Moneyball? Try Powerball. Even the A's are swinging away

BySam Miller ESPN logo
Thursday, April 27, 2017

On April 6, Oakland's Khris Davis had a 3-0 count, got a pitch he liked and swung at it. Four days later, he had a 3-0 count, got a pitch he liked and swung at it. Three days after that, he swung again on 3-0. Earlier this week, he did it twice more.

So far this season, Davis has swung at 3-0 pitches with a frequency about 100 times that of the average hitter. You might wonder what this says about Davis or about the A's. You might be right to wonder.

Most baseball tactics are considered "stathead" or "traditional." To oversimplify: Statheads don't pitch out, and they bat their best hitters second in the lineup; traditionalists use closers rigidly and sacrifice bunt. Those are all caricatures -- exceptions abound and "rules" change -- but we know where on the stathead/traditionalist spectrum most tactics land.

However, 3-0 swings have recently confounded this easy sorting.There is seemingly no correlation between green lights given and a club's stathead reputation. Joe Maddon has always given a lot of green lights; the A's have always not. In fact, the A's, more than any other team, represented the position that swinging at 3-0 was a waste of a perfectly good 3-0 count. Ten years ago, the A's swung three times on 3-0 as a team in all of 2007.

"We typically don't allow guys to swing 3-0," David Forst, now the A's general manager, said that year. In 2014, Josh Donaldson joked (perhaps?) that Billy Beane would be upset with him if he homered, rather than walked, on a 3-0 count.

But a clearer picture is coming into focus, and this year's 3-0 swing rates, Oakland's in particular, are helping. So far, 105 batters, including light-hitting AthleticsJed Lowrie and Trevor Plouffe,have swung on 3-0. That works out to swings on 9.6 percent of all 3-0 pitches. Is that a lot, historically? Yes and no:

You'll notice that trough in the mid-2000s, when 3-0 swing rates halved from a decade earlier. Those were the years right after Moneyball, when analytics were spreading rapidly across the league. Each era of analytics has its own gold rush, and that era's was plate discipline. These teams valued not just OBP but the ability to grind down a pitcher, to work counts and get starters out of the game. Kevin Youkilis became the icon of the era; he didn't swing at a 3-0 pitch his entire career.

Then the line started moving back up. (Oddly, Forst predicted that it would in that same 2007 interview.) Patient hitters saw plate discipline not just as a means of drawing a walk, though those are nice, but as a way to get into friendly counts, when they could count on fastballs in the fat part of the zone. By swinging more often, they forced pitchers to be a little more cautious on 3-0. Meanwhile, as bullpens swelled with effective relievers, the benefits of getting a starter's pitch count up became irrelevant. After seven years moving gradually upward, the 3-0 swing rate is now pretty close to that of the pre-Moneyball days. This is not the first time, incidentally, that analytics have spun around a few times and ended up right back where baseball strategy started.

How do we feel now when a hitter swings at 3-0? If he homers, as Davis did on April 10, we feel pretty swell. If he grounds into a double play, as Davis did on April 13, we're pretty frustrated. There's no "always" answer, which is increasingly true of almost every tactical philosophy. After tilting toward dogma for a few years, the 3-0 swing has settled into what it probably should be: an occasional tool in a hitter's toolbox.

When is it no longer 'early' in the season?

At a certain point in the season, we'll have to stop saying everything is a small sample. When will that be? The sabermetric writer Russell Carleton has done the math to find the stabilization rates for all sorts of key metrics, from swing rates to exit velocity to batting average. Here's what we can say this far into a season:

  • Swing rate stabilizes at about 40 plate appearances. That means we can say with some confidence that Chris Davis really has been much more aggressive at the plate this year, swinging at 52 percent of pitches after swinging at 43 percent last year. That's the biggest jump of any hitter, and it comes after Davis was uncharacteristically patient last year.
  • A hitter's strikeout rate stabilizes at about 60 plate appearances. That means we can say that preseason sleeper Keon Broxton really has been lost at the plate, with 25 strikeouts in 65 plate appearances -- the third-worst frequency in baseball.
  • A hitter's walk rate stabilizes at about 120 plate appearances. That means we can almost say that breakout slugger Miguel Sano has added a new skill to his game. Sano walked 54 times last year, a healthy but unexceptional free pass rate. This year, he leads the majors with 18 in 84 plate appearances, nearly double last year's rate.
  • A hitter's exit velocity stabilizes at about 40 balls in play. That means we can say -- well, hey, it's Sano again, with an average exit velocity more than 2 mph faster than that of any other qualified hitter.

We can say all these things, but what do they mean? These stats don't necessarily represent these players' true talent or what they'll do going forward. As Carleton warned in an email: "Rule #1 -- every statistic faces backwards." Baseball players go through hot and cold streaks, just as we all do.

But stabilization points mean we can reasonably conclude a player is the primary cause of his statistical performance, that his stats are due to him -- not to noise and circumstances. For instance, a hitter's swing rate is heavily influenced by how many strikes he is thrown and by the quality of the pitcher. After a certain amount of plate appearances, though, the hitter's personality itself is the strongest variable.

So that's where we are: We have learned a few things. We have a lot of small sample songs still to sing.

Keuchel doesn't need no stinkin' strike zone

In his complete-game victory Tuesday, Dallas Keuchel got 30 called strikes. According to the deep-voiced narrator, that's the most in a start in the Statcast era:

After that start, Keuchel led the majors in Baseball Reference's wins above replacement. The most remarkable feature of his success, though, is that he has walked only nine batters in 37 innings despite throwing a record-low number of pitches in the strike zone. Fewer than one in three of his pitches -- 31.7 percent -- finds the rulebook zone.

If it held up for a full season, that would be the lowest rate by a starter since at least 2008, when full PITCHf/x data became available -- by a lot. The lowest zone rate by a starting pitcher since then is 35 percent, by Tom Glavine in 2008 and Derek Lowe in 2011. Last year, Kyle Gibson threw the fewest strikes, at 40 percent. Keuchel, who lives at the bottom of the zone to induce groundballs with his sinker, was at 43 percent.

In most cases, a swinging strike is more telling than a called strike. A swinging strike tells us that a pitcher has swing-and-miss stuff; a called strike might tell us that the umpire did him a favor or that his catcher framed well. But even after accounting for the umpires, batters and catchers involved, Keuchel has "earned" 8 percent more called strikes than an average pitcher would, according to Baseball Prospectus.

He has found an extreme way of pitching, and we shouldn't be completely surprised if he continues to find extreme ways to succeed.