osh Johnson is most definitely an underrated young pitcher. I don’t think there’s any questioning the notion that his playing in Florida has been an obstacle to his getting the attention his results would warrant. But elite? I’m not yet convinced.
Why is he not convinced? Sun Life Stadium.
The ability’s all there, but I’ll need to see him show it in more than one ballpark.
Essentially, there seems to be something of an effect on strikeouts in Sun Life Stadium, and that effect may be inflating strikeouts for Johnson above his supposed true talent level. This effect, which seems to be real based not only on this 2008 David Gassko study but with subsequent work on StatCorner showing similar numbers, is probably positively altering the perception of our pitching staff as a result, and Sullivan thinks that it has helped to make Josh Johnson a bit overrated among those who say he’s vastly underrated (what a mouthful).
And why wouldn’t he rack up big numbers in the park he is most familiar with? Should he be penalized because others haven’t learned to take advantaged of their home parks?
Another thing: when did strikeout rates become a litmus test for being an elite pitcher?
Jeff continues on in the article and brings up career FIP at home and on the road, never taking into account that JJ’s career numbers start when he was 21 years-old or the fact he is now 27 and healthy. Weighing every year the same is a fallacy.
This sounds like a heated argument, one into which I should probably jump in, being both a “stats guy” and a diehard Marlins fan. What follows is my initial digging into this issue; I do not know if this is anything conclusive or if it is a start to any major analysis, but it is something to look at and wonder. I’m going to present the statistics here and discuss my first impressions after looking at the stats.
Park factors for strikeouts?
Some readers of Sullivan’s article and Craig’s response expressed initial doubts regarding how parks can possibly affect strikeouts, a stat that appears to be independent of the pitcher / batter confrontation. The truth is that strikeouts are absolutely affected by the park and more importantly the location of the park. In addition to all the potential differences regarding lighting during night games and shadows during late afternoon starts, the biggest aspect of a park that could affect strikeouts is the humidity of its location. Commeter dcfish over at Fish Stripes says it best.
The humidity. Humid conditions tends to make the ball heavier, meaning that breaking balls will have a lot more sink and “bite” to them. It’s the opposite effect from Coors Field (though not nearly as strong), where the conditions are very dry, so curveballs and such do not break as much.
The humidity increases the movement of pitches and thus likely assists pitchers in getting strikeouts. One can see the effect in the opposite end of the spectrum in Colorado, as the Rockies have the lowest strikeout park factor in baseball according to that study. These effects are definitely real, and these numbers are simply a way of representing their potential magnitude.
Josh Johnson’s home / road splits
One of the concerns a lot of readers brought up was that Johnson’s career statistics encompass a time when he was younger and not as good. To address that concern, I figured we should look at just his past two seasons worth of work, between the 2009 and 2010 years, to consider if the effect has persisted into his supposed good seasons. Note that these two seasons are both years following his Tommy John surgery and recovery; if these are not considered his prime seasons, then I do not know what is.
Johnson, Home / Road
This does paint a picture of what Sullivan said, especially when you compare it to the home / road split that has been observed in the majors in that same time period.
Major Leagues. Home / Road
As you can see, there isn’t a huge split between strikeouts home and away, and the run differentials via both ERA and FIP are also significantly lower than Josh Johnson’s observed difference. The difference between run totals normally seems to be between 0.3 and 0.5 runs per game, while Johnson’s differentials on both ERA and FIP are on the order of 1.5 runs per game. That difference seems pretty extreme, and it does seem to be heavily tied to Johnson’s strikeout rate, as his other home / road splits are not nearly as extreme.
Johnson, Home / Road (other)
The walk rates are fairly similar (probably within league average splits), and the ground ball rates and BABIP are almost identical. The home run rates are lower and that is understandable (if a bit extreme) given the home run park factors for Sun Life Stadium (0.97 five-year regressed park factor according to Patriot). I’d be willing to say that maybe half of the difference between the normal home / road split and Johnson’s home road split is explained by home runs, while the other half may be explained by strikeouts.
Does this mean that we should use only the road numbers to determine Johnson’s skill? Absolutely not. Yes, we know that the park is inflating his strikeouts a bit, but if we know the general extent of the strikeout effect in Florida, then we have an estimate of how muC. Johnson is adding outside of what we think is the park’s effect. If Sun Life Stadium indeed adding approximately 11 percent more strikeouts on average and that the average pitcher at home throws about 1.1 percent more strikeouts (assuming the road numbers are in overall neutral parks), then we would expect Johnson’s home strikeout rate to be closer to 21.6 percent.
If all those assumptions are true, where are those remaining strikeouts coming from? Is he leveraging his park better than usual? As Craig mentions in his post, is that something for which we should discredit him? What’s the difference between “leveraging your park better” and “needing your park to succeed?” These are all interesting questions.
But are they even important. Consider the type of pitcher we’d be seeing if Johnson had struck out the number of hitters we expected him to get out. Calculating his strikeout rate using that 21.6 percent home strikeout rate, we get a two-year strikeout rate of 20.0 percent, different than his actual 23.5 percent. He lost just about 55 strikeouts in the “correction.” What was his FIP in that span using that corrected rate? He gets a FIP of 3.07. As I mentioned in my season preview article for JJ, most of the projection systems have him between an ERA of 2.87 and 3.30, which corresponds fairly well with an ERA of 2.94 from 2008 to 2010. Essentially, this “correction” brings Johnson closer to his projected performance, meaning that the projection systems almost already took this into account. The projections all came out with an ERA in the high 2.00 and low 3.00, which is exactly where we pegged after correcting his strikeout rate.
So is this interesting find potentially important? Yeah, I think so. But I do not believe that it excludes Johnson from being “elite.” What it does is show that there is something about being home, whether it is solely Sun Life Stadium or something about the status of “home” in general, that gets Johnson to another gear. I don’t know what it is, and I don’t think we can tease it out or make any conclusions from it. It simply is what it is.