Blue Jays: Sacrifice bunts and stealing third

OMAHA, NE - JUNE 24: Jordan Nwogu #42 of the Michigan Wolverines gets tagged out at third base in the first inning by Austin Martin #16 of the Vanderbilt Commodores during game one of the College World Series Championship Series on June 24, 2019 at TD Ameritrade Park Omaha in Omaha, Nebraska. (Photo by Peter Aiken/Getty Images)
OMAHA, NE - JUNE 24: Jordan Nwogu #42 of the Michigan Wolverines gets tagged out at third base in the first inning by Austin Martin #16 of the Vanderbilt Commodores during game one of the College World Series Championship Series on June 24, 2019 at TD Ameritrade Park Omaha in Omaha, Nebraska. (Photo by Peter Aiken/Getty Images)

One of the MLB rule changes for the 2020 season will be that the top and bottom of all extra innings will start with a runner on second. Should a team employ sacrifice bunts and the stealing of third to win games that last beyond the ninth inning?

Of the 208 games that went into extra innings in 2019, approximately one-quarter of those games had doubles hit with no outs after the ninth inning. Accordingly, although the runner-on-second-in-extra-innings scenario is not new, Field Managers will have to address the situation in every extra-inning game in 2020. Two tactical options are sacrifice bunts and stealing third.

The subject of sacrifice bunts has long been a debating topic between the analytics/Sabermetrics community and old-school advocates. The value of stealing bases is also a topic that often lands on the debate schedule. However, the new extra-innings rule creates a particular situation such that the subject of sacrifice bunts and base stealing is worth re-visiting.

Spoiler alert

In specific circumstances under the new rule, the use of sacrifice bunts and the stealing of third base is justifiable.

More from Toronto Blue Jays News

Overview

I am tired of writing runner on second in extra innings and variations of the same. Therefore, I will create the acronym ROSIE (runner on second, innings extra).

One of the reasons why MLB implemented the ROSIE rule was to expedite the ending of games given the compressed 2020 schedule. Accordingly, we can focus upon the effectiveness of sacrifice bunts and steals of third base in terms of the impact on a team’s chance of ending the game. Put another way, assess the usefulness of these two tactics in terms of winning the contest.

Win Expectancy

Tom Tango of MLBAM, a subsidiary of MLB, produces the Win Expectancy table on his website, Tangotier.com. What is Win Expectancy? MLB.com defines it as follows:

Win Expectancy (WE), otherwise known as Win Probability, indicates the chance a team has to win a particular game at a specific point in that game.Expressed as a percentage, Win Expectancy is calculated by comparing the current game situation — with the score, inning, number of outs, men on base and run environment all considered — to similar historical situations. More specifically, the percentage is derived from the number of teams that faced a comparable situation in the past and went on to win the game.

Why is the Win Expectancy resource ideally-suited to help analyze bunts and steals in the new extra-innings situation? Because it checks all the boxes.

  • Indicates the chance a team has to win a particular game at a specific point in the game
  • Compares the current game situation to similar historical situations
  • Considers the score, inning, number of outs, men on base and run environment
  • The percentage is a product of the teams that faced a comparable situation and went on to win the game

To quote Kramer, Giddy Up!

The ROSIE version

Tango adapted the Win Expectancy Table for use in the ROSIE situation (see Table 1). There are some items of note.

  • In the top of the tenth, the away team has a 50% chance of winning (see Row 3, Column 1)
  • If the away team does not score in their half of the inning, the home team has an 81.2% probability of winning (Row 11, Column 1)
  • The reason why the home team’s odds of winning is higher than 50% to start the inning is that Tango’s probability model showed that runs occurred in 72% of tenth innings
  • The experience of the minor leagues, who used the ROSIE rule in 2018 and 2019, supports the 72% figure
  • Therefore, Tango had to adjust the table to reflect that, if the away team did not score, the home team’s probability of winning must be higher than 50%
  • This adjustment is also consistent with the home team’s probability of winning in the bottom of the ninth in a tie game (63.4%)

To demonstrate how to use Table 1, consider the following situation. The home team comes to the plate in the bottom of the tenth in a tie game. The club successfully executes a sacrifice bunt, which lands a runner on third with one out. The team has increased its chance of winning from 0.812 (Row 11, Column 1) to 0.832 (Row 13, Column 2). If the team instead stole third base, the club improved its probability of winning from 0.812 to 0.919 (Row 13, Column 1).

The probabilities in Table 1 reflect the home team’s perspective. For example, when the away team has a man on third and no outs in a tie game (Row 5, Column 1), the probability of winning for the away team is 0.579 (1.000 less 0.421).

General comment

Table 1 should only be used as a guideline because it depicts probabilities given the performance of the average batter, pitcher, defender, and baserunner.

Consider the situation where there is a runner on first base with no outs in the bottom of the eighth. The game is tied, Mike Trout is next up, followed by Al “Whiff King” Jones and Jimmy “Pop-Up” Clark. In this case, don’t ask Trout to attempt a sacrifice bunt because of the probability of winning declines with a successful sacrifice. Also, because Trout is an elite hitter, and Jones and Clark are below-average. On the other hand, ponder the same scenario except that Johnny “Always Hits A Slow Infield Roller” Smith is due up, followed by Trout and Trout 2.0. In this case,  Johnny lay down that bunt.

Summary

The conclusions from the scenario analysis are summarized below:

Away team

  • The away team should not attempt a sacrifice bunt to move the runner on second to third because the probability of winning with a runner on third and one out is less than the probability of a runner on second and none out
  • The away team should attempt the stealing of third if the expected success rate is 73.8% or higher (the MLB average for thefts of third was 79.0% in 2019)

Home team

The game is tied going into the bottom of the tenth

  • The home side should attempt a sacrifice bunt to move the runner on second to third base because of the increased chance of winning
  • The home club should try to steal third if the expected success rate is 68.5% or higher

The home team trails by one run

  • The home side should not attempt a sacrifice bunt to move the runner on second to third because the probability of winning with a runner on third and one out is less than the probability of a runner on second and none out
  • The home club should not attempt to steal third. Although the success-rate threshold is 74.5%, the margin-of-error compared to the MLB average (79.0%) is thin. Exceptionally thin given that, if the steal attempt fails, the chance of winning is only 11.1%

The home team trails by more than one run

  • No bunting (reduced probability of winning)
  • No stealing of third (a success rate of 81.8% or better is required)

It is important to remember that these conclusions reflect probabilities that are averages. There may be circumstances where applying these recommendations is not appropriate. However, similar to a Field Manager using batting averages or other batting metrics as a basis to make a pinch-hitting decision, these conclusions are data-based guidelines.

Ben Clemens wrote an article at FanGraphs that examined sacrifice bunts and steals of third in the ROSIE environment. One of his observations was that the success rate of sacrifice bunts relating to a runner on second was approximately 71%. The other outcomes were as follows: the lead runner was out (4.1%); the bunter was out, but there was no advancement (10.7%), and a hit/error occurred (14.3%). For those interested, Table 2 shows that the weighted incremental value of a sacrifice bunt is slightly less than Table 1 indicates. However, the difference is small and does not affect the conclusions above.

Next. Toronto Blue Jays infield makes history. dark

The last word

A new MLB rule stipulates that each half of an extra-inning will begin with a runner on second. As a result, we may see teams’ use sacrifice bunts and steals of third in an attempt to win games in the tenth inning and after. In certain circumstances, these two tactics would be justifiable. I wonder if the Blue Jays will use sacrifice bunts and base stealing in extra-inning games? If the answer is yes, who are the candidates who could successfully execute the strategy?