Expanding the Strike Zone
About a year ago, I wrote an article entitled Pitchers, Pitch by Pitch. Using pitch location data from Baseball Info Solutions, I looked at how often individual pitchers get batters to chase pitches outside the strike zone, or in other words, swing at pitches they shouldn't be swinging at. I called this stat Outside Swing Percentage or OSwing, for short. To recap: OSwing correlated with a pitcher's strikeout-to-walk ratio and just for fun, let's look at the starting pitchers and relievers who had the highest OSwing the past two seasons. Name 2005 Name 2006 Brad Radke 31.76% John Smoltz 33.00% Johan Santana 30.21% C.C. Sabathia 31.47% Curt Schilling 29.60% Jeremy Bonderman 31.01% Felix Hernandez 28.43% Roger Clemens 30.88% John Smoltz 27.53% Curt Schilling 30.76% Odalis Perez 27.47% Johan Santana 30.72% Jon Lieber 26.44% Jake Peavy 30.61% Rich Harden 26.00% Roy Halladay 29.55% Andy Pettite 25.97% Odalis Perez 29.16% Paul Wilson 25.33% Aaron Harang 28.91% Name 2005 Name 2006 Brad Lidge 32.41% Cla Meredith 37.88% Rudy Seanez 30.93% Patrick Neshek 35.92% Mike Wuertz 28.67% Mariano Rivera 34.13% Derrick Turnbow 28.40% Mike Wuertz 32.57% Jonathan Papelbon 28.27% J.J. Putz 32.18% Julio Santana 28.01% Brian Sikorski 32.00% Bobby Jenks 27.68% Francisco Rodriguez 31.65% Kyle Snyder 27.06% Dennys Reyes 31.56% Tyler Walker 26.71% Scott Proctor 31.54% Eddie Guardado 26.69% Ramon Ramirez 31.31% All in all, some fairly prestigious lists, especially the 2006 list of starting pitchers, but OSwing is hardly a "magic bullet stat" since there are some players on these lists I'm sure you're scratching your head at. Let's take a step back for a moment and think about the relationship between batters and pitchers. Ideally, a pitcher is going to try and get ahead in the count and when this happens the pitcher has effectively "expanded the strike zone" since the batter is now on the defensive and will be more prone to chase pitches outside the strike zone. Conversely, when a pitcher is behind in the count, a batter will be less prone to chasing bad pitches. Looking at OSwing by count this becomes fairly evident. Count OSwing ZRatio 0-0 11.33% 1.15 0-1 22.54% 0.83 0-2 31.57% 0.51 1-0 18.61% 1.31 1-1 26.78% 1.05 1-2 37.37% 0.70 2-0 16.38% 1.61 2-1 28.58% 1.41 2-2 41.40% 0.98 3-0 2.69% 1.73 3-1 23.37% 1.67 3-2 44.86% 1.57 Furthermore, if you look at the ratio of balls thrown in the strike zone to those outside the strike zone (ZRatio), you can see that when a pitcher is ahead in the count, he's much less likely to pitch in the strike zone. So, is how often a pitcher gets a batter to chase a ball outside the strike zone merely a matter of his skill at managing the count? Not exactly. While the two are hardly mutually exclusive, some pitchers do seem to be able to make batters chase pitches outside the strike zone more than others in the various counts. Here are the 15 starting pitchers who get ahead in the count the most and their OSwing above (or below) the MLB average when ahead in the count. Starter OSwing(Above Average) Ahead% Johan Santana 22.89% 33.91% Curt Schilling 19.34% 33.74% John Smoltz 27.29% 33.55% Mike Mussina -4.35% 33.27% Roy Oswalt 6.43% 33.21% Elizardo Ramirez -1.56% 32.84% C.C. Sabathia 37.21% 32.81% Brad Radke 4.99% 32.35% Paul Byrd 3.93% 31.97% Jon Lieber 1.01% 31.68% Francisco Liriano 20.35% 31.37% Randy Johnson 19.13% 30.73% Pedro Martinez 3.40% 30.72% John Lackey 17.62% 30.64% Cliff Lee 1.93% 30.55% As you can see, many of the top pitchers in baseball get ahead in the count, but they don't always have the same success getting batters to swing at their pitches outside the strike zone. Let's have a look at the relievers: Starter OSwing(Above Average) Ahead% Patrick Neshek 30.22% 39.82% Rafael Betancourt -1.81% 39.64% Chad Bradford -3.44% 36.69% Jose Valverde -12.64% 36.26% Mike Timlin 9.34% 35.78% Trever Miller 14.74% 34.99% J.J. Putz 18.50% 34.93% B.J. Ryan 10.92% 34.92% Todd Jones 8.87% 34.75% Justin Duchscherer 5.01% 34.62% Rafael Soriano 25.02% 34.49% Mariano Rivera 30.22% 33.73% Matt Capps 4.88% 33.60% Jonathan Papelbon 22.20% 32.78% Cla Meredith 71.75% 32.78% Once again, pretty much the same deal with the relievers. It's worth noting that Cla Meredith's 71.75% above average OSwing was by far the highest in baseball when ahead in the count. Batters swung at just over half the pitches he threw out of the strike zone. Just looking at these two lists, I'd venture to say that the pitchers who more often than not get ahead in the count, and those who were also able to get batters to chase pitches more than the MLB average would probably be the most dominant pitchers in baseball. Let's look at one final list where the pitchers are at least 15% above the MLB average in OSwing and percent ahead in the count. Name OSwing(Above AVG) Ahead%(Above AVG) ERA K9 BB9 Patrick Neshek 30.22% 46.24% 2.18 12.89 1.45 J.J. Putz 18.50% 28.26% 2.29 11.94 1.49 Rafael Soriano 25.02% 26.65% 2.25 9.75 3.15 Johan Santana 22.89% 24.54% 2.77 9.43 1.81 Curt Schilling 19.34% 23.92% 3.97 8.07 1.23 Mariano Rivera 30.22% 23.88% 1.80 6.60 1.32 John Smoltz 27.29% 23.22% 3.49 8.18 2.13 C.C. Sabathia 37.21% 20.48% 3.22 8.03 2.05 Jonathan Papelbon 22.20% 20.38% 0.92 9.87 1.71 Cla Meredith 71.75% 20.37% 1.06 6.57 1.06 Matt Thornton 15.10% 18.31% 3.33 8.16 3.50 Scott Proctor 25.63% 17.25% 3.51 7.82 2.90 Bill Bray 29.43% 17.08% 4.08 6.92 3.19 Brian Sikorski 27.47% 16.14% 5.02 10.05 1.85 Alan Embree 36.32% 15.81% 3.26 9.11 2.57 Joe Nathan 20.58% 15.48% 1.58 12.51 2.10 Francisco Liriano 20.35% 15.20% 2.15 10.71 2.38 Brandon League 23.21% 15.09% 2.53 6.11 1.89 A pitcher's ability to expand the strike zone is something that is widely talked about, but rarely quantified. What I find most fascinating about all this is that by using location data, it should be possible to see exactly how far and where the strike zone expands/contracts for individual pitchers and batters depending on the count. It sounds like the type of information that could really make a difference in statistical scouting reports. But that's another project for another day. David Appelman is the creator of FanGraphs.com. You can contact him via e-mail. |
Comments
Very cool.
Quick question: only two of the starters from the '05 top 10 made it to the '06 top 10. Do we know how well this correlates year-to-year?
Posted by: Nolan at December 7, 2006 4:49 AM
Just looking at 2005 & 2006, it has an r^2 of about .27 with pitchers who threw at least 500 pitches in both seasons.
Posted by: dkappelman at December 7, 2006 7:37 AM
I was going to make the exact same comment as Nolan, so I'll ask the follow up instead.
Quick question: are we disappointed by the lower r^2 value?
ps. Very Cool!
Posted by: Jacob at December 7, 2006 7:55 AM
I don't have multiple year r^2's for walk rate and strikeout rate on hand... but just looking at 2005-2006 with the same group of pitchers:
K/9 - .5
Unintentional BB/9 - .3
ERA - .09
It's probably worth mentioning that I found the average 2006 OSwing about 3% higher than 2005. I don't know if this is due to improvements in pitch location charting or just random year to year variance. I'll be sure to take another look at this next year. These pitch location studies are hardly "set in stone" and very much an ongoing project.
Posted by: dkappelman at December 7, 2006 8:27 AM
Brad Lidge led relievers in 05. Obviously he had his problems in 06. What was Lidge's percentage in 06? I'm just curious to see how far it dropped between 05 and 06. A key to Lidge's strike outs is the swing and miss at his slider in the dirt. People frequently stated that batters quit swinging at Lidge's slider in 06, thereby increasing his walks and causing him to have to throw fastballs in bad counts.
Posted by: CJ at December 7, 2006 6:29 PM
Lidge saw a slight decline from 2005 - 2006 in actual OSwing, but it was only 2%. But, considering the baselines are different from 05 and 06, if you do % over league average, he went from about 59% better than the average pitcher to 26% better.
It'd be interesting to see if he was throwing his fastball more and in what counts. His strikeouts are still pretty off the charts and his walk rate isn't that bad. It doesn't look like OSwing really sheds any new light on his troubles.
Posted by: dkappelman at December 8, 2006 8:21 AM