In part one of these posts, I took a look at what effects walk rate. Part two was Nick’s general swing rates and splits. Part three was some more detail of swing rates on the fastball in particular. That’s a piece of the puzzle, but it’s the end result – the 45% drop in walk rate, from working a BB in 13.3%* of his PA to just 7.9% – that was the actual hit to Markakis’ value to the team. How did that happen?
* He actually walked in 14.2% of his plate appearances, but 7 of those were intentional walks and I took those out.
First off, to draw a walk you need four balls. And to get to four balls, you need to get to three balls*. I think that implies** pretty strongly that a batter who can get to a 3-something count more often is going to be able to work a walk at an increased rate.
* Advanced mathematics!
** Didn’t feel like checking. If that’s wrong, feel free to let me know.
Here is the proportion of total plate appearances in which Markakis reached a three-ball count as well as the AL averages (excluding intentional walks):
and compare that to his walk rates versus the league average (again excluding intentional walks):
You can see that both the plate appearances getting to three balls and those getting to four have both slightly increased over the past few years overall. In 2006 Nick was below average in both respects, and in 2007 he was about average. In 2008, there was an increase in three-ball counts per plate appearance to 33% better than the league average (that’s comparing the two rates, and not a straight difference), and a big spike in walk rate to 66% better than league average. Had I seen that previously, my inclination would have been to think that the walk rate wasn’t really sustainable at such a high level given that disparity. In 2009 though, things went the other way. Nick got to a three-ball count a touch more than average, but ended up walking 6%* less than the average hitter. That seems low to me, and Markakis “should” have actually walked around 9% of the time. It’s nice that 2008 and 2008 are the years were there were disparities, since those are the ones that I have Pitch/FX data for.
* The difference wasn’t as big as it originally appeared, since Nick didn’t get a single intentional walk in ’09. It seemed to be a league wide thing. After years of 530, 532, and 529 intentional walks, that number was down all the way to 409. Weird.
Here’s the pitches Nick saw in 2008 with three balls. The ones he swung at are in red, and the ones he took are in blue.
Of the 63 pitches out of the zone, Nick only swung at 14 (22%). Those pitches that he did swing at were mostly quite close to the strike-zone (especially if you were to account for the fact that lefties get pitches off the outside corner called strikes more often than right-handed hitters), and 10 of them were on 3-2 counts. That discipline helps explain how Nick was able to convert 52% of his plate appearances getting to three balls into walks (well above the league average of 42%).
Now here’s 2009:
Nick swung at 21 of 51 pitches out of the strike-zone, which at over 41% is nearly double the rate from 2008. Those pitches that he swung at tended to be a little farther away from the zone than the comparable pitches were from the previous year as well (though umps did appear to be calling pitches that were a little farther off the corner in ’09, so I’d forgive him for some of those outside pitches). Interestingly,the same proportion of those 21 swings came with two strikes (15 out of 21, versus 10 out of 14 above). Overall, Markakis was able to turn a three-ball count into a base on balls only 38% of the time (league average still 42%).
In 2008 it looks like pitchers were either going to challenge Nick with a pitch in the middle of the plate (in which case he’d swing) or go outside the zone (in which case he’d take). Last year, it appears that they were trying harder to throw strikes on the outside portion of the plate, and if Markakis picked up on that then he might be more inclined to swing at pitches that weren’t as close as in ’08 because it was harder to differentiate a ball from a strike.
So since pitchers were willing to throw strikes in three-ball counts more often in 2009 (63.3% of pitches in the zone versus 59.6% in 2008), perhaps Nick thought he better start swinging more overall – and perhaps it just so happened that the more aggressive approach resulted in the jump in balls swung at (since the rate of strikes swung at was already high).
Even in the best case scenario, with Nick not swinging at any pitches outside the zone in 2009, the ump calling all of those balls, and assuming none of those swings occurred in the same plate appearance (which some did, but go with it), Markakis’ walk total would still only have gone up by 21. That is a lot of walks, but it only brings his total up to 77 – that’s a 10.8% walk rate (almost exactly half-way between his ’08 and ’09 walk rates) – and would have been a walk in 52% of his three-ball counts (the same as 2008). Even if Nick was relatively perfect when he got to ball three in 2009, he still wouldn’t have come close to repeating his ’08 OBP. Realistically, he would have upped his walk rate by ~1% by taking an approach more consistent with 2008. Part of that is on the pitchers in these situations, and part of that is on Nick working the count to get to three balls. That’ll have to wait until part five.