It is sensible then to assume that improved plate discipline statistics accompany this ascendant OBP figure. It is perplexing then that Cano both swung more overall in 2010 (52.6% compared to 51.6%), and at more pitches out of the zone (36.5% compared to 30.9%) than in 2009*. This would seem to be counter intuitive; generally, a more patient approach at the plate yields more walks, not a more aggressive one. To gain greater understanding of Cano’s approach at the plate, it is useful to visualize the data (chart to the right).
This graph shows where Cano liked to swing in 2010 (red) and 2009 (blue). All of the area within each of the circles represents the locations where Cano swung over 50% of the time. All of the area outside of each of the circles shows the locations where Cano swung less than 50% of the time. The graph is from the catcher’s perspective, so the right side of the graph is close to Cano and the left side of the graph is away from Cano. The black box is the strikezone.
(click “view full post” to read more)
As you can see, overall his approach at the plate in 2010 strongly resembles his approach from 2009. In 2010, Cano was a little more willing to swing at a pitch on the outer edge of the plate, but other than that there isn’t much of a difference.It seems likely then that his plate discipline did not really improve in 2010; overall, his approach was quite similar to 2009 and he was even a little more aggressive. If the explanation for Cano’s spike in walks cannot come from improved plate discipline, it must come from how pitchers deal with Cano.
In 2010, Cano was intentionally walked (IBB) 14 times, which is equivalent to the sum of all his previous IBB. That was good for second in the American League and first among all Yankee players. If we remove this from the BB% equation and only look at unintentional walks, Cano had rates of 6.2% in 2010, 4.2% in 2009, and a previous career high of 5.1% from 2007. Pitchers also both threw him less strikes overall in 2010 (43% compared to 49.6%) and less fastballs (57.4% compared to 62.8%) than in 2009. Here is a graph that shows the differences in the locations of pitches thrown to Cano* (chart to the right).
In this graph, blue represents locations where pitchers threw the ball less to Cano in 2010 than they did in 2009. Red shows where pitchers threw the ball more to Cano in 2010 than 2009.The graph is from the catcher’s perspective, so the right side of the graph is close to Cano and the left side of the graph is away from Cano. The black box represents the strikezone.
What is so striking about this graph is the huge blue mass located all across the inner part of the strikezone. Relative to 2009, pitchers abandoned the entire inner third of the zone when pitching to Cano; the risk of throwing a meatball and having it hammered was just too great. Also of note is that nearly all of the area outside of the strikezone is light red, meaning that pitchers were much more willing to pitch around Cano in 2010 than in 2009.
We can conclude then that pitcher fear is the answer. While Cano experienced an explosion in his walk rate in 2010, he was not actually better in terms of plate discipline. Pitchers were simply more afraid of him. They attempted to starve Cano of fastballs and pitches for him to turn on, and Cano prevailed. Not only did he sustain his batting average on balls in play skill (the main driver behind batting average), he actually improved his rate of power and his overall production. Who knows what pitchers will try next?

Great job, Josh!
Very nice, putting walk rate into a context that is understandable is great, and this short analytical explanation is well done. I was waiting to find out what the circle represented as I forcasted some cherrypicked number, but over 50% is nonarbitrary and makes a ton of sense. Good work!
This actually got me thinking that being a SABR Scout could be a lot of fun with the kind of data that's available now, almost too much to process, and the number of nonlinear relationships keep increasing. I imagine the pitch f/x data is proprietary?
The pitch f/x system is owned by major league baseball advanced media (MLBAM) and you see the data used mostly in gameday, so in that sense it's proprietary. However, they do make it possible to get the pitch f/x data for free.
This is great stuff. Just one nit to pick. Red and blue are used twice on the same page to differentiate two different variables. It might be more useful to use a different scheme in different graphs if they're not referring to the same distinction.
Thanks for the suggestion. I'll keep that in mind in the future.
Nice piece! I don't comment often, but I'm a constant reader and I'm glad to have this type of analysis in the bullpen as well.
Nice post Josh.
So, can you now go back and show how pitchers feared Jim Rice?
Haha. I would but I'm too afraid to do the analysis. He was that scary.
Great stuff, Josh. You know it really felt like Cano chased fewer pitches (especially that high one) at least in the first 2/3 of 2010. But the data and conclusion are unmistakable. I really look forward to more of these types of analysis.
Looks like Jason is on his way to becoming the Billy Beane of sports blog GMs; all the other sites passed on Josh and Anna but Jason utilized advanced writing statistics and nabbed them. I'll be looking for the forthcoming book: "It's About The Moneyball."
Well we did see him chase less high pitches in 2010 than 2009, but this is mainly because pitchers weren't challenging him up in the zone as much in 2010 as 2009. In other words, he did not have as many opportunities to chase high fastballs, so we saw him chase less high fastballs.