Opportunity for a guest post (college baseball)

by Bud Elliott on May 11, 2009 4:33 PM EDT Comment 16 comments

Greetings from TomahawkNation, the FSU wing of SBNation.

We have a problem. Our coach is in the stone age.

On Saturday, against Georgia Tech, the 'Noles had men on 1st and 2nd with no outs, in the first inning.

Mike McGee, the #3 hitter, is OPS'ing 1.229

http://www.seminoles.com/sports/m-basebl/stats/2008-2009/teamcume.html

I understand bunting here is wrong, that it slightly increases the chance to score 1 and only 1 run, and really decreases the chance to score more than 1 run. What I am looking for is someone to tell our readership just how stupid this move is. How much does it decrease your chance of winning? How much worse is it to do this in the college game than the pro game?

Any help would be really appreciated. Our readership is very stat heavy and they'd enjoy something like this.

P.S. I'm also interested in hosting a baseball v. football: what's worse, punting on 4th and manageableor bunting in some select situations.

0 recs | Comment 16 comments |

Story-email Email Printer Print

Comments

This shouldn't be too hard, but I need a bit. I did it for H.S. teams. Will do it for college.

by Jeff Zimmerman on May 11, 2009 4:39 PM EDT reply actions

Here you go and the following can be used for any baseball situation (college, HS, LL, etc)

1. Goto http://www.tangotiger.net/markov.html and enter in your team data and/or league. Here is a link with discussion on doing Markov for a High School Team
2. The results will be calculate and the table you want is labled: “Run Expectancy Matrix” and “Run Frequency Matrix”
3. Using FSU’s team stats I get:

1st and 2nd, 0 outs: 2.246 runs scored on average
2nd and 3rd, 1 out: 1.723 runs scored on average

His move cost you .5 runs on average.

If you go down to the Run Frequency Matrix, you will find this pretty disheartening:

1st and 2nd, 0 outs: 0.728 chance of scoring
1st and 2nd, 0 outs: 0.705 chance of scoring

So by bunting, the team will score less runs (~.5) and has less chance of scoring (~2%).

There is no reason to bunt in that situation. Here is all the information from the Markov:

8.685 : Runs Scored per Game

AVG / OBP / SLG
0.315 / 0.419 / 0.532

Chance of Scoring
Event 0 outs 1 out 2 outs AVERAGE
1B/BB 0.553 0.376 0.181 0.370
Double 0.669 0.495 0.286 0.484
Triple 0.860 0.690 0.317 0.622

Run Expectancy Matrix
Bases 0 outs 1 out 2 outs
xxx 0.965 0.523 0.196
1xx 1.518 0.899 0.377
x2x 1.634 1.018 0.482
xx3 1.825 1.212 0.513
12x 2.246 1.445 0.683
1×3 2.390 1.604 0.707
x23 2.506 1.723 0.812
123 3.158 2.219 1.102

Run Frequency Matrix
Bases 0 outs 1 out 2 outs
xxx 0.406 0.256 0.114
1xx 0.553 0.376 0.181
x2x 0.669 0.495 0.286
xx3 0.860 0.690 0.317
12x 0.728 0.546 0.307
1×3 0.872 0.705 0.331
x23 0.872 0.705 0.331
123 0.912 0.774 0.419

Marginal Run Values by Event,
Event Markov BaseRuns Runs Created
Walk 0.506 0.403 0.262
Single 0.631 0.581 0.681
Double 0.895 0.886 1.100
Triple 1.147 1.191 1.519
Homerun 1.524 1.457 1.938
Out, sans K -0.512
Strikeout -0.512

by Jeff Zimmerman on May 11, 2009 4:59 PM EDT reply actions 3 recs

Let me know how it goes and here is another point to consider:

Another point looking at the team is that they don’t steal successfully enough considering the run environment.

The team is currently stealing bases at 76% clip 70 for 92

If a person goes from 1st to 2nd with 0 outs, they increase the amount of runs scored by 0.116 (1.634-1.518). If the the person gets caught, the amount of runs scored drops by .995 runs (1.518 – .523). So for the team to at least break even running the bases, the need to have a 905 or better success rate.

For FSU this year, they have gained 8.352 (.116 70) runs stealing bases, but lost 21.89 ((92-70).995) because they have gotten caught stealing.

by Jeff Zimmerman on May 11, 2009 7:57 PM EDT up reply actions