Allen Webster’s Zone% and Strikeout Rate

Allen Webster was on my “to investigate” list because his strikeout rate was low compared to the number of swings-and-misses he got in 2014. The numbers intrigued me enough to watch one of his starts for one of my Quick Looks (full report on Friday).

What I found is he can’t/doesn’t throw pitches in the strike zone. Hitters need to chase his pitches out of the strike zone if they want to make contact. Most of the time they don’t though.

In an effort to figure him out further, I found out how much to adjust a pitcher’s predicted strikeout rate knowing his strike zone percentage and swinging strike rate, but it doesn’t make a huge difference.

Time for a little math. A simple way to estimate a pitcher’s strikeout rate (K%) is to double their (SwStr%). Taking all of the 2014 pitchers (min 50 IP), the r-squared between K% and two times the SwStr% (pK%) is 0.66. Good, but not great. Usually early in the season, I look for pitchers with high differences between these two values to find potential break out or bust candidates.

Now we get to Allen Webster. Here are the ten pitchers with the biggest negative differences between their K% and their predictive strikeout rate (pK%)

Name	Team	IP	K%	SwStr%	pK%	Zone%
Zach Britton	Orioles	76	21.8%	13.1%	26.2%	48.8%
Koji Uehara	Red Sox	64	32.1%	18.8%	37.6%	51.1%
Zach Putnam	White Sox	54	21.6%	13.7%	27.4%	36.2%
Luke Gregerson	Athletics	72	20.8%	13.3%	26.6%	45.9%
David Hale	Braves	87	11.5%	8.7%	17.4%	43.2%
Jared Burton	Twins	64	16.9%	11.4%	22.8%	51.7%
Phil Coke	Tigers	58	16.0%	11.0%	22.0%	43.4%
Jared Hughes	Pirates	64	14.1%	10.2%	20.4%	42.0%
Allen Webster	Red Sox	59	13.9%	11.0%	22.0%	44.9%
Bryan Morris	– – –	64	18.4%	14.0%	28.0%	44.0%

Two items stick out on this list. I used 50 IP to included relievers, but found no pitcher with over 100 IP way at the bottom. Here are the three pitchers with over 100 IP with the largest negative difference.

Kyle Gibson (179 IP): -3.5%
Jeff Locke (131 IP): -2.6%
Edwin Jackson (140 IP): -2.2%

The extreme differences begin to lessen as the number of innings pitched increases. Some regression towards a difference of zero should be expected (possible article for a later date).

Today, I focus on the second item of interest … a low Zone% for these pitchers. The league median Zone% value is around 49%. What I wanted to find is if a low Zone% could cause a pitcher to have a K% lower than the value predicted by their SwStr%.

After several different looks into seeing if Zone% has an effect and the extent of the effect, I kept finding the same answer. There is an effect, but it is generally pretty small. Here is the basic rule of thumb I came up with:

Using 49% Zone% as the break even point: Every 5 percentage point increase in Zone% there is on average a 1 percentage point increase in pK%. It works also with a drop in pK% with a Zone% under 49%.

Adjusted pK% = ((Zone% – 49%) * 0.2) + pK%

For an example, here is how the formula would work with Allen Webster:

Zone% = 45%
SwStr% = 11%
pK% (SwStr*2) = 22%
K% = 14%

Adjusted K% = (((44% – 49%) * 0.2)+22%) = 21.2%

Allen Webster’s lack of Zone% would likely force his pK% down ~1% point, but not enough to explain the huge difference between pK% and K%.

The inability to throw pitches in the zone can cause a pitcher to have a lower strikeout rate than normal given a known amount of swing-and-miss. The amount of change is just not much. Allen Webster’s large discrepancy between K% and pK% is probably a small sample size issue than an issue with him throwing strikes.