Intro to Small Samples 101
Through the first eight games of the 2021 season, nobody in the majors was worse at scoring than the Oakland A’s. They managed only 19 runs, slightly over two per game, ranking last out of 30 teams.
However, a peek under the hood suggested things weren’t as bad as they looked. There were a few hot bats, and the overall group was making reasonably hard contact and not striking out too much, and they did face a few Cy Young winners along the way. In a community poll, the majority of Athletics Nation agreed things would get at least slightly better right away, and around half were confident the lineup would return to normal output within the next week.
To be clear, that’s not the same as saying everyone on the A’s will definitely have an awesome full season. Just that we can now write off the worrisome early small-sample at the full-team level, and reset back to whatever normal expectations we had for each hitter. Nobody forgot how to play over the winter. If you thought last month that a particular player was going to be good, then you can return to regularly scheduled optimism.
There are also still a few individuals who haven’t settled in yet, and it’s always possible they never will. But as an overall group, they’re back in business.
MLB Ranks (out of 30 teams)
- 15th in xwOBA
- 16th in Exit Velocity
- 17th in Strikeout rate (or 14th-highest)
They didn’t even hit the ball any harder, as their EV went down by a couple tenths, and their strikeouts remained basically the same. But a couple extra homers found seats instead of hooking foul, and on the real-world side more of the successes were sequenced together properly (6-for-11 with runners in scoring position on Fri/Sat) instead of being scattered harmlessly throughout the games.
The hot hitter list has grown. In real terms, here’s everyone who rates average or better so far:
- Ramon Laureano, 158 wRC+
- Mark Canha, 156 wRC+
- Jed Lowrie, 132 wRC+
- Matt Olson, 108 wRC+
- Stephen Piscotty, 100 wRC+
- Seth Brown, 100 wRC+
And in Statcast terms, here’s who has the best underlying performances. League average is .333 right now, which is a bit higher than a normal full season:
- Jed Lowrie, .441 xwOBA
- Mark Canha, .437 xwOBA
- Stephen Piscotty, .402 xwOBA
- Ramon Laureano, .401 xwOBA
- Seth Brown, .355 xwOBA
- Matt Olson, .349 xwOBA
For context, among MLB players with at least 20 plate appearances, Lowrie ranks 32nd, Canha 33rd, and Piscotty 55th, out of 256 hitters. And while Olson trails this particular list, he does lead the team with a 94.2 mph exit velocity, which is Top 20 in the majors (min 20 PAs). We all saw what happened when Olson finally found the barrel on Friday.
Also remember that, for all intents and purposes, Piscotty homered off Clayton Kershaw (109.8 mph EV, 408 feet), and Brown homered off Trevor Bauer (98.8 mph EV, 364 feet). But they both went foul by a couple feet, so they aren’t included in the metrics above. That kind of thing evens out over a full season, but in a tiny 10-game sample it’s highly relevant, especially when it’s not the only great contact they’ve made. If you want to analyze how they might perform moving forward, you’re gonna want to know about those rockets regardless of what the box score says.
That’s the good news. The A’s finally scored some runs, Statcast says the bad days were almost certainly a fluke based on how everyone has been swinging, and there are at least six scalding-hot hitters who can all fit in the same lineup if desired (DH, 1B, 2B, three OF). And now we get to see them face last-place pitching from Arizona and Detroit instead of World Series rotations.
Now for the even better news! There is actual reason to believe things will continue to get better, and that the most recent two-game sample is only the tip of the iceberg. And I don’t just mean “we expected them to be good so let’s selectively cling to the happier and more recent sample.”
Statcast’s xwOBA metric is a theoretical “expected” stat, modeled to look like a real-life wOBA. For those more familiar with OPS, it would be like saying, “this is what his OPS should be based on how well he’s actually hitting the ball.” That means you can compare the expected stat against the real one to see how reality lines up with true talent, like with a pitcher’s ERA and FIP.
Nobody has been unluckier than the A’s so far. Their .328 xwOBA is right around average, but it’s only resulted in a .265 wOBA on the field. For perspective, that’s roughly what Jonathan Lucroy (.271) and Dustin Fowler (.263) each posted in 2018. Oakland has swung like an average team, but produced like a light-hitting catcher or a prospect who washes out back to the minors.
The .063 difference between the A’s marks is the highest in the majors, by far. The distance from them to second place is around the same as the difference from second place to sixth, and no team has ever finished a 162-game season worse than minus-.012 in this regard (in the short 2020 the worst was minus-0.18).
Add in that the worst full-season wOBA during that span is .288 (the 2019 Marlins), and that the 2021 A’s probably aren’t the worst-hitting team of the last half-decade and so their totally normal xwOBA probably isn’t the fluke in the equation, and it becomes nearly a statistical fact that their results will continue to catch up. In other words, we will see more days like Friday and Saturday at the plate, and far fewer like the first eight snoozefests.
What’s more, on an individual level, not a single hitter on the A’s is outperforming his batted balls, as everybody’s xwOBA is higher than their wOBA. Everyone with good numbers is truly earning them, and everyone with bad numbers is either truly not hitting well or is quietly getting unlucky. Nobody has a big OPS based on lucky bloops or cheap singles, and nobody has a triple-slash batting line (avg/obp/slg) better than that which they could reasonably sustain all summer.
Again, all of this isn’t green-and-gold-tinted Alex assuring you everything is probably fine. It’s Statcast telling you hard data, in no uncertain terms. The A’s lineup is not bad, and it’s probably good, just like we thought two weeks ago.