Pounding the Zone: Walk Rate Peripherals

When we look at a hitter that’s struggling to produce, we have plenty of peripherals at our disposal. When we look at a pitcher that’s struggling with his control, we have… two? We have his zone percentage, and we have his first-strike percentage. We can compare those to the league average and hope we have a sense of how important either is to his walk rate going forward.

Well, let’s see how well these things correlate to walk rate. Why not.

Seemingly, zone percentage is the most important number. You want to have a better walk rate? Throw the ball in the strike zone. Of course the two are related, but you might be surprised about the slope of the line describing their relationship:

Yeah that’s not a heck of a slope. The r-squared value for this relationship is .0875, meaning that zone percentage describes just short of 9% of the variance in walk rate. There’s a general relationship between the two, but this means that there are plenty of wild guys like Danys Baez in 2002, who hit the zone 56% of the time while walking 11.3%, and plenty of good control guys like Mark Buehrle last season, who hit the zone 43.1% of the time while walking 5.2% of the batters he faced. Zone percentage is just a general guide.

Why is this? It might be because pitchers with exceptional stuff can get batters to chase on pitches — that’s not a pitch in the strike zone, but it’s a strike, not a ball. Maybe we can find our low-zone-percentage, low-walk-rate hurlers and look at their o-swing percentage? Taking pitchers with a below-average zone percentage (we’ll use 49%) and an above-average walk percentage (we’ll use 8.5%), we get a group of 91 pitchers that averaged swings on 29.57% of their pitches outside the zone compared to the league average o-swing of 28.06%. But if you limit the list to only those with a better-than-average contact percentage, the list gets (smaller and) more interesting:

Now that we’ve narrowed our list to those pitchers that pitch in the zone less than average but still have better-than-average walk rates — and also garner more whiffs than average — we get a stronger result. These pitchers get whiffs on swings outside the zone, and that helps mute the effect of zone percentage on walk rate.

We still have first-strike rate to fall back on. It makes sense that fewer at-bats that start 0-1 will end up in a walk (4.6% vs the league average of 8.2%), but it also seems that if it was that easy, everyone would just throw strikes on their first pitch.

Here, the r-squared is .435, meaning that first-strike percentage is much more important to walk rate. It explains almost half of the variance in walk rate! That’s pretty impressive. After all those years hearing about the importance of getting strike one and pounding the zone — now we know which is more important.

As a bonus, the size of the circles in this last graph was determined by the pitcher’s zone percentage. As you can see, the bigger circles (better zone percentage) are clustered on the lower part of the graph. Perhaps, with the two together (and o-swing% and contact%?), we can find an equation for expected walk percentage. I’ll just have to go math up a little before we can get there.