King Indeed

I have mentioned a few times about how Zack Greinke gets nowhere near enough press attention, especially for his level of success this year. Well, as far as the latter point goes, move over Greinke because here comes Felix Hernandez.

Take a look at Felix’s last eleven starts, from a Matthew point of view (that is, looking at the numbers that I care about):

3752816231_d7dd51dc0b_o

I prefer tRA to FIP, but even if you do not, you should get the general picture here. A 12% swinging strike rate coupled with over 50% of batted balls being on the ground? Yee-haw!

For years and years, the Seattle blogosphere cherished and also admonished Felix Hernandez, cursing him for throwing too many fastballs, which many of us felt, and the pitch type values back up, was/is Felix’s worst offering. Which is not to say that it is bad, certainly not, but rather that it is a testament to just how good his off speed and breaking pitches are.

A funny thing happened though, instead of throwing fewer fastballs this year (64.2% versus a career mark of 60.8%), Felix has instead simply thrown better ones. It was always one of the possible paths of improvement that we in Seattle talked about, it just seemed the least likely and more difficult of the two. Perhaps throwing all those fastballs has finally paid off with improved command. We will never know. Frankly, I do not much care. I know I get to watch this guy pitch every five days or so, hopefully for a long while, and that is all I need right now.





Matthew Carruth is a software engineer who has been fascinated with baseball statistics since age five. When not dissecting baseball, he is watching hockey or playing soccer.

23 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
truth
14 years ago

great article Matthew. I was unfamiliar with tRA whilst reading this article, so I did some research and found some research done by tango tiger and dave cameron on it. tRA appears to be an even more accurate barometer than FIP of a pitcher’s true abilities. Any chance you guys add that to the player pages?

Toffer Peak
14 years ago
Reply to  truth

tRA uses HRs allowed and LD%, both of which have been proven to be beyond the pitcher’s control, (outside of FBs allowed) and therefore it’s not necessarily the best luck-neutral run estimator. Though FIP uses HRs allowed as well. To keep things simple it would probably be best if FanGraphs simply replaced FIP with xFIP.

Fett42
14 years ago
Reply to  Toffer Peak

There’s also tRA* however.

Toffer Peak
14 years ago
Reply to  Toffer Peak

The problem with tRA* is that as of now on StatCorner it only regresses the current season stats toward the league mean. There are a few things wrong with this. HRs and LD% need to be, or should be, regressed significantly more than stats such as BBs and Ks which have been proven to be under the control of pitchers unlike HRs and LDs. Also at the beginning of each year tRA* throws away everything that it knows about a pitcher’s previous skills. Should we really be regressing Lincecum’s K rate toward the same league average as we do Carlos Silva’s?

Alex
14 years ago
Reply to  Toffer Peak

Correct me if I’m wrong, but doesn’t tRA* regress numbers based on how quickly they stabilize (meaning HR and LD would be regressed more than BB and K)? I agree wholeheartedly with the second part of Toffer’s comment though – it seems as though there has to be a better way to take into account what we know about the pitcher in the regression. Wouldn’t the current method of doing artificially inflate top pitchers’ tRA*s (as they are regressed each year as though they were rookies, I think)

B
14 years ago
Reply to  Toffer Peak

“tRA uses HRs allowed and LD%, both of which have been proven to be beyond the pitcher’s control, (outside of FBs allowed)”

I’d be interested in this research if anyone wants to share the links. I’m pretty skeptical of the conclusions that certain results are totally “beyond the pitcher’s control”. Just because a number tends to regress towards a league wide mean does not necessarily mean a pitcher has no control over it (or that it’s actually regressing towards the league mean, each pitcher may have a slightly different mean of their own) – so it may be they have a very limited amount of control over it. I’d like to read the research before judging whether this conclusion is off base or not, though.

I see a lot of problems with LD% in general, so you may be right that it’s not a good stat to include. My guess is there will often be balls hit where the only difference between whether it’s classified as a LD or FB is whether it happens to fall in for a hit. In those situations, rather than LD% causing hits, hits are causing LD%.

“To keep things simple it would probably be best if FanGraphs simply replaced FIP with xFIP.”

I’m under the impression FIP has been tested to be a better stat than xFIP. If someone thinks otherwise, I’d be interested to see the evidence.

Graham
14 years ago
Reply to  Toffer Peak

“HRs and LD% need to be, or should be, regressed significantly more than stats such as BBs and Ks which have been proven to be under the control of pitchers unlike HRs and LDs.”

Perhaps it would be better to understand the metric more deeply rather than making false accusations which are quite frankly an insult to my intelligence.

Toffer Peak
14 years ago
Reply to  Toffer Peak

Graham – As I admit above I may have exaggerated to make a quick, simple point. I’ve also linked to lots of sources that have led me to these conclusions. Since you’re the creator of tRA maybe you wouldn’t mind explaining my stupidity rather than pouting? Links to previous work would be fine.