Should Barry Bonds Be
Intentionally Walked?  Part 2
by Martell

(Editor's note:  MATH ALERT!  This is a very technical article.)

When we left off, I mentioned that I was going to set up a distribution table to determine whether or not intentionally walking Barry Bonds is the correct strategy.  Actually, I will be writing an equation.  Using Bonds' season-to-date stats, I will determine the odds of Bonds making an out, hitting a single, etc. were he allowed to hit.  I will then do the same for the rest of the Giants' lineup, multiplying the odds of each outcome with one another, recording the results and noting how many runs the team would score in each scenario.  Finally, I will go through this process once more, but this time with Bonds automatically on first base.  In the end, we should be able to compare the two options to see which one allows the Giants to score less runs--this will be the proper strategy for an opponent to use.

First off, some assumptions:  1) I will be using the lineup from the Giants' most recent game (August 3 vs. Cincinnati).  2) Intentionally walking other players on the team will not be allowed.  3) There will be no double plays or sacrifice bunts/sacrifice fly balls.  4) A player will always go from first to third and second to home when there is a single, and will always score from first on a double.

A truly thorough study would take everything into consideration, and perhaps later on I will be able to do all that.  That being said, since all of these details are removed from both equations, the effect should be fairly negligible.

Also, as mentioned last time, I will go through the whole process in detail for one scenario--two outs, nobody on base, Bonds is up.  That way, anyone interested can check my work and make sure I haven't screwed up the arithmetic or left anything important out.

Here are the lineup and stats I will be using (stats are through games of 8/1):

LINEUP
======
Bonds
Feliz
Pierzynski
Grissom
D. Cruz

                   BONDS                FELIZ            PIERZYNSKI           GRISSOM               CRUZ
               ========     ========     ==========      =========     ========
1B s       40  (12.97)      53  (15.10)           64  (19.94)           72  (17.70)        31  (15.90)
2B s       17  (5.52)         19  (5.41)              18  (5.61)              17  (4.18)            19  (9.74)
3B s         0  (0)                  2  (0.57)                1  (0.31)                2  (0.49)              1  (0.51)
HRs       27  (8.77)        14  (3.99)                8  (2.49)              15  (3.69)              3  (1.54)
BBs        67  (21.75)      13  (3.70)             13  (4.05)             20  (4.90)           10  (5.13)
Outs    157  (50.97)  250  (71.23)        217  (67.60)        281  (69.04)      131  (67.18)

The numbers in parentheses are the percentage chance of each occurrence.

Next, I'll calculate the odds of the Giants not scoring.  Here are the ways the inning could end without a run scoring:  1) Bonds gets out, 2) Bonds walks, singles or doubles and Feliz gets out, 3) Bonds walks or singles, Feliz walks or singles and Pierzynski gets out, 4) Bonds doubles, Feliz walks and Pierzynski gets out, 5) Bonds walks or singles, Feliz walks or singles, Pierzynski walks and Grissom gets out, and 6) Bonds doubles, Feliz walks, Pierzynski walks and Grissom gets out.

And here's the math:  (.5097) + (.4026 * .7123) + (.3474 * .1880 * .6760) + (.0552 * .0370 * .6760) + (.4026 * .1880 * .0405 * .6904) + (.0552 * .0370 * .0405 * .6904) = .84418

Here are the ways the inning could end without a run scoring, but now with Bonds intentionally walked:  1) Feliz gets out, 2) Feliz walks or singles and Pierzynski gets out, and 3) Feliz walks or singles, Pierzynski walks and Grissom gets out.

(.7123) + (.1880 * .6760) + (.1880 * .0405 * .6904) = .84465

Well, that's pretty amazing.  It makes almost no difference which decision an opposing coach makes.  There's a slight advantage to walking Bonds in this scenario, about .05% greater chance of getting out of the inning without allowing a run--that is, about 1 out of 2000 times it will make a difference--but it sure isn't much.  Perhaps it will make a difference in the other innings, effecting whether the Giants have a "big inning" or not.  Here are those numbers:

BONDS HITS
=========
No runs --(.84418)
1 run--(.11206)
2 runs--(.02693)
3 runs--(.00533)
4 runs--(.00113)
Inning continues--(.01037)

BONDS WALKS
=========
No runs --(.84465)
1 run--(.06830)
2 runs--(.05615)
3 runs--(.01099)
4 runs--(.00230)
Inning continues--(.01761)

Okay, here's a little something.  If you're a coach that lets Bonds hit, there's a 4.4% greater chance that you will give up exactly 1 run, but also a 3.0% SMALLER chance that you will give up exactly 2 runs.

So for this scenario--once again, two outs, nobody on--if you have a two-run lead, you should let Bonds hit.  If you have a one-run lead, you have a choice.  If you prefer a greater chance of the Giants tying the game, but an equally lower chance of them taking the lead, then you should let Bonds hit.  If not, you should walk him.

In Part 3, I'll show the whole table (thankfully without the math) and identify the other times when Bonds should be walked.  And after that, maybe I'll actually write about something interesting.

 

Martell can be reached at martell@babblog.com.

 

© Copyright Babblog and the Respective Authors, 2004-06
Web Design and Graphics by Jeff Lewis.