Pitching stat(s)

[TooLiveBrew]

I'm wondering if a pitching stat exists that incorporates batted ball data like flyballs, line drives, etc. I tend to prefer to look at xFIP & K/BB rates when evaluating pitchers, but lately I've been thinking that it would seem useful to incorporate at least a guy's LD% allowed. Basically, my vague idea of a stat would include just about everything from FanGraphs' "Batted Ball" segment of a player page -- GB/FB, LD%, GB%, FB%, IFFB%, HR/FB, IFH%, BUH%. Is there a way for rates like those to be included with something like xFIP to make a more informative descriptive stat?

[TheCrew07]

I know where you're going, I'm just not sure you can get there, nor am I aware of a metric that takes into account batted ball data.

 

No matter how pitching is evaluated and what rates are factored into a metric you're still going to end up having to explain my least favorite concept in modern sabermetrics, luck. I firmly believe players and teams make their own luck, most events which are attributed to luck in baseball aren't luck at all, but rather a complex system of variables which is incredibly difficult to quantify yielding a result that appears to be random on the surface. I'll use Manny Parra as an example, other than his BB rate just about every metric you'll find says he should be a better pitcher than he is, so he gets labeled as unlucky. However when you watch the games his mistakes tend to get hit very hard, on the ground or otherwise. I think he's predictable in that hitters believe he'll fall behind in the count so they wait for a FB that they can handle and then make solid contact with the pitch. How do you account for that mental chess game between pitcher and batter in a metric?

 

I'm sure someone will spend the time to figure out a way to incorporate pitch f/x data into a metric which would be better than what's out there now, but then we're still losing the context of the situation and the intended vs the actual location. There's such a fine line between being good in an outing and being bad (6-10) pitches, that it's very difficult to find a smoking gun anywhere good or bad, in the end it's just easiest to measure the result.

[cwolf2012]

"but rather a complex system of variables which is incredibly difficult to quantify yielding a result that appears to be random on the surface"

 

Not to cherry pick your post too much, but this is essentially luck. I agree that the situations had specific variables that are absolutely unquantifiable or predictable that led to the specific (if inevitable) outcome that is based 100% on those specific circumstances. It's called determinism in philosophy.

 

Instead of trying to act like those variables can be controlled or manipulated, it's much much easier to refer to that as "luck."

[brewjihad]

Your first post basically describes tRA. It's on the player pages here, and here's the main explanation.

[TooLiveBrew]

Thanks brewjihad, I'm off to check that out

[logan82]

There are outliers though like Manny Parra specifically how carry a higher BABIP despite what his GB% and LD% show.

[rluzinski]

"No matter how pitching is evaluated and what rates are factored into a metric you're still going to end up having to explain my least favorite concept in modern sabermetrics, luck."

 

When sabermetric talks about luck, they are referring to statistical uncertainty. Statistical uncertainty is not a "concept of modern sabermetrics". It is an indisputable fact of probability statistics. Everything we observe in baseball is the sum of the signal (true talent) and noise (luck). Anyone who claims the signal or noise in any situation is 0% or 100% is probably wrong.

 

If I flip a coin 10 times and get 10 heads, you can tell me I made my own luck in doing so but it's just another sports cliche, IMO. Some portion of the outcome is dependant on skill and some on luck. Anyone who watches baseball has to realize luck is a HUGE factor.

[pete]

If luck is used to refer to uncertainty, that's pretty damning, because those are two different things.

[rluzinski]

If luck is used to refer to uncertainty, that's pretty damning, because those are two different things.

What the hell IS luck then?

[TheCrew07]

Luck has become a convenient fall back option for statisticians. Instead of digging harder to quantify all of the variables, or admit that there isn't enough data to make a definitive conclusion, we end up debating the merits of statistical noise and luck, which is garbage in my opinion. It's become a convenient excuse used when production doesn't come close to the approximation projected by whatever metric people are trying to use.

 

Take my least favorite metric ever... BABIP, and one of my favorite pitchers, Manny Parra. For 2 years now BABIP supporters have been riding the "he's just unlucky" card when if you watch the games it's pretty clear Manny left way too many pitches up and over the plate, and they were getting crushed. Instead of giving the hitters credit for doing exactly what they are supposed to do in crushing a meatball, excuses about luck are made to explain an extraordinarily high BABIP. If MLB hitters cannot turn on a mid thigh to waist high FB... when then they wouldn't be MLB hitters to begin with because everyone can hit that pitch, except maybe Ben Sheets. Even though I really like Manny's talent, he's stunk, but the difference this season is just obvious as his failings were in the past. This season he's been mostly working down in the zone with much better results. He still misses up and he still gets hit hard, but any pitcher that misses up is going to get hit hard.

 

I'm not saying Parra is fixed yet, but what I am saying is that I don't believe he was unlucky for a season and a half and now he's suddenly lucky which is leading to better results.

 

Basically, luck is garbage, the end.

[cwolf2012]

"we end up debating the merits of statistical noise and luck, which is garbage in my opinion. It's become a convenient excuse used when production doesn't come close to the approximation projected by whatever metric people are trying to use."

 

Do you think "luck" or "statistical noise" or "variance" plays any part in the game at all? Are you completely deterministic? The projections are based upon regressing towards the mean with higher variance statistics and looking at past data, nothing more.

 

"For 2 years now BABIP supporters have been riding the "he's just unlucky" card when if you watch the games it's pretty clear Manny left way too many pitches up and over the plate, and they were getting crushed."

 

This is completely subjective and open to confirmation bias. Some pitchers get away with the high stuff, but when they do, you don't bat an eyelash. But if Manny gets hit, since you seem to have a bias against him, you make a mental note and remember it. If he gets away with it though, I bet you say he was lucky...

 

"Instead of giving the hitters credit for doing exactly what they are supposed to do in crushing a meatball, excuses about luck are made to explain an extraordinarily high BABIP. If MLB hitters cannot turn on a mid thigh to waist high FB... when then they wouldn't be MLB hitters to begin with because everyone can hit that pitch, except maybe Ben Sheets."

 

It doesn't matter how bad the pitch is, if your swing is literally a half an inch too high or too low, it's an easy out. That's precision that fails well with the bounds of human error. That's why you see pitchers get away with stuff all the time. Manny was absolutely giving better pitches to hit than he should have, so he likely deserves a higher BABIP, but his astronomical levels were simply too high to be sustained. That is, unless you think Manny Parra is the worst pitcher in the league.

 

"This season he's been mostly working down in the zone with much better results. He still misses up and he still gets hit hard, but any pitcher that misses up is going to get hit hard."

 

Once again, confirmation bias. I would love for you to bring me some pitch f/x data to back this up, not anecdotal evidence. His LD% is almost the same as last year (18.1% in 09 to 17.6% this year), his GB% is almost the same (48.1% to 49.3%). His LOB% is almost 10% better, and he's giving up slightly fewer home runs. His BABIP is still .362. He's performing almost EXACTLY the same as last year, except his ERA and his results are better.

 

How else can you explain it except he's getting the better end of variance?

 

"I don't believe he was unlucky for a season and a half and now he's suddenly lucky which is leading to better results.

 

Basically, luck is garbage, the end. "

 

His peripherals disagree with you. I know you're a smart guy, so I strongly suggest you take this in and really think about it. LOB%, HR/FB%, and BABIP are 3 stats that have a lot of variance. There are some guys who have the ability to supress BABIP and HR/FB%, and those guys are mostly heavy GB pitchers with good defense behind them (Webb and Jimenez come to mind). There are some guys, like Parra, who are worse with BABIP due to bad defense and bad pitching. That doesn't make him 2 standard deviations from the mean worse than everyone else in a high variance stat.

 

Manny Parra is almost exactly the same pitcher as last year, and the results are better. If that's not luck (or whatever you want to call it), I don't know what is.

Edit: Fixed some formatting and grammar.

[rluzinski]

I'm not saying Parra is fixed yet, but what I am saying is that I don't believe he was unlucky for a season and a half and now he's suddenly lucky which is leading to better results.

 

There's nothing wrong with your assertion that it wasn't luck. It might actually be more likely that Parra's true BABIP (or SLg against or whatever you want to use to measure effectiveness) trully is well above average. The thing about statistical uncertainty is that you just can't PROVE it conclusively, using sample based statistics. A true .320 BABIP pitcher could pretty easily have a .360 BABIP for a season and half ONLY because the coinflips went against him. Or it could be because his location stunk.

 

Now, if you KNOW that Parra is consistently thowing his pitches in a location that on average, causes a worse than average results, that's altogether different. More than a couple of studies have verified the pretty obvious connection between location and pitch effectiveness. Here's one:

 

http://baseballanalysts.c...9/07/can_pitchers_co.php

 

Like cwolf, I am quiet skeptical that you have accurately kept track of Parra's pitch locations, though. I don't know how anyone could, just by watching games. Here is Parra's pitch f/x data by starts:

 

http://www.fangraphs.com/...010&date=2010-06-18&dh=0

 

Even having all the data in front of me at one time, it's hard to make any objective observations. I would need to dump all the raw data intoa spreadsheet and do a study. But apparently, you've apparently stored all that in your head and came to a definitive conclusion? I think you are WAY overestimating your observational abilities. Thousands of pitches! Thousands of results! The notion that "all you have to do is watch him pitch and you know" HAS to be just starting with the results and working your way backwards. That or you are Rainman.

 

Also, I think it's a lot more complicated than just looking at pitch location, anyway. Pitch velocity, break, location, sequence, deception, etc... I don't think anyone has claimed to be able to understand everything that goes into understanding the art of pitching.

 

"Basically, luck is garbage, the end."

 

To apparently completely reject the whole concept of luck in sports is pretty silly. It's like rejecting that that water is wet and the sky is blue. I think you just don't find the non-deterministic nature of sports to be emotionally satisfying so you try to dismiss it. If you flip heads, it's because you were supposed to flip heads?

[pete]

If luck is used to refer to uncertainty, that's pretty damning, because those are two different things.

What the hell IS luck then?

 

Uncertainty is not knowing something. Luck usually refers to a positive random event.

[sbrylski]

If luck is used to refer to uncertainty, that's pretty damning, because those are two different things.

What the hell IS luck then?

 

Uncertainty is not knowing something. Luck usually refers to a positive random event.

That would be good luck. Luck itself simply refers to unanticipated events or circumstances that operate for or against an individual.

[rluzinski]

We are talking about the luck of the coin flip, which evens out over the long haul. A batter has a certain percent chance of getting a hit against a particular pitcher. Unless you think that it's 0% or 100%, you've implicitly accepted the luck aspect in baseball.