Kawasaki, Arencibia and the Two Strike Approach

One of the great goals of the sabermetric revolution in baseball is to scientifically evaluate assumed truths within the sport. Baseball is a game absolutely steeped in tradition and because the sport looks so fondly upon its past it can become set in its ways. Conventional wisdom often rules the day from the press box to dugout to the general manager’s office. The leadoff man has to be fast. The cleanup hitter has to hit for 100 RBI’s. Having a proven closer is essential. Et cetera. Those who are fluent in advanced statistics often take great pleasure in poking holes in old school logic that can be disproved with objective quantitative information. There is a sense of discovery to uncovering truth hidden behind years and years of misleading platitudes and it’s always nice to feel smarter than people getting paid a great deal of money to understand baseball.

This kind of myth-busting is satisfying for the writer, interesting for the reader and provocative for the baseball coomunity so it is understandable that there is so much of it in circulation. However, it feels like less effort is put into trying to prove the older wisdom to be true. I’m not one for deferring to the authority of blanket statements without examining them further, but usually behind each cliché is at least a kernel of truth. Today I thought I would examine one of said kernels of truth. Specifically, I thought I would analyze the oft discussed but rarely quantified concept of the "two strike approach".

The phrase "two strike approach" is likely painfully familiar to Blue Jays fans in that they have often been reminded that Anthony Gose does not have one. Or if he does have one, it is not very advanced. The idea is that when a hitter reaches a two strike count he must significantly alter the way in which he goes about his business. The two strike approach has two elements that make it differ from the approach in other counts: The expansion of the zone and the fighting off of pitches.

When you are watching a baseball game when a batter reaches a two strike count you can bet your bottom dollar on the play-by-play guy making a comment about that batter having to "expand the zone". The idea of expanding the zone is simple and fairly intuitive. Due to the fact that taking a pitch called for a strike will result in an out, the hitter no longer has the luxury of taking borderline pitches that may or may not be strikes. Instead he must swing away at any pitch that he thinks the umpire could conceivably call as a strike, in order to avoid the strikeout. This is not a novel or difficult concept. I have no illusions of believing that I am explaining this to anyone for the first time and I don’t think it is a concept widely thought to be false by baseball fans anywhere.

When it comes to "fighting off pitches" the two Blue Jays that come to mind are Reed Johnson and Munenori Kawasaki (more on him in a minute). The fact of the matter is that with two strikes contact is an absolute must. Those who can raise their contact rate with two strikes demonstrate the ability to "fight off pitches" This usually manifests itself as fouling balls off, but it can also take the form of poking the ball up the middle or the other way.

While both these concepts seem inherently true, it has never been made clear to me exactly how true. Given that the information is out there to answer that question in a fairly precise manner (thanks Brooks Baseball) I figured I would give that a shot.

The first target was zone expansion. In order to quantify the expansion the strike zone with two strikes I had to come up with a new statistic which I creatively dubbed "Z-Expansion%. Z-Expansion% takes the percentage of pitches swung for outside the strike zone on two strike counts and compares it to the percentage of pitches swung for outside the zone on all other counts. For example, if a player were to swing for pitches outside the zone 30% of the time in 0 and 1 strike counts but 45% with two strikes his Z-Expansion % would be 50% because he swings for pitches outside the zone 50% more often with two strikes. In theory, the Z-Expansion% could be negative if a player were to swing for fewer pitches outside the zone with two strikes, but that did not come up in my research and I can’t think of any reason why it would.

The concept of fighting off pitches was a bit tricky to figure out how to quantify. After considering the problem for a while I could not come up with a metric that entirely satisfied me. Nothing I came up with was worthy of use, let alone a fancy name. Ultimately I decided to keep it simple. In order to simulate the ability to fight off pitches and stay alive in at-bats I looked at whether batters had a decline in Whiffs per Swing with two strikes that would indicate a purposeful and effective change in approach.

Now that I had an idea of how I might quantify that two strike approach I wanted to apply it to the Blue Jays. This is a Blue Jays blog after all. I planned to examine the data and determine who had the most and least advanced two strike approaches on this team in 2013 (minimum of 250 PA, which means that I actually didn't even research the aforementioned Anthony Gose). My first instinct was to assume that a low Z-Expansion% and a big decline in Whiffs per Swing with 2 strikes would denote the best approach. I thought that hitters that maintained a consistent strike zone were the ones most likely to avoid strike outs and be effective with two strikes. However, as I dug into the numbers I saw the error of my thinking. In fact, the lowest Z-Expansion% went to players who already swung for far too many pitches outside the strike zone earlier in the count, making the difference with two strikes far less significant. I ultimately decided that a high Z-Expansion% was indicative of an advanced 2 strike approach because it indicated a purposeful change to protect the plate after being disciplined earlier in the at bat. With the criteria in mind it was not difficult to spot the most and least advanced two strike approaches on this squad in 2013 (it was also not difficult to predict them, which is why I don't feel bad about the spoilers in the headline). The choices will not surprise anyone, but the numbers just might. Without further ado the award for most advanced two strike approach goes to none other than the greatest .229/.326/.308 of the modern era:

Munenori "Two Strike Ninja" Kawasaki

Part 1: Zone Expansion

They say a picture is worth a thousand words. If that’s the case than it’s definitely worth a few thousand charts full of numbers, so we’ll start with a couple of beautiful Brooks Baseball offerings before we get to the hard data. Below is Munenori Kawasaki’s Swing % on all counts:

Notice how well he stays within the strike zone and keep in mind that this actually includes two strike counts; it would be much more intense of an effect if you just looked at specifically non two strike counts. Now we take a look at how he swings with 2 strikes:

You can see a pretty profound difference there, especially on pitches off the plate outside. The overall numbers look like this:

Count	O-Zone Pitches	O-Zone Swings	O-Zone Swinging %
All Counts	739	162	21.9%
0 or 1 Strike Counts	497	63	12.7%
2 Strike Counts	242	99	40.9%

Z-Expansion %: 222%

You don’t need a frame of reference to know that those numbers are insane. Kawasaki almost never swings for pitches outside the zone unless there are two strikes. When he reaches two strikes he is more than three times as likely to do so. The second highest Z-Expansion% on the Jays was 127% (belonging to Jose Reyes) which gives you an idea of how special Kawasaki is. It was well known that Kawasaki had excellent plate discipline and was a tough out with two strikes but the extent to which that’s true is pretty staggering.

Part Two: Fighting Off Pitches

Unfortunately, the effect for this aspect of the two strike approach was not nearly as significant as the first, but it is still fairly interesting. Below is Kawasaki's Whiffs per Swing in all counts:

While that is hardly a problematic picture, it is apparent that Muni does an even better job of making contact when his back is against the wall:

Count	Swings	Whiffs	Whiffs per Swing
All Counts	437	67	15.3%
0 or 1 Strike Counts	208	38	18.2%
2 Strike Counts	229	29	12.6%

While this effect isn't nearly as dramatic as the first one, there appears to be something here. Anyone who has seen Kawasaki foul off pitch after pitch refusing to strikeout might predict an even bigger discrepancy here but the problem is that the Japanese shortstop is always very difficult to whiff so there isn't a ton better he can do with two strikes.

Moving on we reach the man who surprises no one by winning the prize for least advanced two strike approach:

J.P. "Every Count is the Same" Arencibia

Part One: Zone Expansion

The diagrams below shows Arencibia’s swing rate in all counts:

This diagram shows his swing rate in two strike counts:

As you can see there really isn't that big a difference between these two pictures. The numbers bear out that the way Arencibia wins the distinction of least advanced two strike approach is by declining to have a two strike approach.

Count	O-Zone Pitches	O-Zone Swings	O-Zone Swinging %
All Counts	1195	511	42.8%
0 or 1 Strike Counts	779	305	39.1%
2 Strike Counts	416	206	49.1%

Z-Expansion %: 25.6%

While it’s unfair to say Arencibia doesn’t change his approach at all when he nears the inevitable strikeout, the difference isn’t nearly as much as one would think. To give a sense of perspective, even noted hack artist Rajai Davis had a Z-Expansion of 63.4%. The reality is that Arencibia’s strike zone judgment is so poor in every count that he couldn’t swing much more by the time it reaches two strikes. Kind of sad really.

Part Two: Fighting Off Pitches

If you were a big fan of the very similar J.P. Arencibia pictures above you are going to love the next two. We begin with Arencibia's Whiffs per Swing on all counts:

And now with two strikes:

And for good measure here are the numbers:

Count	Swings	Whiffs	Whiffs per Swing
All Counts	1025	330	32.2%
0 or 1 Strike Counts	644	212	32.9%
2 Strike Counts	381	118	31%

The theory that J.P. Arencibia does not have a two strike approach is alive and kicking with these numbers. I'm not entirely sure what else you can say about it, he just doesn't demonstrate any ability to cut down on his whiffs in two strike situations, which helps account for his mammoth strikeout totals.

Taking a stab at quantifying the effect of two strike approach gave rise to some fairly predictable conclusions. Munenori Kawasaki is a wizard-ninja and J.P. Arencibia is not. While I don’t think anyone is going to be seeing Z-Expansion% on the back of a baseball card any time soon, it hopefully serves as an example of how statistics can serve to help increase our understanding of what we already know to be true as opposed to taking aim at the baseball establishment. They’ve taken enough abuse already.