Nate Schierholtz and the back-foot slider

[gp956]

I've noticed throughout this year that Nate has done a good job of laying off the back-foot slider from right handed pitchers.

In 2009, and most of 2010, you could count on a right-handed pitcher attacking Nate with the down and in slider whenever they needed a strike. He flailed so badly, at times, that he actually swung at a couple of sliders that ended up hitting him.

Recently, Grant, at McCovey Chronicles wrote an article where the thesis was the Giants continuing to find a way to win despite their obvious flaws. And, while I agree with the theme, one point Grant tried to put out there was Nate hitting "well" despite flailing at back-foot sliders. I objected to that, and remarked "does Grant even watch the Giants anymore?"

Anyway, as is usual at McC, Grant's sycophants circled the wagons to assert that, yes, Nate still has a huge hole down and in. A short discussion followed in which I said I would post some heatmaps to attempt to show that Nate has closed that hole.

Well, here they are.

 

[gp956]

 

[gp956]

What these maps show is the location of sliders that would have been "Balls" had Nate not swung at them to turn them into "Strikes". It's pretty obvious from these charts that there has been a major shift in the way Nate handles the down and in slider.

What is not readily evident in the maps is how few down and in sliders pitchers throw to Nate anymore. It's inconceivable to me that pitchers would simply abandon a hole if it's still getting positive results. And so, it's consistent that as Nate stopped helping the pitcher out, the pitchers started looking for other holes.

That's the way the game works.

 

[filosofy29]

What these maps show is the location of sliders that would have been "Balls" had Nate not swung at them to turn them into "Strikes". It's pretty obvious from these charts that there has been a major shift in the way Nate handles the down and in slider.

What is not readily evident in the maps is how few down and in sliders pitchers throw to Nate anymore. It's inconceivable to me that pitchers would simply abandon a hole if it's still getting positive results. And so, it's consistent that as Nate stopped helping the pitcher out, the pitchers started looking for other holes. That's the way the game works.

Yup, when I pitched (albeit not at a high level at all), it was common knowledge that if you found a weakness, you keep throwing the same pitch until the batter proves that he can hit it (and actually do some damage besides just fouling it off a couple of times). If this happens at the H.S. level, than it sure as hell should happen at the pro level where games are much more important and scouting reports are much more detailed.

 

[tzill]

What these maps show is the location of sliders that would have been "Balls" had Nate not swung at them to turn them into "Strikes". It's pretty obvious from these charts that there has been a major shift in the way Nate handles the down and in slider.

What is not readily evident in the maps is how few down and in sliders pitchers throw to Nate anymore. It's inconceivable to me that pitchers would simply abandon a hole if it's still getting positive results. And so, it's consistent that as Nate stopped helping the pitcher out, the pitchers started looking for other holes.

That's the way the game works.

Great work, GP. Aside: I'd like to post my PQS and QS% data that I've been tracking all season. Any suggestions as to format? Raw numbers aren't as interesting as heat maps and such. IIRC, you mentioned histographs before, any idea on how to make that happen from my data?

 

[gp956]

Great work, GP. Aside: I'd like to post my PQS and QS% data that I've been tracking all season. Any suggestions as to format? Raw numbers aren't as interesting as heat maps and such. IIRC, you mentioned histographs before, any idea on how to make that happen from my data?

I think I mentioned doing a histogram for PQS so that we get a feel for what a typical PQS distribution looks like, then we can calibrate ourselves towards unusually good/poor performances. Once you've calculated all the metrics, a histogram is pretty easy to accomplish in Excel. Just get all your PQS data in one column, determine some min/max values, and in another column create some evenly spaced intervals between your min/max values - these will serve as your "bins" to hold the PQS counts. Once you've done that a histogram is a matter of invoking one function.

Heatmaps are good for spacial data, not so good for pure counting stats. For instance, from the charts above it would appear that Nate simply shifted the holes to another location, and that is partly true, but what the maps don't show is that Nate simply turns Balls into Strikes, on sliders, at a lower rate than previously. I working on how to overcome that without blurring the main signal with too much extra information.

PS if you got your PQS data calculated you can post the raw data here. I can turn it into a histogram using R in just a couple clicks. I'd recommend using R for your future work, but the learning curve is steep.

 

[gp956]

Hotzone Comparison

These are wOBA weighted hotzones for Schierholtz. I'm still testing out these charts. Right now, the wOBA I'm using is really a pseudo-stat. Also the colors in the charts are relative values - that means colors don't corresponds to absolute values of wOBA.

Anyway thought I'd post the prototype charts to get some feedback:

 

[gp956]

Same charts, now side by side:

 

[gp956]

Just for comparison purposes - Joey Votto's Hotzone in 2010

 

[tzill]

Just for comparison purposes - Joey Votto's Hotzone in 2010

I'd reverse the colors....white for cold, yellow for tepid, orange for warm and red for hot.

 

[gp956]

I'd reverse the colors....white for cold, yellow for tepid, orange for warm and red for hot.

Not to be a pedant, but......the colors correspond to the expected spectral content of light emitted by an object as it rises in temperature. In the "real," physical, world, white is hotter than red.

 

[SFAnthem]

Extra credit if you can find Tejada's hot zone

 

[calsnowskier]

Extra credit if you can find Tejada's hot zone

I think is in 2002 Oakland...

 

[BeerMe]

Gp, be honest, are you Bill James?

 

[gp956]

Extra credit if you can find Tejada's hot zone

:typing: ........uh.......


Next is Huff's hotzone 2010 vs. 2011

 

[SFAnthem]

I think is in 2002 Oakland...

ha! You've been wearing out the rep button lately

 

[gp956]

Gp, be honest, are you Bill James?

Could Bill James ever be known as "The Hammer"?

 

[SFAnthem]

Gp, be honest, are you Bill James?

..and can you confirm or disprove the umpjack conspiracy vs. Giants pitching?

(homer alert: I think Cain was squeezed last night)

 

[tzill]

Not to be a pedant, but......the colors correspond to the expected spectral content of light emitted by an object as it rises in temperature. In the "real," physical, world, white is hotter than red.

Understood, hence "white hot." Nonetheless, I think it'd read better with red being super hot and white being cold. I don't suppose blue would work for cold, but maybe.

Just my 2 cents....remember: I'm only firing on an IQ of 99....

You fucking pedant. :lever:

 

[gp956]

Understood, hence "white hot." Nonetheless, I think it'd read better with red being super hot and white being cold. I don't suppose blue would work for cold, but maybe.

Just my 2 cents....remember: I'm only firing on an IQ of 99....

You fucking pedant. :lever:

I'll see how it looks. I'm also working on showing a scale of color, so a number can be associated with a color. At this point I'm just playing with what kind of structure is being revealed, and whether it jibes with what our eyes tell us. Also, I'm still not totally sold on the pseudo-wOBA metric I'm using.