Hitting by Pitch Count and Total

I was watching the Cubs-Pirates game today (April 4th, 2013), which Chicago won 3-2 with no thanks to Carlos Marmol (there's a post waiting to happen when I go through Brooks Baseball PITCHf/x data), but it did give me my first real exposure to the newest member of the Cubs TV broadcast booth, former Astros pitcher Jiim Deshaies. On more than one occasion, he referenced "hitter's counts," as well as the notion that fly balls don't travel as far in cold weather (maybe yes, maybe no). I'll show the data on how well hitters hit by count, all data from the 2009-2012 regular season:

All straight-forward data, except for the %AB column, which is the percent of at-bats in which the ball was put in play on a given count. Several noteworthy items:

1. There is a war between getting deep in a count vs. swinging at that first pitch, especially if that first pitch is a slider that doesn't slid,e a curve that doesn't curve or a fastball that crosses at 82 mph with no movement. I strongly suspect that the high average of balls hit on the first pitch are precisely that--mistakes that hitters take full advantage of. It's the true gift horse--it's there, so why not take it. In addition, the notion of working deep is valid for starting pitchers only, since I'm unaware of relievers that work on any kind of pitch counts other than the one that says that if you threw 30 pitches in an inning, you weren't very effective.

2. Deshaies is correct with the notion of hitters counts, but the data shows quite clearly that anything less than two strikes works for a hitter. I don't think anyone should be surprised by the dropoff between one and two strikes, but the magnitude is rather dramatic.

3. Look at those numbers on 3-0. It's as old as baseball to take on that pitch, and when only .2% of at-bats occur on this count, it must still be in force, but I wonder if it isn't time to re-think that a bit. If a hitter is guaranteed to get a nice strike, it just might behoove him to swing away on anything that is on this side of borderline. When the OPS jumps over 750 points, well...

The number of pitches the batter sees is less favorable to the hitter than I think is commonly believed:

Ignore that lack of wOBP, since the way I calculate it...on second thought, just ignore it. There is nothing in this chart that suggests that going "deep" into a count (however you wish to define deep) works, especially when after 5 pitches (and occasionally after4, 3 or 2) the hitter has two strikes, and the prior chart showed how well that turned out for hitters. In case the drop in batting average isn't obvious enough, let's graph it:

It's a fine line between working a pitch count to potentially shorten how long a pitcher will pitch vs. hitting at the point most advantageous to the hitter, which is usually earlier rather than later in the count. The numbers appear to come down on the side of hitting sooner rather than later.