Friday, August 15, 2008

Is A-Rod Easy To Pitch To (Part II)

In the first part, we began to investigate if there is a predictable pattern behind A-Rod's performance, and whether pitchers can maximize their success by picking the right pitch and location for each count. The main assumption is that if there were an easy way to deal with A-Rod, that lots of pitchers would be aware of this-- it's when they deviate from the game plan that the pitch ends up in the bleachers. Our friend is Josh Kalk's pitch f/x database, and our enemy is sample size.

This time, we'll see how A-Rod performs in the most pressure counts- with two strikes and with three balls. Then we'll see if there's anything interesting in the data related to pitch speed and break.

0-2 Count

Looking at the spreadsheet (available for free here), A-Rod swings at about an average rate in this count, though a higher than average number of those swings go for fouls. One thing you're not likely to see is A-Rod staring at a called strike three on this particular count- just four called strikes in 119 0-2 pitches in the database. The most interesting story in this count is on sliders. A-Rod is twice as likely to swing at an 0-2 slider, and has more swinging strikes on sliders than all other pitch-types combined. The location of the pitch does not seem to matter, although just like other counts he is more likely to swing at the slider away (from righties) and the slider down and in (from lefties).

1-2 Count
Before I saw the data, I had no reason to expect that A-Rod would perform any differently between 0-2 and 1-2 counts; the data disagree with that preconception. First of all, A-Rod is much more likely to swing on 1-2 (54.4% versus 44.5% on 0-2 and 45.7% overall). In the 0-2 count, we saw that the slider was deadly against A-Rod while he had some success with a fastball. The reverse is true on 1-2, he's even hit two home runs on 1-2 sliders. The visual evidence on the 1-2 fastballs is is striking:

Wheras A-Rod only swings and misses on pitches out of the zone on 0-2, there are a number of missed fastballs right in the middle of the plate. I cannot think of any reason for this to be true; it is not true of any other count. It seems at times that he cannot handle the 1-2 fastball. However, A-Rod's remarkable eye continues to show, as he is very unlikely to take a third strike on this count as well.
2-2 Count
A-Rod is even more likely to swing when the count is evened up, at 69.3% the most of any count. This is especially true of fastballs, where the rate goes up to 71%. By far the most common result of these swings is a foul ball... 34 times in 76 2-2 fastballs in the database, well above his normal fastball foul-percentage of 20.3%. Pitchers seem to like to pound the zone on this pitch, not wanting to run the count full:

A-Rod seems more likely to go outside the zone on this pitch than in other counts, but he's still unlikely to take that third strike. The best strategy seems to be that breaking pitch outside the zone or the fastball up.

3-2 Count

With the count full, A-Rod has had a lot of success. His keen batting eye returns from previous two-strike counts, although he will still go out of the zone. He also gets a lot of hits on pitches right over the plate, and with a .571 TB/Swing, it's his best count. If there's a pattern to how to get A-Rod out in these counts, I can't pinpoint it. Occasionally he will swing through fastballs in the zone, but that's not the best plan given his multiple homers on full counts. Once again there are a high number of foul balls:

In general, with two strikes A-Rod does what we would expect most hitters to do: he swings more, he takes strikes less, and he fouls off more pitches. According to a visual look of pitch location, he is more likely to chase pitches out of the zone. Perhaps its his baserunning ability that causes pitchers to pound the zone on 3-2.

Other three-ball counts are not nearly as successful for A-Rod. Even 3-1, a notable hitters pitch, doesn't match the full count in terms of TB/Swing. He is more likely to take a pitch in this count (22%), but not as likely as the 1-0 count (23.1%). Just like the 2-0 count I examined last time, A-Rod fouls off a lot of 3-1 pitches (26.8%). On 3-0, A-Rod doesn't seem to get the take sign all that often, his take percentage is 21.2%, which is near the median. While he's only missed a 3-0 pitch once, he hasn't had much success, with a TB/Swing of just .209. So we continue to have evidence that A-Rod fouls off a high percentage of "hitters pitches." His best pitches, as measured by TB/Swing, are three-ball counts.

Pitch Speed and Break
Location, pitch type, and count are not the only factors that go into A-Rod's performance, the effectiveness of the pitches are also important. Pitch f/x also allows a peek into this world. I've added sheets to the Google Spreadsheet that detail what goes on when the pitcher dials up the speed. First of all, he's more likely to swing:

All Fastballs-- 46.2%
Greater than 90 mph-- 47%
Greater than 93 mph-- 47.8%
Greater than 96 mph-- 51%

His effectiveness goes down as the speed increases. Amazingly, in 104 pitches above 96 mph, he has swung 52 times. He has yet to get a hit. Twenty-seven of those fireballs were fouled off, while he missed swinging at ten of them (chart at right).

Sliders, meanwhile, gain effectiveness not from speed but from break. The break is measured as deviation from a line extending from the front-middle of the plate to a point 40 feet in front of home plate. In the horizontal direction, a positive pfx score is indicative of a pitch which moved from the middle of the plate to the outside (from a righty to A-Rod). In the vertical direction, a positive value is a pitch which drops less than it does from gravity alone (due to backspin); a negative value drops more. The average slider to A-Rod broke 1.97 inches horizontally (away from A-Rod) and 3.03 inches vertically. Compare this to the fastball, which had a horizontal break of -3.44 inches (towards A-Rod) and a vertical break of 10.44 inches. Sliders don't so much slide as they do stay put, relative to the fastball. So is the more effective slider the one which stays put, or the one which dances?

According to the pitch f/x data, A-Rod is most dangerous on a slider which breaks more than three inches horizontally and between 0 and 3 inches vertically.

So what have we learned in this voyage that will help a pitcher? Some of the things the data suggest would be obvious by common sense, others are a bit puzzling. To sum up for readers like Shaun P who like to read the conclusions first:

1) Pitch A-Rod up and in early in the count, don't try to get him to chase out of the zone.
2) Behind in the count, A-Rod is less likely to take a strike and more likely to foul off a pitch.
3) In hitters' counts, A-Rod tends to foul off a high number of pitches.
4) A-Rod's most dangerous counts are 1-0, 3-1, and 3-2.
5) On 0-2, A-Rod will swing at sliders but not fastballs; on 1-2 the reverse is true.
6) A-Rod has trouble with pitches exceeding 96 mph.
7) Sliders which deviate too much from their initial trajectory will be hit a long way.

What do people think of these conclusions? Are there ready explanations for the more puzzling ones? Is there something wrong with my analysis? I'm open to all comments. Thanks for reading!

No comments: