We saw the Braves experiment with batting the pitcher 8th for a few games. The idea was obviously to get more PAs for Albies/Acuna, and to have Ender on base more often for those guys after the 1st inning.
What seemed to happen was the pitcher came up to bat several times with men on base, and killed potential rallies. We can use BRef's play index to see what actually happens with 8th place hitters, and we will limit the data to 2017-2018 so the data doesn't bog down the searches too much.
First, let's take a look at the average MLB #8 hitter (including pitchers who only represent 1.3% of the sample...batting the pitcher 8th is rare):
PAs: 23265
Any On PAs: 10281 (44.2%)
non-self RBI: 1728
non-self RBI per Any On PA: 0.168
So a typical #8 hitter bats with any runner(s) on base 44% of the time. They cash in these runners at a rate of 0.168 runs per PA.
Now let's take a look at pitchers batting 8th:
PAs: 322
Any On PAs: 145 (45.0%)
non-self RBI: 16
non-self RBI per Any On PA: 0.110
Pitchers batting 8th see runners on base at exactly the same rate as non-pitchers batting 8th (so much for lineup protection). Every time a pitcher batted with runners on rather than the average #8 hitter, it cost the team 0.058 runs. Since the 9th batter will still hit 2/3 of the time (when the pitcher doesn't make the 3rd out), we can divide that by 3. The cost is now 0.019 runs every time the pitcher bats with runners on base rather than the 8th place hitter.
In 2017, NL teams sent a pitcher to the plate on average 328 times, mostly from the 9th hole. If a team batted the pitcher 8th, we can add a handful of PAs (17 per lineup spot, but the pitcher won't hit the entire game), so let's call it 340 PAs. In those 340 PAs, we can assume 45% will be with runners on base, or 153 PAs. Each of those PAs cost the team 0.019 runs, for a total of 2.9 runs for the entire season.
That seems like a very small price to pay for the supposed benefit of Albies/Acuna batting with more runners on base. Determining how many more runners are on in front of those hitters is hard to gauge, but we can see how many more runners were on in front of the lead off guy for the teams who batted the pitcher 8th most often: Cubs (124 PAs) and Mets (60 PAs).
But first, the typical lead off batter hit with runners on base 9165 times in 27763 PAs (33.0%).
The Cubs lead off guy hit with runners on base 314 times in 933 PAs (33.7%).
A typical lead off hitter bat 750 times in a season, and batting the pitcher 8th gets him about 5 more PAs with runners on base.
So giving up 2.9 runs to let Albies bat 5 more times with runners on?
As usual, this lineup up stuff doesn't matter much at all.
What seemed to happen was the pitcher came up to bat several times with men on base, and killed potential rallies. We can use BRef's play index to see what actually happens with 8th place hitters, and we will limit the data to 2017-2018 so the data doesn't bog down the searches too much.
First, let's take a look at the average MLB #8 hitter (including pitchers who only represent 1.3% of the sample...batting the pitcher 8th is rare):
PAs: 23265
Any On PAs: 10281 (44.2%)
non-self RBI: 1728
non-self RBI per Any On PA: 0.168
So a typical #8 hitter bats with any runner(s) on base 44% of the time. They cash in these runners at a rate of 0.168 runs per PA.
Now let's take a look at pitchers batting 8th:
PAs: 322
Any On PAs: 145 (45.0%)
non-self RBI: 16
non-self RBI per Any On PA: 0.110
Pitchers batting 8th see runners on base at exactly the same rate as non-pitchers batting 8th (so much for lineup protection). Every time a pitcher batted with runners on rather than the average #8 hitter, it cost the team 0.058 runs. Since the 9th batter will still hit 2/3 of the time (when the pitcher doesn't make the 3rd out), we can divide that by 3. The cost is now 0.019 runs every time the pitcher bats with runners on base rather than the 8th place hitter.
In 2017, NL teams sent a pitcher to the plate on average 328 times, mostly from the 9th hole. If a team batted the pitcher 8th, we can add a handful of PAs (17 per lineup spot, but the pitcher won't hit the entire game), so let's call it 340 PAs. In those 340 PAs, we can assume 45% will be with runners on base, or 153 PAs. Each of those PAs cost the team 0.019 runs, for a total of 2.9 runs for the entire season.
That seems like a very small price to pay for the supposed benefit of Albies/Acuna batting with more runners on base. Determining how many more runners are on in front of those hitters is hard to gauge, but we can see how many more runners were on in front of the lead off guy for the teams who batted the pitcher 8th most often: Cubs (124 PAs) and Mets (60 PAs).
But first, the typical lead off batter hit with runners on base 9165 times in 27763 PAs (33.0%).
The Cubs lead off guy hit with runners on base 314 times in 933 PAs (33.7%).
A typical lead off hitter bat 750 times in a season, and batting the pitcher 8th gets him about 5 more PAs with runners on base.
So giving up 2.9 runs to let Albies bat 5 more times with runners on?
As usual, this lineup up stuff doesn't matter much at all.