some slaps to the other way. And just those, like really just focusing on guys that extremely pull the ball, I think, is what the minor adjustment is what we had to make.¡±
Riley, per OAA, was roughly average or only slightly below in both April (-1 OAA) and May (-2). He was baseball¡¯s weakest fielder in June (-8 OAA) as the shifts ramped up, though that¡¯s not just about advanced metrics; at one point, he made five errors in five days. It settled down in July (+1 OAA) and August (+1 OAA); in September, he¡¯d rebounded so well that at +5 OAA, he looked like one of baseball¡¯s best third basemen.
¡°There's no egos involved, we just try to get better,¡± said Weiss. ¡°That relationship has been really good. You combine the information with the wisdom that comes with experience and that's the sweet spot.¡±
¡°I think that just goes to show upstairs, I think they've done a really good job of game planning for this series,¡± said Riley.
They have. They¡¯ll have a lot to say, too, about whether they get through the next series, to get Atlanta to a most unexpected World Series. The Braves have shifted everything.