Swinging Strike Rates

[shelmow]

I just read the article on second half surges and its been one of several that has used swinging strike rate to evaluate pitchers. Is there a place on the baseball hq website where data on all pitchers swinging strike rates?

[RAY@HQ]

Not yet, but it's coming very soon.

[shelmow]

Thanks ray

[Keano's Magic Hat]

Any other PitchFX type stuff we can look forward to?

[RAY@HQ]

Just the SwStr% stuff in this first pass (and some other Playerlink redesign stuff). Will revisit in the offseason...

[Keano's Magic Hat]

Not critical because its so easily available elsewhere for free.
Fascinating stuff to consider in light of the HQ skills vs results mantra.

No offence but that data makes metrics like BPV pretty meaningless.

[RAY@HQ]

No offence but that data makes metrics like BPV pretty meaningless.

No offense taken, but at the same time I couldn't disagree more, actually. And that's why we haven't jumped into the pitch-type data with both feet. Our approach is to bring in those data elements as those data elements prove to tell us something different than our traditional metrics.

I've used this example before, but it still holds: one of the pieces of data we might look at bringing in to HQ is velocity data. It's pretty widely available now, as you say. But does it tell us anything we don't really know? Here's an exercise: grab that data and run a search on the top 4-seam fastball velocities. Now come back to HQ and run a search on the top k/9 rates. You get pretty much the same list of pitchers. Velocity data on its own isn't telling us anything new.

Now, studying changes in velocity rates gets interesting. If you see a pitcher with a change in k/9, you want to validate that change with some of the other data elements that are out there: for instance, swinging strike data is great for that. That's why we're bringing in that data now. Velocity could have similar utilities, so could the movement data that Stephen Nickrand periodically references in the SP Buyers Guide.

As these new metrics get proven to tell us things that our traditional metrics aren't capturing, we'll bring more of them in to our data set.

But BPV meaningless? Abso-freakin'-lutely not.

[Gantner]

Yes. I like Pitch FX for analyzing smaller sample sizes. Over a handful if starts a SP still throws enough pitches to make for meaningful analysis that can tell us whether a swing in Dom over that period may be significant.

Similarly I like looking at xPX vs PX over small sample sizes for batters. Helps remove the H% noise.

[Keano's Magic Hat]

No offense taken, but at the same time I couldn't disagree more, actually. And that's why we haven't jumped into the pitch-type data with both feet. Our approach is to bring in those data elements as those data elements prove to tell us something different than our traditional metrics.

I've used this example before, but it still holds: one of the pieces of data we might look at bringing in to HQ is velocity data. It's pretty widely available now, as you say. But does it tell us anything we don't really know? Here's an exercise: grab that data and run a search on the top 4-seam fastball velocities. Now come back to HQ and run a search on the top k/9 rates. You get pretty much the same list of pitchers. Velocity data on its own isn't telling us anything new.

Now, studying changes in velocity rates gets interesting. If you see a pitcher with a change in k/9, you want to validate that change with some of the other data elements that are out there: for instance, swinging strike data is great for that. That's why we're bringing in that data now. Velocity could have similar utilities, so could the movement data that Stephen Nickrand periodically references in the SP Buyers Guide.

As these new metrics get proven to tell us things that our traditional metrics aren't capturing, we'll bring more of them in to our data set.

But BPV meaningless? Abso-freakin'-lutely not.

A few years ago Ron wrote here that BPV formula was adjusted to more accurately correlate with results, and in fact was tweaked to heavily favour GB pitchers. From that point it was apparent that BPV *isnt* a tool for evaluating skills, its a tool for evaluating outcomes. It has some value, ie High BPV generally means Good Pitcher, but measure of skills it isnt.

Thats all I meant. Tools like BPV (and era, whip) are important, but the pitch analysis stuff lets us really understand what individual pitchers are doing to get those results, and what criteria might be good lead indicators. I think we're a fair way off that understanding, and the more intelligent people using the data and publishing their conclusions the better.

[RAY@HQ]

I guess I don't understand why you think including GB% in BPV devalues the metric, or makes it less skill-focused.