Nick Markakis’ 2009 Drop In Walks (Part Two)
March 2nd, 2010 | by Daniel Moroz |In part one of these series of posts, I took at look at what factors affect player’s walk rate. The most important component seemed to be the rate at which batter swung at pitches outside the strike-zone, so I suppose that’s a good place to start investigating.
Let’s take a look at the pitches that Nick swung at in 2008 and 2009, first by their height. 2008 is in red; 2009 is in blue, the black lines give an idea of the top and bottom of the strike-zone, and the numbers at the bottom are the percentage of pitches at which he swung.
On lower pitches it’s pretty close, but on the high ones Nick swing a good deal more in 2009 – at the expense of those pitches in the middle zone that one would think would be better to hit. Pitchers did elevate the ball a bit more against Markakis in ‘09, but not nearly to this degree – and they didn’t exactly force him to swing, either. Doesn’t seem like a positive development. It might explain Nick’s increase in flyball rate though, with almost 41% of his balls in play being flyballs compared to 33% in 2008. If he’s putting the bat on the ball higher in the zone, then it would make sense that he’s hitting it into the air more.
The horizontal look:
He chased pitches inside off the plate more (remember; this is from the catcher’s perspective, so Nick would stand on the left there), as well as those outside (though not to the same degree). Again, while swinging at pitches in the middle of the plate less.
I was going to do further break-downs using graphs, but figured that there were so many splits that it wasn’t worth it. Instead, a table!
The rates don’t exactly match FanGraphs since I simplified things a bit, but it’s close. Lots of stuff to look at there. Takeaways?
- Swung at pitches in the strike-zone less and pitches out of the strike-zone more. That was consistent both against right-handed and left-handed pitchers. It’s kind of bizarre actually; did his pitch recognition get messed up or something?
- Swung at strikes less with no one on bases, but balls more.
- Chased pitches out of the zone more with runners on.
- I thought that maybe lefties were getting Nick to chase breaking-balls more, but that didn’t really happen. There was the increase on change-ups, but that was on a very small number of pitches.
- Swung less at fastballs in the zone but more at those out of the zone, which probably explains the height graph above. He wasn’t able to lay off of those pitches as well as he did in 2008. Perhaps he was feeling some pressure as the team’s main run producer and was looking to drive those high pitches.
In part three I’ll look into the fastball issue in a little more detail.
Tags: Nick Markakis, Pitch/FX
By Crawdaddy on Mar 2, 2010
Just wondering . . . are you adjusting the data for each stadium? Pitch f/x numbers differ wildly in some places (upwards of 3 inches) when I was looking at data last year.
By Daniel Moroz on Mar 2, 2010
I did not, and that’s a fair point. Over the course of the season – and especially comparing two seasons played in the same parks – shouldn’t make a very big difference though, I’d think. Do you know of a good way to make such an adjustment?
By strachtman on Mar 3, 2010
This is great, great stuff, Daniel.
By Crawdaddy on Mar 3, 2010
You have to go in and look specifically at the data. Very large differences are noted between seasons in the same park because the equipment as I understand it is taken down or replaced often and replaced with newer units.
What others have done is that you can pick starters and do a simple comparison of pitch locations park to park. Supossedly it becomes very evident after half a season. I figured out a major difference between Baltimore and . . . maybe KC last year. I manually corrected it.
By Daniel Moroz on Mar 3, 2010
Hmmm… I definitely do remember seeing differences looking at them on a game by game basis. I’ll try to keep this in mind going forward, though if it’s a matter of manually making these adjustments every time, then that makes me a sad panda.
By Crawdaddy on Mar 3, 2010
Tell me about it.
What might work best is to figure out the coefficient for each metric and assign it to the home team. That would do it . . . but yeah it is a bit of work if more than a couple variables are off.