How Meaningful are Quality At-Bats?
When you’re a fan of a bad team, you often find yourself searching for optimism in very odd places. For me, this means watching countless young Rangers strikeout, pop out, or ground out but saying “that’s okay, at least he had a good at-bat” and hoping that particular plate appearance was somewhat meaningful in the player’s development.
This perpetual struggle spawned the idea for this article. Quality at-bats aren’t officially tracked at the professional level, but they’re a common tool used by coaches and players in non-professional baseball. The idea behind quality at-bats is simple: every one of your plate appearances should be productive in some way. Most of the time, being productive involves getting on base in some way (obviously, that’s like the whole point of baseball). But quality at-bats also track hard-hit balls and long plate appearances. Hard-hit outs don’t always help the team, but they’re a signal that a player saw the ball well and put a good swing on it, and that’s always a good thing. Likewise, long plate appearances don’t always result in the batter reaching base, but it shows the ability to fight off tough pitches and be a difficult out, in addition to making the pitcher throw more pitches and exert himself more — again, all good things. The full list of criteria for a quality at-bat, for the purposes of this work, is defined like this:
- A hit
- A walk
- A hard-hit ball (using the Statcast definition of a ball hit at least 95 mph)
- A sacrifice hit
- A plate appearance that lasts at least 6 pitches
- An HBP
There are, of course, many different definitions in many different places. Some people include reaching on an error. Some favor recording a Barrel instead of just a hard-hit ball. Personally, I prefer this list, and there’s nothing you can do to stop me. And yes, I know they should technically be called Quality Plate Appearances, but that just doesn’t sound right.
With all that out of the way, let’s see how meaningful quality at-bats really are.
Year-by-Year Breakdown
First, a note on methodology. For this study, I pulled the last pitch of every at-bat from the “Statcast era” — 2015 to 2020. I limited the study to this era because it’s the beginning of when exit velocity data is available. I used Bill Petti’s wonderful Baseballr package for R to scrape this data on a season-by-season level, and I used R for all of the research and analysis. The full code, for anyone interested, is on my Github.
Naturally, the first thing I did was develop a yearly leaderboard for Quality At-Bats in each season. The results are displayed below.
Each year’s QAB% leaders are usually just that year’s best hitters, but there are some incredible surprises that are worth investigation. The ones I really care about are the extreme outliers: Joe Mauer in 2017, Kendrys Morales in 2018, Lorenzo Cain in 2019, and Tommy Pham in 2020. Are these just outliers, or are they underrated kings that make me look like a genius?
Let’s not brand me a genius just yet. Upon investigation, it looks like these players were all really tough outs at the plate. The four “outlier” players each posted long (6+ pitches) plate appearances more often than the average player. Additionally, all but Lorenzo Cain had an above average walk rate, and all but Tommy Pham had an above average strikeout rate. Another separating factor: these players had exceptional hard-hit rates. Perhaps most importantly, however, is the fact that each one of these hitters drastically underperformed in terms of expected wOBA — unsurprising, as players who hit the ball hard often tend to underperform in that regard.
So what does this mean? Well, obviously, these players weren’t exceptionally productive. But they were all experienced hitters who worked the count, got on base, and hit the ball hard often, and that’s exactly the kind of player you want in your lineup every day. More importantly, these cases illustrate that quality at-bat percentage can highlight players who are having good at-bats often, even if that doesn’t always translate to productive offense.
Outliers aside, QAB% does correlate strongly to various offensive metrics. Below are the correlations with wOBA, xwOBA, and OPS.
The correlation isn’t surprising. Quality at-bats and these three stats all essentially track the same things, and good hitters will be rewarded in all three metrics (for the most part, as seen above). Perhaps a more interesting question, then, is whether changes in quality at-bat percentage from year to year correspond to yearly changes in offensive metrics. To answer that inquiry, we turn to more scatter plots:
Once again, there’s a strong correlation, and that’s not surprising. A player who’s improving his QAB% is getting more hits, drawing more walks, and hitting the ball hard more consistently, all of which are reflected in wOBA and xwOBA. So in terms of a player’s development, it’s basically equivalent to look at those two stats or a player’s quality at-bat percentage, as they’re essentially measuring the same things.
The Long At-Bats Question
Are there any advantages to quality at-bats, then? Well, the one thing quality at-bats track that wOBA and xwOBA don’t is long at-bats. So when we ask “how meaningful are quality at-bats”, in a sense, we’re really trying to see how meaningful long at-bats are, since wOBA and xwOBA can tell us how meaningful all the other stuff is. It may seem reductive to put a six-pitch strikeout on the same level as a 450-foot home run, and it probably is, but long at-bats do have some value. Qualitatively, being a tough out and making the pitcher throw as many pitches as possible will always be valuable on both an individual and a team level. A team full of players that work long at-bats will get the starting pitcher out of the game sooner, which is a good thing. A player who works an 8-pitch single will be praised much more than a player who hits a single on the first pitch of an at-bat. Also, if you’re going to strike out, better to do it on six pitches than three or four pitches.
Quantitatively, long at-bats are a high-risk, low-reward proposition. These sorts of battles are quite difficult for pitchers and hitters, but it appears they lean slightly in the pitcher’s favor. 21 percent of plate appearances that last at least 6 pitches end in a walk, whereas 31 percent of these plate appearances end in a strikeout. Both of these figures are much higher than their short plate appearance counterparts. The overall slash line for long at-bats is 0.200/0.370/0.335 and 0.251/0.301/0.415 for short at-bats. The implications here are interesting — generally, more hittable pitches are thrown earlier in counts, but a hitter in a long at-bat is more likely to see a pitcher miss out of the zone after such a long battle Additionally, as seen by the inflated strikeout rate, hitters become much more likely to whiff as an at-bat goes on. The lower slugging percentage also reflects the higher quality of pitches seen in these prolonged two-strike counts, and these are often much harder to drive for extra-base hits. The large amount of walks nearly make up for it, however, as the OPS between the two situations is nearly identical.
While long at-bats may not be especially conducive to high-value outcomes, they’re still quite valuable to a team. There’s obviously the fact that one-fifth of 6+ pitch plate appearances end in walks, but I think the qualitative portion is just as important. Just three of these plate appearances in an inning already puts a pitcher at 18 pitches in an inning, and regardless of the outcome, that’s on the high end of a workload for an inning. When a pitcher has to throw that many pitches in an inning, it’s more likely he’ll make a mistake, at least from a mathematical perspective.
Conclusion
Now, you may be thinking: gosh, Austin, QAB% sure sounds like it’s unnecessary when wOBA, xwOBA, wRC+, and WAR already exist and measure overall productivity much better than this fake stat you’re clinging to from your childhood. And you’re probably right. I thought of this too. So, to finish this piece off, I was going to present a better way to think about quality at-bats. I would’ve given each plate appearance an importance based on the runners on, number of outs, and the result, and used that to make a weighted number that reflected who had quality at-bats in important situations most often. What a great idea, I thought. That is, until I checked Fangraphs, and realized I had just re-invented RE24. Way to go, Austin. Anyway, RE24 measures the productivity of a hitter based on how they change the run expectancy of an inning. It’s a bit complicated, but it basically rewards a hitter more for a bases-loaded double than that same double hit with nobody on base, whereas other stats like wOBA treat the hit equally (“context-neutral”, they’re called).
The top hitters by RE24 each year are usually just the year’s best hitters, as they’re the guys who most often hit with guys on base, and they’re the guys who are most capable of those game-changing extra base hits with guys on. It’s not the major revelation I thought it would be. Turns out being a good hitter is more important than anything else.
So what are we to make of quality at-bats? I like to think of it as a “good outcome percentage”. While it essentially measures the same exact things as wOBA and xwOBA, perhaps it may be easier for a casual fan to understand than those two advanced metrics. It’s an extremely simple way to gauge a player’s acumen, and while there are exceptions to that, those exceptions can sometimes be players who are just underperforming at the plate while still working good at-bats.
I can’t say there’s any real statistical advantage to QAB% over the plethora of advanced metrics available today. If you’re looking for productive hitters, wOBA and wRC+ are your friend. HardHit% is already tracked on Baseball Savant. On-base percentage is easy to find. And for players who are seeing the ball well but aren’t having good results, there’s the “x” family of stats — xwOBA, xBA, xSLG, and more. There’s the long at-bat factor, which is interesting, but ultimately is just analogous to BB%, which is included in the more comprehensive offensive metrics. So while Quality At-Bats are interesting to think about, and a nice way to gauge the amount of productive plate appearances for a player, there’s a reason they aren’t relied on for serious statistical analysis or player evaluation.
Thanks for reading. Find me on Twitter, or my Rangers Analytics Twitter, if you’re into that kind of thing.