clock menu more-arrow no yes

Filed under:

Three weird facts about the mighty 2019 Braves’ offense

New, 12 comments

Make no mistake: this has been a good offense. It’s gotten there in some weird ways.

MLB: Atlanta Braves at Chicago Cubs Matt Marton-USA TODAY Sports

After Wednesday’s victory over the Cubs, the Atlanta Braves had amassed 48 wins to 33 losses, the team’s best half-season mark since the 2010 campaign. (You have to go back to the 101-win 2003 team to find a better half-season mark.) The production by component, however, has been anything but even, however:

  • Offensive value — fourth in MLB (excluding pitchers);
  • Defensive value — 14th to 25th in MLB, depending on the metric used;
  • Starting rotation value — 20th in MLB (or 18th by RA9-WAR); and
  • Bullpen value — 22nd in MLB (or 15th by RA9-WAR).

That’s... that’s pretty lopsided! Even in terms of top half placement of a component, you’ve got the team’s offense and then just team DRS. But, the offense has been enough, and it would be enough even if the team weren’t outplaying its run differential or net outcomes, as the Braves have the best record-predictor metrics in the division, no matter which particular one you look at.

On the surface, there’s nothing specifically notable about the Braves’ offensive production. They’re top 10 in homers, runs scored, walk rate, strikeout avoidance, and ISO, and just about there in baserunning, but never actually in the top five until you get to the aggregate measures, including the basic stuff like batting average and OBP as well as the more holistic wOBA and wRC+. Yet, drill a little bit deeper, and things get weird. Let’s take a look.

(All stats through Wednesday’s 5-3 victory over the Cubs.)

Weird Fact #1 — The Second Time Through the Order

As pitching roles evolve in baseball, we talk a lot about the need to insulate pitchers from facing an opposing lineup a third time. However, that doesn’t mean that pitchers suffer no times-through-the-order (TTO) penalty for their second trek through opposing hitters. Rather, the effect is fairly consistent between the first and second times through, as it is between the second and third. As Mitchel Lichtman puts it:

In summary, we can say this: The first time facing the lineup, the starting pitcher has the advantage, as compared to his overall “true talent.” The second time, the battle between the pitcher and batter is roughly neutral. The third time through the order, the batter gains the advantage.

Now, that article was written in 2013, and things have changed since then. Check out the below.

(Note: The minus on each stat indicates that each value is X% above or below league average, where below is better.)

While some of this might be partial-season, smaller-sample weirdness, the reality is that while the third TTO penalty has shrunk, the second TTO penalty remains mostly the same. Interestingly, for wOBA rather than fielding-independent measures, there doesn’t even appear to be a specific third TTO penalty any longer. (The reason is hopefully clear even without me writing this: the fewer pitchers allowed to pitch third TTO, the more that sample over-represents better pitchers or pitchers facing weaker offenses/cruising on that particular day.)

But, what does this have to do with the Braves? Well, let’s swap out the 2018 chart with a Braves 2019 chart instead.

(Note: wOBA- for the Braves is not adjusted for league or park and simply calculated by dividing the relevant wOBA figure by league wOBA.)

That’s a really weird graph. Sure, the xFIP looks “right” but what’s going on with wOBA and FIP? Results-wise, the Braves have somehow been worse second TTO. Sure, they’ve made it back third TTO, and the xFIP suggests that this all may work itself out in the end, but even the xFIP for second TTO is odd — the fourth-best offense in baseball is letting pitchers produce above-average peripherals the second time through the order?

The 2019 Braves have gotten better results than average the first time through the lineup, and woe be upon any manager that lets their pitcher work through this lineup a third time. The second time through, though? Well, it’s weird. Mystifying, even. The three teams offensively ahead of the Braves at this point (Dodgers, Astros, Twins) don’t exhibit anything like this — if anything, their most pronounced feature is that their third TTO results fall off, probably because opposing managers have no choice but to lift their starters against these teams relatively early. The Braves, somehow, haven’t quite forced this situation (they’ve batted third TTO 457 times; the highest among the other three teams is the Dodgers at 416; league average is 407) but have made the most of it when it’s occurred. Are they simply lulling opposing starters and managers into a false sense of complacency, the way my cat sometimes does what my wife and I call “fake belly rub” pose? Well, probably not. But still, it’s weird. And now you know.

Weird Fact #2 — O-Contact and Z-Contact, All Backwards

What’s the basic tenet of batting? Take balls, swing at strikes, right? If you want anecdata that validates that truism, look no further than Freddie Freeman, who has made a living at the top of offensive leaderboards by doing this exact thing. The 2019 Braves have taken this ethos to another level.

(Note: Numbers in this section use the Baseball Info Solutions data hosted on Fangraphs. While there are many different plate discipline datasets, including ones from Pitch Info/Brooks Baseball and Statcast, they all mostly tell the same story — if this is of interest to you, I encourage you to cross-reference my analysis against the other datasets to see if there’s anything different or contradictory there.)

Just as a quick refresher:

  • O-swing: rate of swings at pitches outside the zone;
  • Z-swing: rate of swings at pitches inside the zone;
  • O-contact: rate of contact on pitches swung at outside the zone;
  • Z-contact: rate of contact on pitches swung at inside the zone.

If we plot a team’s o-swing versus its z-swing, the results are fairly predictable: teams that swing more swing at everything more, teams that swing less swing at everything less. There’s not too much of a value judgment here about how to go about having a good offense — both the Dodgers and Astros don’t swing much, while the Twins swing at everything. It’s the Braves, highlighted in red below, that are weird.

(Note: What’s a z-score, you ask? It’s just a normalized measure of a stat that tells you how many standard deviations from the mean it is. A 1.00 z-score means a team’s stat is one standard deviation above the mean, a z-score of -0.75 means a team’s stat is three-fourths of a standard deviation below the mean.)

As you can see, most teams fall within either the third (bottom left) or first (top right) quadrant. If you imagine a diagonal line that goes through both of them (the y = x line), you can see that most (but not all) teams are really close to it. While the Braves are not the furthest team from that line, they’re the team that’s furthest from the line in a way that makes good baseball sense.

  • The Tigers are technically the furthest from the line, but they’re the dot all the way over on the right. They swing at tons of balls (z-score = 2.48) but an average number of strikes (z-score = 0.40). The result? The majors’ worst offense by wRC+ so far.
  • The Marlins are technically the second-furthest from the line, but they’re all backwards. Their dot is the one in the second (bottom-right) quadrant, furthest to the right. They swing at tons of balls (z-score = 1.27), but actually fail to swing at strikes at a great above league-average as well (z-score for z-swing = -0.61). The Marlins have the majors’ third-worst offense.

So then, you get the Braves. The Braves take swing disparity to an extreme, but in a positive way. Their z-score for o-swing is -0.12, that is, they basically chase at an average rate. But their z-score for z-swing is 1.63, reflecting a league-best 72.5 percent mark. The team’s “net swing” (z-swing less o-swing) is a league-best 41.8 percent; the other teams above 40 percent are an offensive mixed bag featuring the Rays and Athletics (average-y offensive teams) as well as the Reds and Indians (well below-average offensive teams). So, the Braves have made this approach work for them beyond their compatriots in “aggressive selectivity,” and in slight contrast to the Dodgers (-1.28, -0.43 on the graph above), Twins (1.05, 1.36, pretty generic elevated swing rates), and Astros (-1.06, -0.91, a very generic patient approach). By the way, check out the Angels — a top 10 offensive club so far with by far the lowest o-swing and z-swing rates among all teams.

Alright, so, the Braves take balls and swing at strikes. That’s not weird. That’s just good. But wait, I said this was a weird fact, right? Well, here we go!

Below is the same chart, but this time for o-contact and z-contact. Unlike the swing rates, there’s less of a clear relationship between what is and isn’t a good contact mix. Z-contact is obviously good, because z-whiff is terrible for hitters. But o-contact is weirder — on the one hand it helps you avoid strikeouts, but on the other hand, putting balls in play on pitcher’s pitches seems rather ineffectual for good contact and outcomes. In any case, we can probably agree that of the two, what you really want is high z-contact if nothing else. And yet, and yet...

(Remember, the red dot is the Braves).

The Braves’ z-contact is... really quite bad. They’re over a standard deviation worse than the league. Their o-contact, meanwhile, is really high... but the only two teams ahead of them there (Angels, Astros) make tons of contact on everything.

Again, if you figure contact skill is true regardless of where the ball ends up being (imagine the y = x line again), the chart above makes sense — teams that make more o-contact tend to make more z-contact. There are a few outliers, for sure, but they’re mostly not egregious... except the Braves.

Now, in trying to reason through why the Braves were offensively successful despite making relatively more of the “wrong kind” of contact, here’s what I came up with:

  • Not all z-contact is good. Whiffing on a pitch in the zone can be advantageous if you would have make better contact on a different pitch. With the Braves having such a high z-swing, there may be some z-contact that’s inefficient and better off being a strike instead.
  • Not all o-contact is bad. Fouling stuff off and getting a better pitch to hit can be an advantageous strategy, especially if you’re the Braves and pretty selective about which balls you swing at in the first place.

An important piece of evidence to consider in the context of above is that the Braves’ xwOBACON (i.e., their quality of contact only, excluding walks and strikeouts) is the best in baseball. Yes, better than that of the Twins, even though the Twins get to use a DH, and the Braves have pitchers’ PAs baked in to their number. They are second to the Twins (again, DH vs. pitchers) in xSLG and barrel rate. In other words, we have to believe that whatever the Braves are doing in terms of their contact rates is leading to, or at least not interfering with, really good contact, and therefore the joint ideas of the Braves’ eschewing worse z-contact via whiff and avoiding weak-in-play o-contact make sense.

Or, at least, that was my attempt to reason through it. Looking at the data, though, the reality is a little weirder. Check it out.

From the above, we essentially see that as far as team offense goes, the biggest driver is a low o-swing. Weirdly/surprisingly, this matters but z-swing rates really don’t. Reducing whiff rates also matters, but this stems largely from making o-contact, rather than z-contact. It’s also interesting that fewer first-pitch strikes correlates with better offense, but this in some ways is not under the team’s control, but rather the purview of their opponents.

In any case, it appears that my attempted reasoning through all of the above has (weirdly or not) been shattered upon the shore of really basic correlations, kind of. From the above two scatterplots, we see that the Braves are: A) good at z-swing and okay at o-swing; and B) good at o-contact and bad at z-contact. From the above table, we see that being good at z-swing and z-contact doesn’t really matter, but being good at (not doing) o-swing and o-contact does. The Braves are okay at (not doing) o-swing and great at o-contact, so there you have it.

Just for the sake of comparison:

  • Dodgers — good at (not) o-swing, okay at o-contact;
  • Astros — good at (not) o-swing, great at o-contact; and
  • Twins — bad at (not) o-swing, okay at o-contact.

So, maybe the Twins are the weirdos here, not the Braves. Anyway, hopefully you found that interesting.

Weird Fact #3 — Pobrecito Braves

Unlike the other two weird facts, this one just kind of is. It’s not really specifically weird in a baseball sense, it’s just something that kind of hangs out there. Maybe it’ll change over the next half-season. Maybe it won’t. Either way, it is what it is right now.

As mentioned before, the Braves have the second-best xwOBA in baseball, behind only the Twins. However, their wOBA (including pitchers) lags somewhat, ranking fifth overall. In essence, this means the Braves’ lack of luck on balls in play has affected its ranking of offensive output. Not every team is going to have a wOBA equivalent to its xwOBA, especially not halfway through a season. And yet...

None of the teams ahead of the Braves in wOBA have even been unlucky. Of the teams suffering similar or worse luck than the Braves, all but the Braves have been horrid offensively. Weirdly enough, it actually looks like there’s some degree of correlation between a team’s offensive production, and the extent to which that offensive production deviates from its underlying contact quality. At a first guess, I’d actually expect this not to occur — with wOBA-xwOBA gaps being randomly distributed, there’s no real reason for them to correlate with anything. However, it does occur. Possible reasons could be: A) park effects, which aren’t picked up by xwOBA but serve to make batted balls events at Coors Field and similar stadiums overperformers while also inflating home team wOBA disproportionately; B) lack of defensive flexibility to cut down xwOBA when dealing with additional baserunners; and/or C) teams that employ better hitters are also more keen to employ hitters that don’t underperform their xwOBAs, i.e., hitters that aren’t shift-prone/etc. Either that or the Baseball Gods just want to empower the teams with actual good hitters (but probably not this).

In any case, this is all fine and dandy... except for the Braves, who currently have the sixth-worst xwOBA underperformance in baseball. Of the five teams (Giants, Tigers, Marlins, Blue Jays, Cardinals) with more rotten batted ball luck, four are the four worst offensive teams in the game, and the fifth (Cardinals) is also below-average offensively. And then you have the Braves.

Are there explanations for this? One place I immediately went to is homer propensity. Homers lead to xwOBA outperformance by definition; fewer homers, therefore, means less of an ability to outperform xwOBA, all other things equal. However, there’s actually not that much correlation between homer rates and xwOBA gaps, and moreover, the Braves are top 10 in homer rate right now, so they’re not exactly getting hurt in this regard because they don’t hit it out of the park sufficiently often.

Another explanation could be shifts, but a gander at Baseball Savant doesn’t lead to much joy here either. The three heavily-shifted (i.e., more than half the time) Braves hitters so far this year are Brian McCann, Freddie Freeman, and Matt Joyce. Yet, these three guys are not underperforming their xwOBAs, and across all PAs with the shift on, all three actually have better results (i.e., higher wOBAs) than when no shift is put on. Furthermore, McCann, who has been shifted the most often among Braves hitters, actually has a slightly higher wOBA on “infield-level” balls in play (launch angle of nine degrees or lower) with the shift on than his xwOBA.

In any case, it’s pretty cool-but-weird that the Braves have managed to hit so well for the first half of 2019 while also getting fairly unlucky in the process. Really, it’s all Dansby Swanson’s fault (read: not his fault at all), as his .032 xwOBA underperformance is a top-50 rotten luck mark in baseball among all players with 100+ PAs so far. Just imagine what might happen if that luck starts to turn, or if the Braves actually outperform their xwOBA for a change. Actually, you don’t need to imagine — that’s what June 2019 has been, with the Braves outperforming their xwOBA by 0.012 en route to by far the best offensive performance (.373 team wOBA) for the month, although it’s come with the best offensive inputs as well (.361 team xwOBA, next highest is the Twins at .348). More of the June 2019 offense? Sign me up for that.