Tag Archives: run expectancy

Ramichanalytics Part II

Last week I posted DeNA manager Alex Ramirez’s responses to questions about batting his pitchers eighth. Here is my analysis of how that’s working out for him.

I’ll give you a hint: The haters probably need to shut up or at least think before they tell you what’s wrong with it.


The experiment began on April 14, 2017, when starting pitcher Joe Weiland batted at eighth at home against the Yakult Swallows. Weiland struck out twice, but the BayStars beat the Swallows 4-3. Three weeks later, Ramirez tried again, batting Weiland eighth on the road against the Giants. He went 0-for-3 with two more strikeouts in a 5-2 win at Tokyo Dome.

After that, every DeNA starting pitcher in the lineup batted eighth until the end of the 2018 season, when having finished out of the postseason for the first time in his four years in charge, Ramirez suggested batting his pitchers eighth was a thing of the past.

The manager’s rationale

To summarize briefly what Ramirez said, he argued that the value of having a pitcher — who must be a competent bunter — is to sacrifice runners on base so that the No. 9 hitter can drive them in with a single.

A sabermetric rationale

In their “The Book,” Tom Tango, Mitchel Lichtman, and Andrew Dolphin, reason that over the course of a 162-game season, a major league team can gain about 2-1/2 runs a season by batting the pitcher eighth.

Why? Because replacing your worst hitter in the ninth spot with a half-way decent hitter gives your best hitters at the top of the order a much better chance to hit with runners on base. The logic is that outweighs the pitcher’s lessened ability to finish an offensive sequence started by the middle of the order, and giving the No. 9 hitter lessened RBI opportunities than he’d have in the No. 8 spot.

How can we evaluate the experiment?

This is kind of a tricky puzzle and I don’t pretend to have the answer, but there are a few different ways of looking at it. I’d tried looking at how many RBIs and runs were being scored in the different lineup spots, but the problem is that the BayStars had very few games in 2017 when the pitcher batted ninth. We could compare 2016 when none did, or 2018 when everyone did, but those run contexts were quite different.

We’re also faced with the reality that from 2017 to 2018, the only time BayStars pitchers batted ninth was in April when fewer runs are scored.

I settled on using run expectancies. Batting your worst hitter eighth should decrease run expectancies for the No. 9 hitter, but increase them for the No. 1 hitter — who ostensibly should be one of your better hitters.

The questions then are: How much does batting pitchers eighth…

  1. Increase the run expectancies for the No. 1 hitters?
  2. Decrease the positive results from the No. 8 spot?
  3. Decrease the run expectancies for the No. 9 hitters?
  4. Increase the positive results from the No. 9 spot?

Since it’s silly to compare offensive numbers from March and April with offensive numbers from May to October when runs in BayStars games in 2017 were 4.5 percent more frequent, I’ve decreased the run expectancies and average production for all the hitters in the lineups with the pitchers batting eighth by 4.5 percent to level the playing field a bit.

Cleaning up in the ninth spot

According to this analysis, the BayStars made out like bandits by batting the pitchers eighth in 2017 because the No. 9 hitters were extremely productive. Here’s the breakdown to the four questions above (including the run adjustment mentioned above):

  1. BayStars leadoff hitters came to the plate with a run expectancy 0.0067 runs higher over the course of 463 plate appearances, adding about 3.11 runs to the team’s expected scoring.
  2. No. 8 hitters (pitchers and pinch-hitters) produced 0.039 runs fewer per PA over 382 plate appearances, a total decrease of about 14.8 runs over the season.
  3. No. 9 hitters came to the plate with a run expectancy of 0.36 runs less, for a decrease of about 13.3 runs over the season.
  4. No. 9 hitters (position players) produced 0.096 more runs per plate appearance than the pitchers and pinch-hitters who previously occupied the No. 9 spot for a total of about 35 runs over the course of the season.

In summary, those figures are:

  1. +3.11 runs — Setting up the leadoff hitter:
  2. -14.8 runs — Lost production from the No. 8 spot
  3. 13.3 runs — Lost run expectation ahead of No. 9 spot
  4. +35 runs — Gained production from the No. 9 spot.

The estimated total for the 2017 season was + 10.03 runs.

One of the unexpected benefits of having the pitcher bat eighth is bringing in a pinch-hitter slightly earlier. Normally, pinch-hitters add nothing to the expected run output, but replacing a pitcher with a pinch-hitter is a big plus.