As promised in the update from yesterday’s post here are the results from the test of hitting power on walks for 2001. The format is close to the same as before, with only a few modifications. First, the coefficients I report result from regressions without the control variables. I did this because I want show you the strongest support I can find for power preceding walks (recall that I have been biased in the opposite direction). I included the controls last time, because they led to the strongest results. It is important to note that the controls change very little about the result so I think they are likely unimportant. The second change is that I was able to pull intentional walks (IBBs) out of the analysis with this data. I could not do that for the 2002 analysis. As you will see, the results do change some when looking only at non-intentional walks.
Here are the coefficient estimates when IBBs are included in the walk-rate.
Var. Coef. T-stat Elasticity SLG 0.413 1.92 1.28 Iso-P 0.573 1.7 1.14 HR 0.052 0.64 0.10
SLG is not-quite statistically significant at the 5% level. Iso-Power is significant at only the 10% level. The marginal impact of power on walks is much larger in 2001 than in 2002 for SLG and Iso-Power. 2001 is the year that Bonds really, I mean REALLY, stepped it up. I remember watching that Braves series in May when he hit 6 HRs in 3 games. I am sure many pitchers took note of such power jumps (although HRs alone don’t seem to explain pitcher behavior). By 2002, Bonds’s reputation was at a whole new level, which is why the effect may not be so pronounced. In 2002 Bonds, pitchers percieved Bonds to be dangerous at all times.
But, how much did pitchers respond by simply giving Bonds a free pass, and how much of it was pitching around Bonds? That is, when pitchers pitched to Bonds, did they nibble or go after him based on his recnt history of hitting power. In this table I look at the walk-rate excluding all intentional walks.
Var. Coef. T-stat Elasticity SLG 0.300 1.48 1.10 Iso-P 0.420 1.32 0.99 HR 0.046 0.6 0.10
All of the coefficients shrink some, and the results are no longer statistically significant. Thus, it seems that while there may have been some pitching around, pitchers were responding to Bonds’s power surges with IBBs. Here is a link to the 2001 game-log data. Thanks again to Doug and Alec for pointing me to it.