Giving Away At-Bats
One piece from FanGraphs this season has stayed with me more than any other article on the website. In early September, Kevin Goldstein wrote a piece called The Rays’ Unique Ability To Mitigate Risk.
For most of the piece, Goldstein examined why the Rays pitch effectively even though they use so many relief pitchers. Most of the time, a team that cycles though relief pitchers in bunches is a bad one, like the Baltimore Orioles this year. But the Rays, as they often do, defy common practice.
I actually did not remember that part of Goldstein’s article; I only remembered it when I re-read it before writing this. What stuck with me was a short section at the beginning in which he explained why the Rays score so many runs.
Goldstein’s question was how does a team that has no high-priced free agent slugger, like Bryce Harper or Manny Machado, or no home-grown young stud, like Juan Soto or Fernando Tatis Jr., score so many runs? (You will see in a moment why I am ignoring the Rays’ young phenom Wander Franco.)
Goldstein’s answer is simple. While the Rays do not have Harper or Soto, they do have a deep roster of above-average hitters. In fact, they have more than any other team in the American League East. When the article was written, the Rays had 11 hitters on their roster who had at least 200 plate appearances and a wRC+ of more than 100. Now you know why I left out Franco: they had 11 hitters at that level as of September 9 even though Franco had yet to have 200 PAs.
The rest of the AL East pretty much follows the standings. As of September 9, the Red Sox had eight hitters with 200 PAs and a wRC+ of more than 100, the Blue Jays had seven, and the Yankees and the Orioles had six each.
My boyhood team, the Oakland A’s, finished the season with seven: Matt Olson, Starling Marte, Tony Kemp, Mark Canha, Ramón Laureano, Seth Brown, and Matt Chapman (barely with a wRC+ of 101). And Brown only lifted his wRC+ above 100 with two home runs on the last game of the season.
Goldstein’s answer as to why the Rays offense is so good is that they have no soft spot in their lineup. They can field nine players with a wRC+ of more than 100. In other words, even if you don’t have a superstar hitter to anchor your lineup, you can still have an excellent offense if it’s full of above-average hitters and if — and here’s the important part — you minimize the players in the lineup with a wRC+ of less than 100.
All of which brings me back to my beloved A’s. What stuck with me after reading Goldstein’s article is not how many PAs the A’s gave good hitters, but the opposite. How many PAs did the A’s give to hitters whose wRC+ was less than 100? Or rather, how many players did the A’s give substantial playing time to who, relative to the rest of the league, were easy outs? The answer is four.
Player | PAs | wRC+ |
---|---|---|
Elvis Andrus | 541 | 72 |
Sean Murphy | 448 | 99 |
Mitch Moreland | 252 | 93 |
Chad Pinder | 233 | 97 |
Two more players just missed joining the group above.
Player | PAs | wRC+ |
---|---|---|
Josh Harrison | 199 | 79 |
Stephen Piscotty | 188 | 78 |
One player on the A’s finished in limbo: Jed Lowrie had 512 plate appearances with a wRC+ of exactly 100.
The Rays have three players with more than 200 PAs and a wRC+ of less than 100. What about the A’s main division rivals? The Astros had three, but one is catcher Martín Maldonado, who is a big contributor defensively, and another was Myles Straw, whom the Astros traded away. The Mariners had five, with 670 PAs going to Kyle Seager and his 99 wRC+, and 377 PAs to Jarred Kelenic, who totaled a 73 wRC+.
The more I thought about giving plate appearances to weak hitters the more I realized that baseball really needs another stat, one which would measure a team’s GAAB%: its Giving Away At-Bats percentage. I define this as the percentage of a team’s total plate appearances (not counting those taken by pitchers) that went to hitters with a wRC+ of less than 100.
To compute these numbers, I only looked at a few AL teams and I did not include PAs by pitchers. I took PAs by position players with a wRC+ of less than 100 and divided it by the team’s total plate appearances.
Before I give the A’s GAAB% this year, I will give you some other teams’ GAAB% so you have a context for guessing Oakland’s percentage.
Team | GAAB% |
---|---|
Rays | 16.0% |
Astros | 17.7% |
Mariners | 36.7% |
What about the worst team in the American League this year, the Baltimore Orioles? They came in at 35.7%. How about a team that finished around .500, such as Cleveland? They gave 56.7% of their PAs to below-average hitters.
The Mariners’ GAAB% makes you wonder further if the team played above its head this year, while Cleveland has an astonishing 11 players with more that 200 at-bats and a wRC+ below 100. It makes you question how much worse the team’s record would be if not for the efforts of a handful of their talented pitchers.
Finally, the Oakland A’s came in at 24.2%.
No team will have a GAAB% of zero. Every team has a back-up catcher, utility infielder, or fourth outfielder who provides in ways other than offense. Every team calls up injury replacements from Triple-A. But the GAAB% quantifies why some teams score more runs than other teams from a perspective that fans sometimes overlook. The A’s GAAB% gives the team a yardstick to measure how much the lineup needs to improve to match the offense prowess of the Astros or the Rays. Teams like Oakland must transfer about 500 PAs from batters with a wRC+ of less than 100 to those that are above-average. How the team does that is part of the offseason debate.
Really interesting work. Very concisely written, with sprinkles of person touch/personality, and a narrative that wraps around the analysis beautifully.
One weird thing about the stat is it’s evaluating a team based on what a player will do in the future. To your very point… a player who hits close to 100 wrc+ all year but then either has an amazing or terrible final week will shift the GAAB% drastically.
It’s almost like you need GAAB to look at wrc PRIOR to the season. Or at least prior to the time when they’re in the lineup.
Still… great work!
Thank you for the feedback and compliment.
There is very little difference between someone just above or just below a 100. Sequencing matters at that point (ie. the Mariners were ver clutch in 2021). Plus, not facing the Orioles pitching cost the Orioles hitters points in wRC. So, I would only count players below 95 or find a way to normalize the stat for who they faced. Moreover, a guy could be an effective hitter until an injury then the guys performance could plummet relatively quickly. In other words, 300 AB at 105wRC then 50 AB at 0 wRC and you get a guy below 100. That’s not exactly the same as someone who is consistently below 100. Lastly, it’s not a very good predictor of future performance on most teams.
It is a nice historical stat though. I would argue its a way to measure a GMs effectiveness.
You are right about injuries causing a player’s wRC to drop and skew his end of the year wRC. Jed Lowrie is a good example of that this year. He was above average until he played hurt in September and his wRC dropped to 100. I agree with your last point too. Past performance is not a guarantee of future results. Still, it gives you a goal in terms of building a roster. If you want a championship caliber offense, you have to build a roster that has a good chance to give away only 16%-18% of its at bats.
Nice article. I think this should be heavily weighted when thinking about law of regression and how as an example it could possibly show just how Seattle may come back to earth in # of wins. This among other components like pitchers added and subtracted, injuries, prior season workloads. Is there a way to see the team by team listing?
I agree with all your points. The article makes two main points. One, a team does not necessarily have to have a superstar hitter to build its line-up around. A team with a small budget, like the Rays, can still build a playoff competitive team if it has a lot of above average hitters and if (here is the second point) it minimizes the at-bats it gives to poor hitters. That makes all your points crucial. How, for example, does Seattle project Jarred Kelenic next year? If playoff caliber teams only give 16-18% of their at bats to bad hitters, how many of those at bats does Seattle want Kelenic to take? What does Seattle think Kelenic’s wRC+ will be next year.? Projecting that is a skill that I do not have. But Seattle fans should hope that the team’s front office does possess it.