Examining Division I College Baseball's Run Scoring Environment

Recently I helped Tomahawk Nation author FSUncensored determine if Florida State baseball coach Mike Martin should sacrifice bunt in the second inning of a game. In that article, I used FSU's specific run-scoring environment, not the one for Division I baseball as a whole. With the College World Series starting this weekend, I decided to revisit the issue, tackling the Division I run-scoring environment.

First, I went to Tom Tango's website and enter the numbers for the 2009 season into his to Maklov calculator to generate the 2009 college baseball scoring matrix. Then I made another one for the 2009 MLB season so far and compared the two.

As you can see, there is much more offense in the college game compared to the majors. Next we need to look into how this higher run scoring environment should effect a manager's game time decisions.

Should a team sacrifice bunt when not in a close (one run or less) and late (ninth inning or later) game?

The premise of a sacrifice bunt is to be able to put a runner in scoring position and thereby score a run. To look at this idea, we need to start by using the Run Expectancy Matrix generated by Tango's markov:

Run Expectancy Matrix

Let's take the two most common situations and see if it pays off to bunt (all situations assume an average batter).

Runner on 1^st, no outs and runner sacrificed to 2^nd base. The run expectancy before the bunt was 1.19 runs for the inning and it drops to 0.83 runs afterwards. Don't Sacrifice

Runner on 2^nd, no outs and runner sacrificed to 3^rd base. The run expectancy before the bunt was1.34 runs for the inning and drops to 1.06 runs. Don't Sacrifice

There is actually no case in college ball when an average batter should bunt before the 8^th inning or if behind by 2 or more runs in the ninth.

Sacrifice bunting in a close (1 run or less) and late (ninth inning or later) game

For this explanation, I will use the run frequency chart that shows the chances of scoring at least one run, not the one showing total expected runs scored:

Run Frequency Chart

Lets look at the same two situations, but using the run frequency chart

Runnner on 1^st, no outs and sacrificed to 2^nd. The chances of scoring a run before the sacrifice is .49 (or 49%) and it drops to 0.466 about the sacrifice. Don't Sacrifice

Runner on 2^nd, no outs and sacrificed to 3^rd.  The chance of scoring a run is .641 and jumps to .690. Sacrifice

Using the run frequency chart, the only two times you should sacrifice with the goal of scoring one run is with 0 outs and a runner on 2^nd with third base open.

During this season, the average college team sacrificed 32 times. I am pretty sure most teams weren't behind or tied in the ninth inning in that many games.

Base stealing

Base stealing has always been a way to move runners into scoring position, but the success rate of stealing is important, because the cost of a failed stolen base is an out plus a lost runner.  Outs are especially important in higher run scoring environments like college ball, because they happen less often.

To determine the success rate, the Run Expectancy chart will be used again.

Runner on 1^st base, 0 outs and runner tries to steal 2^{nd base}. Runs gained by the steal is 0.15 runs (1.34 runs - 1.19 runs). Runs lost by getting caught: 0.82 runs (1.19 runs - 0.37 runs). Stealing success rate needed to break even = 84% (0.82/(0.82+0.15) *100%)

Runner on 1^st base, 1 outs and runner tries to steal 2nd base. Runs gained by the steal is 0.14 runs (0.83 runs - 0.69 runs). Runs lost by getting caught: 0.56 runs (0.69 runs - 0.13 runs). Stealing success rate needed to break even = 80% (0.56/(0.56+0.14) *100%)

Runner on 1^st base, 2 outs and runner tries to steal 2nd base. Runs gained by the steal is 0.12 runs (0.40 runs - 0.28 runs). Runs lost by getting caught: 0.28 runs (0.28 runs - 0.00 runs). Stealing success rate needed to break even = 70% (0.28/(0.28+0.12) *100%)

For this past season, college teams stole 73% of the bases they attempted. Again, I am pretty sure team weren't stealing with only 2 outs, even though they should. Only 16 out of the 288 teams stole bases at a rate (84%) that generated positive runs for their team or at least broke even.

Closing

There you have it, a quick look at the college run scoring environment along with what this means in certain game situations. I plan on doing this for additional leagues in the future.

For additional reading on this topic:

• The Book by Andrew Dolphin, Mitchel Lichtman, and Tom Tango

• Tim Kniker's Week 2 entry in Baseball Propectus Idol

• High School version created at The Book Blog.