SB Nation - Login for mobile commenting

Lookout Landing

Fastballs, the 2009 M's, and the New Acquisitions

Jeff and Matthew both have good posts below this one, so make sure to read them too.

---------

I've been wondering lately whether it's possible for a team to be vulnerable to a certain pitch type. It certainly seems possible. Team overloaded with left-handed bats will fare badly against changeups, a right-handed lineup might have problems with breaking balls. A bad lineup might have trouble with fastballs.

Poking around Casey Kotchman's Fangraphs page today led me to thinking about using their pitch value data in order to take a quick look at whether a team might exhibit splits of this nature. Kotchman's still pretty good at hitting fastballs, even if he can't hit anything else, so, I reasoned, perhaps the 2009 Mariners were deficient at hitting the heater, thus making Kotchman a better fit? Well, beyond their usual standards, that is. Chone Figgins and Milton Bradley are both above average fastball hitters too (hardly surprising since they're both still good at hitting, which is more than our new first baseman can say), but I suspected that there might be something to the idea that the team might not have been very good at hitting fastballs.

And boy, was I right. The table below is a set of rankings for the AL teams for three pitch types: fastballs (four seamers, two seamers, and cutters), breaking balls (sliders and curves), and changeups (changes and splitters).

Table 1: AL Rankings by pitch type, denoted r(XX) where XX is fastball (FB), breaking ball (BB), or changeup (CH).

The Mariners were dead last in the American League in hitting fastballs last year. The Royals can't hit breaking balls or offspeed pitches worth anything, but they can still hit a fastball much better than we could. The table's interesting in other ways, too. The Indians, for example, were collectively excellent fastball hitters but awful at hitting anything that moved. The Yankees? Good at everything. The Rangers were below average when it came to hitting fastballs or breaking pitches, but god help you if you throw a changeup to them. Etc, etc, etc. For the curious, the Mariners contrived to be a magnificent 44 runs below league average at hitting fastballs and cutters, a number rivaled only by Oakland.

So, why would a team be bad at hitting bendy/movey things and awful at hitting fastballs? Part of it is that fastballs are easier to hit and so the bar is higher, but I wonder if it might be because our poorer hitters are the ones with a semblance of plate discipline but nothing else. We can lay off sliders in the dirt, or what have you, but we can't punish a fastball over the plate (well, nobody but Lopez, Guti, and Ichiro can). That certainly seems to fit the type of hitter the Athletics carry too. Regardless of what caused our fastballitis last year, the hitters we've brought in this offseason have no such difficulty with heat, with the group coming in at around 40 runs above average against fastballs last year (in a down year for both Kotchman and Bradley). With any luck, the M's ranking last in the AL against fastballs will be a curiosity limited to the past.

4 recs  |  39 comments

Comments

Bravo

You guys never fail to impress with your detective work

Graham wins! Graham wins!
You should also note Graham

That the run values attached to pitch type rankings do have a lot of noise in them. Basically, they are linear weights based off of the actual outcomes of each pitch. Accordingly, they will be heavily influenced by BABIP and sample size issues.

That doesn’t change the conclusions of your post, that the Mariners will be a better fastball hitting team next year – however, the reason has a lot to do with simple regression to the mean as well as the additions of new players.

The issues you raise are going to be a lot more true of individual players than teams.

15,000 fastballs in a season is not a small sample size.

Isn't this more a reflection of good offenses hitting everything?

I mean, Yankees, Angels, and Red Sox all had good offenses last year and obviously had no problems with anything. Mariners and Royals had a shitty offense last year and have problems with everything.

I could be mis-reading you,

but it seems you’re saying that those three teams hit everything well BECAUSE they had a good offense, when really it would make more sense to flip the cause-and-effect and say that the ability to hit anything is what caused them to have a good offense.

Which may seem like pointless semantics at first,

until you look at it from a perspective of “how do we improve the offense?” If we treat the “good offense” as the cause and the ability-to-hit-any/all-pitches as the effect, then we’re not provided with any useful information at all as to how to improve the offense and can only logically deduce that to improve the offense we must simply improve the offense. Not wildly helpful. However, if we flip the cause and effect, we’re able to break it down to smaller component pieces which we can more readily understand and more easily address. Much like how we break down offense in terms of average, on-base percentage, slugging, and baserunning so we can find out exactly what we’re lacking (OBP) and where we can most readily improve.

Or I'm reading too much into it and you're simply stating that "good offenses hit well and bad offenses don't"

which is both obvious and uninteresting, so I don’t suspect it to be the case.

I think the keyword in his writing was "everything."

Both the Royals and the Mariners had awful offenses, thus everything was bad (14th, 10th, 13th). Ergo, perhaps in their case, an upgrade at hitting fastballs specifically may not be necessary, but rather simply hitters that hit anything well.

I have no opinion either way but apparently it’s become my schtick to clarify things for people.

Basically your performance against fastballs and other stuff is strongly correlated.
Actually you are correct

When you graph the rankings of one category vs another you can see that if you do better in any category, you will do better in the other. If you do a simple regression using the table of rankings Graham has above you can get:

FB = 0.52*BB + 3.6
R^2 = 0.27

FB = 0.44*CH + 4.2
R^2 = 0.20

BB = 0.52*CH + 3.6
R^2 = 0.27

FB = 0.48*(BB+CH)/2 +3.9
R^2 = 0.30

These don’t tell you what are the causes of good performance but it does show a trend in the data.

I did a full blown regression between all pitch types on fangraphs using team data

(excluding knuckleballs and only used AL teams)

There is only one relationship that is statistically significant which is between fastballs and curveballs (p = 0.02). If you are good at hitting fastballs then you are good at hitting curveballs.

The next closest relationship says that hitters good at hitting sliders are also good at hitting cutters (p = 0.12) but this isn’t a very significant relationship. It might also be caused by misclassifying sliders as cutters and vis versa (I seem to remember these sometimes get confused).

Everything else that the regression analysis spits out isn’t too interesting because it isn’t statistically significant however you can still kind of pick out some trends.

Teams that tend to hit one pitch well almost always hit other pitches well.
Teams which are good at hitting sliders are bad at hitting fastballs.
Teams that are good at hitting fastballs or sliders are bad at hitting splitfinger fastballs.
Teams can hit change ups well except if you are good at hitting cut fastballs.

One thought of mine is that a lot of this analysis is skewed by the handedness of the teams which prevents us from really drawing the interesting conclusions we are interested in.

I can post more if people are interested but its a lot of numbers that get pretty overwelming.
This is more what I was striving for.

It seems self evident that good offenses would hit most / every type of pitch well while bad offenses wouldn’t. The real interesting ones are the teams like the Indians, who clearly have a penchant for hitting the fastball over the other pitch types.

Yep

The only “outliers” I see are the indians, angels, blue jays and royals. The only really big outlier is the indians.

This is awesome!

How did you get the data for this?

Fangraphs has all this data in their splits
Thank you Graham.

This was interesting.

I think the Yankees' fastball numbers are skewed.

They never have to face Mariano Rivera’s cutter.

Very interesting angle and very well conveyed.

Because runs occur when hits and walks come in bunches, could improving a specific weakness help the overall offense? Or make the Mariners offense more consistent—if such a thing is even possible in baseball—and less feast or famine?

Is there a difference

between your r(XX) and Fangraph’s pitch value w(XX)? If so, where did you get r(XX) from?

Other than the consolidation of all pitches into 3 major categories...:)
What I'm curious about is...

Could this be a new way to arrange your lineup? It seems like most lineup decisions revolve around whether the pitcher/players are a lefty or a righty and then whether they’re more OBP oriented or SLG oriented. We all know, at a more intuitive level, that certain teams are going to struggle against certain types of pitchers. Could this take it to the next level?

Or, just for simple predictive purposes, should Scott Feldman try to pitch every series against the M’s?

No team would be so much worse against a given type of pitch in one game that it would make sense to pitch an inferior pitcher with a better pitch (x) to take advantage of it
But,

“our poorer hitters are the ones with a semblance of plate discipline but nothing else”

Isn’t that more of a possible statement for 2010? It’s not like we were drowning in hitters with plate discipline last season. Yuni, Beltre, Kenji, etc weren’t people laying off sliders but unable to hit fastballs.

I was thinking more

Chavez, Langerhans, R. Johnson, Wilson, Hannhan, Sweeney etc.

From watching Toronto last year

that rCH ranking isn’t surprising. Never throw those guys changeups.

What's really the best part of this is that the front office probably knew all of this already and planned accordingly.
Kotchman's wSL/100, wCT/100, and wCB/100 all dropped around 2 runs from 2007-2008 levels

This was combined with a decrease in his contact rate by 2-3 % from his 2007-2008 levels. The only area that he improved in was that his swing % went down for both inside and outside the zone by quite a bit, 5 % from 2008, more from 2007.

The numbers suggest that he has become crappy at hitting breaking balls when before he was half-decent. He is taking on the model you put up; of laying off out of zone breaking balls more as he makes less contact on them.

I think there is too much noise in the w(pitch)/100 type of stats to try to

get too much out of them when we talk about yearly figures.

now we just need to sign Pedro Cerrano

he can hit the fastball like nothing I have ever seen before

We have a few Pedro Cerranos in the minors.
I seriously doubt that
Greg Halman has slightly lighter skin, I guess.
Is this assuming a RHP...?

I consider myself pretty damn knowledgable but it’s never been explained to me why a LH would be vulnerable to a change, regardless of the pitcher’s handedness. Ergh.

*"More inherently" vulnerable.
Changeups are only really dangerous to opposite-handed batters due to the way they move

Therefore left handed bats are vulnerable to right handed changeups, and there are more of those.

You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.