Yesterday worked out perfectly. Late Tuesday night, I sent the link to this very interesting fanpost from Dr. Rick out to the rest of the TCB staff. What it did was spurn a pretty in-depth discussion among a few of us about Effective Velocity and the value of a change up.
I just think that those stats are not truly definitive on his pitch due to improper usage. Statistics outcomes are only as useful as the quality of the input data. If a pitch is not used properly, the input is flawed which results in a flawed outcome measure.
This is a quote from me in the email in discussion about Bud Norris' change up since it was discussed several time as to whether he should stop using it. All of the outcome measures pointed to it as a poor pitch. However, the change up is more reliant on pitch sequencing and the previous pitch than any other pitch and if used improperly, it's risk of being unsuccessful increases.
Dr. Rick did a good job of explaining Effective Velocity in that fan post, so be sure to read it. But, I'm going to apply it to the change up.
Take a pitcher who throws a 91 MPH fastball. If that pitch is located in the "7" zone (picture a phone key pad), it is perceived as 88 MPH. Now, take a change up that averages 85 MPH and consider where that pitch can be thrown. That's a 6 MPH difference from standard velocity but is only 3 MPH difference from the perceived velocity. That means The pitch can only be located in the 4, 7, and 8 zones in order to not be within the 6 MPH range (3 MPH above and below previous pitch's perceived velocity). That requires a very good feel and command of the change up.
The discussion evolved into Chris pointing out that the after a large enough sample size, all pitch values have to have some predictive weight to an extent. I have to agree to a degree only in part that after awhile, if there's not an improvement in sequencing, the pitcher is unlikely to do so without a change in the organizational philosophy, pitching coach, ect.
So, I asked Chris what made a change up a bad/good pitch.
What could make a changeup bad, in my opinion:--too small a variation in velocity compared to fastball (mostly would be a problem for guys with low FB velo)--no break at all (coming in flat)--unable to locate it for whatever reason (poor command/control)...maybe because the grip is uncomfortable or unpracticed?
Break is a factor, but not a huge factor. The movement can assist in increasing the whiff rate and even ground ball rates, but I don't think it's required to be effective. Lack of movement won't make a change up bad.
The other thing to consider is something that Clack pointed out.
The pitch values are based on the linear weights (like wRC+) of the outcomes for the pitch. So it is susceptible to both BABIP-luck issues (depending on sample size), as well as pitch sequencing issues. For example, the change up might be more effective than it’s pitch value because it’s real value for the particular pitcher is to set up the next pitch. So, the unrevealed value of the change up might be reflected in the values of other pitches which are set up by the change up.
I'll let ashitaka explain it to you:
Something that jumped out from Sickles' piece on him:
"In the minors he has shown a decent changeup, although he didn't use it much in his first major league outing according to both pitch f/x and people who saw the game."
So I thought I would look that up:
--> Total pitches: 725
--> Change ups: 61 (8.4%)
--> Total pitches: 1509
--> Change ups: 213 (14.1%)
--> Total pitches: 692
--> Change ups: 61 (8.8%)
Though it's a limited sample of just three partial seasons, it's worth noting that in 2012 he posted his best K/9, BB/9, GB% and xFIP when throwing his change up significantly more. Maybe it's not so much the change up, or how good it us, but his talent for sequencing it, like you guys were talking about earlier. No way to know, but I thought it was worth pointing out once I saw it.