Friday, October 15, 2010
Fluke improvement watch: Gio GonzalezPosted by Josh Smolow at 3:55am
When looking at a player whose numbers have seemingly improved a great deal, it's easy to convince yourself that the improvement is clearly real. And after you do, for every warning sign that the improvement is simply mirage, you come up with an excuse for why those warning signs don't apply. But at some point you have to realize that the player you've become fascinated with is probably not going to continue being the player he was in the past.
That paragraph can apply to any statistical way that you look at a fantasy player. In this article, I look at Gio Gonzalez, a pitcher who seems to have made humongous strides this year. Moreover, he's a pitcher who has a single dominant pitch, making him extremely easy to want to like as a (fantasy) player. Unfortunately, as I'll explain below, It doesn't look like, if things remain the same, he can be a pitcher who can sustain an ERA under 3.50.
Gonzalez' numbers this year are drastically up from last year, as seen in the following table:
As you can see, in several relevant fantasy statistics (walks, ERA, WHIP), Gonzalez clearly improved this year, while decreasing in only one fantasy metric (strikeouts). However, his context-neutral statistics (FIP, xFIP) and his BABIP would indicate that at least part of this improvement has been a mirage (xFIP would suggest he's actually gotten worse). What should we believe? Well, let's look at his pitches and see what's changed from year to year:
Figure 1: Graph of the movement of Gonzalez' pitches in 2009 and 2010.
Vertical movement: the number of inches the ball drops/"rises" as compared to how we would expect gravity to make a pitch drop. So a fastball with Positive 10 Vertical Movement "rises" 10 inches more than it should if gravity was the only force acting on it and a curveball with -10 vertical movement drops 10 inches more than a pitch thrown that is just acted on by gravity.
Horizontal movement: The graph is from the view of a catcher or umpire behind home plate. So a pitch that's on the left side of the graph (and has "negative horizontal movement") moves in on righties and away from lefties. A pitch that's on the right side of the graph moves in on lefites and away from righties.
Gonzalez uses four pitches: A four-seam fastball (colored red on the graph), a two-seam fastball (dark yellow on the graph), a change-up (blue), and a curveball (green). As we can see on the graph, in 2009, if he used the two-seam fastball at all, it wasn't very different from his four-seam fastball. This year, Gonzalez managed to make his two-seam fastball clearly distinguishable from his four-seam pitch. The two-seam fastball now drops three more inches than the four-seam fastball. Gonzalez also has started to use this two-seam fastball far more frequently than last year.
Aside from the new, improved two-seam fastball, the rest of Gonzalez' repertoire remains unchanged. Even the velocity of these pitches has more or less remained constant. So is this change enough to explain his improvement? Let's look at the results of these pitches the last two years:
|Year||Pitch type||Batter handedness||Number thrown||Whiff rate||Swing rate||Swinging strike rate||Ground ball percentage||Strike zone %||BABIP||SLGBIP||HR/FB||RV100||RVe100|
Table 2: The results of Gio Gonzalez' pitches each of the last two years.
Whiff rate = Swinging strikes/swings
Swing rate = Swings/total pitches
Swinging strike rate = Swinging strikes/total pitches
Strike zone percentage = Percentage of pitches located within a wide (two foot) strike zone
SLGBIP = Slugging percentage on balls put in play (Also called SLGCON)
RV100 = Run value per 100 pitches (A measure of pitch effectiveness)
RVe100 = Expected run vllue per 100 pitches (A measure of pitch effectiveness, controlling for luck on certain batted ball types.)
I apologize for this table being a little messy, but I'll highlight the interesting points. Gonzalez faces a lot more right-handed batters (635 batters) than left-handed batters (186 batters) because he's left-handed. Lets go through these results against each type of batter to see what's changed result-wise in this latest year.
Against left-handed hitters, he uses mainly two pitches, the four-seam fastball (51.3 percent of the percent). Regardless of what pitch he uses, Gonzalez has always thrown almost xclusively away from left-handed hitters. (He pitches up with the four-seam fastball and down with the curveball). This is true both this year and last year.
Result-wise, there are a few changes in 2010 from 2009. On a positive note: His four-seam fastball was amazingly effective against LHBs, with a swinging strike rate of 9.66 percent (compared to a rate of 5.32 percent in 2009), which is much higher than that of a typical fastball (usually around 6 percent). The two-seam fastball has also been fairly effective because of its high GB Rate (60 percent). On a negative note: Gonzalez' curve ball had its swinging strike rate decrease from 19.15 percent in 2009 to a minuscule (for a curveball) 7.44 percent rate in 2010. The pitch has been a better groundball pitch this year against LHBs, but it's not been enough to make up for the large loss of swing-and-misses.
But remember, the question is: are any of these changes likely to continue in 2011? The answer for these changes would seem to be mostly no; there's basically no change in the movement of any of these pitches (except for the two-seam fastball, which is not used frequently against LHBs). Moreover, there's basically no change in the location of these pitches, either: Gonzales' fastball, for example, hit the strike zone against LHBs at the same rate as last year and in the same basic location. The only result that might be explainable is that of the curveball: While Gonzalez' aim with the pitch was basically at the same place as last year, he hit the middle of the strike zone more frequently this year. (You can see that in his increased strike zone percentage), which might be the reason batters didn't miss as often against the pitch. Thus, his results against left-handed batters next year would seem likely to decline somewhat. This is because we should expect his fastball to come down to earth a little, while his curve ball shouldn't regress as much to match the 2009 numbers.
Against right-handed hitters, Gonzalez uses all four of his pitches. He throws the curve ball (29.84 percent of the time), four-seam fastball (29.04 percent) and two-seam fastball (31.32 percent) equally often, while going to the change-up 9.78 percent of the time. In practice, the curveball is used as his out pitch on 0-2, 1-2, and 2-2 counts. Meanwhile, the two-seam fastball is used often with no balls or one in the count, but gives way to the four-seam fastball in two- and three-ball counts. Gonzalez aimed the curveball middle-low in 2010. He used both of his two fastballs differently in 2010: the four-seam fastball aimed clearly inside on right-handed batters, while the two-seam fastball aimed away.
It's clear that Gonzalez possesses one of the nastiest pitches a right-handed batter will ever face in his curve ball. The pitch gets a swinging strike 17.4 percent of the time, and when it's put into play, results in a ground ball 82.47 percent of the time. That's just amazingly effective, with a total run value of -21.06. To put that into perspective, it's basically equivalent to Adam Wainwright's level of dominance with the pitch.
Gonzalez also uses his new and improved two-seam fastball extremely effectively against right-handed batters. The pitch has a barely mediocre swinging strike rate of 4.72 percent, but gets ground balls 46.92 percent of the time, which is far better than his fastball ever had gotten previously. It's this pitch that has resulted in his increased ground ball rate. Moreover, the pitch has a HR/FB rate of 3.44 percent, which is tiny.
Gonzalez's other pitches also showed changes in their results this year. Gonzalez' four-seam fastball decreased its swinging strike rate from an average 6.4 percent to an abysmal 2.76 percent, while getting fewer ground balls than ever (25.7 percent). His change-up has also lost some swinging strikes but has increased in ground balls.
But remember once again, the question is: Are any of these improvements likely to continue in 2011? The answer, once again, is not positive. Gonzalez' curve ball, this year had a .247 BABIP compared to last year's .333 BABIP. You'd expect the BABIP on the pitch, especially as it's an extreme groundball pitch, would increase at the very least to .300. The curveball is likely to stay a terrific pitch next year, but regress a little. Similarly, the two-seam fastball's HR/FB rate is unlikely to remain so incredibly low. As a result that pitch should get worse as well.
Meanwhile, while the four-seam fastball has gotten worse, it's the only pitch whose location Gonzalez significantly changed from last year. Last year, he'd throw the pitch middle-away, but this year he threw it almost entirely inside. This could explain the decrease in the swinging strike rate on the pitch. The pitch's performance should regress a little because of its .364 BABIP, but not as much as the other two pitches.
One last thing: Remember how Gonzalez's walk rate had declined this year? Well he hit the strike zone at the same rate as last year... so even that rate might decrease back to his abysmal 2009 walk rate.
The result of this look at Gonzalez' pitches is to conclude the same thing we might have if we looked at nothing but his xFIP and his ERA: The kid is clearly due for a decline. If I had to project, I'd think that he's likely to regress to the 4.20 ERA we might expect from his xFIP this year. That said, it is important to remember that Gonzalez is young and is still a developing pitcher. His pitches could emerge from the offseason different from how they've been the last two years. If they do, then the above analysis should be redone. So keep an eye on that. But for now... don't expect him to put up another 3.35 ERA next year.
Josh Smolow wishes that one day pitching coaches would seem to actually, you know, do something. You can contact him on twitter under the name garik16.