If two pitchers had the same gb/fb ratio and the same number of innings pitched (or maybe batters faced would be better), would it be expected that they give up the same amount of HRs?  Assuming you control for park effects and have a large sample size.

Is gb/fb ratio determine HR allowed rate in and of itself, or is there more to it?

I think that certain pitchers can change what happens to their flyballs, which is really what matters here. Tim Wakefield, with his knuckleball, has a 8.9% HR/flyball percentage, which is pretty low.

To better address your question about gb/fb predicting HRs, we can look at the first pitcher I thought of, Johan Santana in his 2004 and 2005 seasons, when everything would presumably be very similar in terms of ballparks, opponents, etc.

2004: 40.6 GB% | 43.0 FB% | 16.4 LD% ==> 15.0% IF/Fly | 12.7% HR/Fly ==> 1.06 HR/G
2005: 40.4 GB% | 43.1 FB% | 16.5 LD% ==> 18.8% IF/Fly | 10.9% HR/Fly ==> 0.93 HR/G

So, despite his GB/FB/LD numbers staying pretty similar, he managed to keep more fly balls in the infield and keep them from leaving the field. It resulted in less home runs per game—over the course of 10 games, about 1 HR was “gained,” which for someone in a Cy Young chase, is actually a lot.

Of course, this is a terribly small sample size, and may change from pitcher to pitcher. I know that Hardball Times tracks these numbers, and the numbers I quoted were from HT (Santana’s page is at http://www.hardballtimes.com/main/stats/players/index.php?playerId=755&firstName=Johan&lastName=Santana) but you may find FB/GB HR/G-adjusted numbers in a database or something, and you can run a query or two there. You may want to talk to Studes at HT who “has the keys” to the “Baseball Info Solutions” data, which tracks batted ball types for him. He may not be able to just give you numbers, but perhaps he’d be interested in writing a column on this topic. I think it’s a very interesting question.

For what it’s worth, I looked closer at this question here: http://garthsears.blogspot.com/2006/07/can-groundballflyball-rates-predict.html

For 2006, pitchers with 80+ IP showed a 16.3% correlation between GB/FB and HR/9. Without adjusting for home parks, etc, I think that one could comfortably say that GB/FB cannot predict HRs allowed. They have a lot to do with each other, but they aren’t nearly close enough to show a strong correlation.

Interesting.  Thanks for taking the time out to look into that.

A few questions.

First, I’d have to wonder how much park factors play a part in those numbers.  Not sure how easy that would be to factor out of the numbers.  I would think they could potentially skew the data.

Secondly, would it be better to look at HR per batter faced?  Just looking at HR/9 kind of gives the numbers a team defense skew doesn’t it?  Good defenses get more outs and therefore the batters faced in the inning would be less.  Also it probably would favor good pitchers over bad ones (good pitchers get out of the inning quicker).

Thirdly, how much randomness is in this number?  BIPA seems to have a lot of randomness.  I think “Baseball Between the Numbers” had something like 40% is randomness.  I’m not even sure how to figure out that number.

Does anyone keep track of gb/fb ratio compared to actual hits instead of outs?  Maybe I’m wrong but I think the espn.com numbers are tracked on outs only.  Which would introduce some more defensive skew.

After these long hours at work end, I’m going to have to sit down and figure out how I can run some of these numbers myself.

Oh yeah, I had one more question.

Does anyone know the formula for DIPS?  I saw one with hits included, which doesn’t make much sense to me, because hits given up is _not_ defense independent.

I guess I’m more curious if they take into consideration HR’s allowed, or GB/FB ratio.

Big Mafia - 02 July 2006 04:46 PM

Oh yeah, I had one more question.

Does anyone know the formula for DIPS?  I saw one with hits included, which doesn’t make much sense to me, because hits given up is _not_ defense independent.

I guess I’m more curious if they take into consideration HR’s allowed, or GB/FB ratio.

Do you mean dERA (McCracken’s measure) or something like FIP (by Tom Tango)? dERA (I think it’s listed as DIPS at espn.com) is very difficult to calculate. There was a page formerly hosted at McCracken’s homepage, but doesn’t exist anymore. I found it at Wikipedia, however (I can’t verify whether it’s exactly the same, but it looks like it is):

http://en.wikipedia.org/wiki/Defense-Independent_ERA

Keep in mind, that’s Version 2.0.

FIP is much easier to calculate, and works very well:

(HR*13+(BB+HBP)*3-K*2)/IP

What I don’t like about that formula is that it uses innings pitched, clearly not defense-independent.

Big Mafia - 02 July 2006 04:33 PM

Interesting.  Thanks for taking the time out to look into that.

A few questions.

First, I’d have to wonder how much park factors play a part in those numbers.  Not sure how easy that would be to factor out of the numbers.  I would think they could potentially skew the data.

Secondly, would it be better to look at HR per batter faced?  Just looking at HR/9 kind of gives the numbers a team defense skew doesn’t it?  Good defenses get more outs and therefore the batters faced in the inning would be less.  Also it probably would favor good pitchers over bad ones (good pitchers get out of the inning quicker).

Thirdly, how much randomness is in this number?  BIPA seems to have a lot of randomness.  I think “Baseball Between the Numbers” had something like 40% is randomness.  I’m not even sure how to figure out that number.

Does anyone keep track of gb/fb ratio compared to actual hits instead of outs?  Maybe I’m wrong but I think the espn.com numbers are tracked on outs only.  Which would introduce some more defensive skew.


After these long hours at work end, I’m going to have to sit down and figure out how I can run some of these numbers myself.

Park Factors—I think that with 91 pitchers (averaging just very slightly over 3 pitchers per home park), and how everybody is on the road 50% of the time, that the park factors are virtually cancelled out.

HR/batter faced—Yes, this would be better. Again, with 91 pitchers involved, I think any outliers are negated here. However, it’s a silly complication that could be avoided easily.

Randomness—I know that line drives become hits 75% of the time, but I don’t know the rates for FB and GB, other than FBs are incredibly unsuccessful for the batter. FBs are either home runs, outs, or the 1-in-500 error. I have no idea about the GB hit/no hit ratio. Over a smaller sample size, this would wreak havoc over the regression, and even 91 is getting close to skew-vulnerable, but again, I think the large sample size helps. The 80+ innings also would, in theory, limit the luck involved from pitcher to pitcher, let alone over the whole thing.

However, to more closely determine GB/FB predicting HR/batter on a pitcher-to-pitcher basis, these problems need to be addressed. The PFs are pretty easily adressed, HR/batter faced should be easy enough to deal with (just make it actual batters faced rather than 27), and randomness can be addressed with larger sample sizes. Unfortunately, I haven’t attended any college math classes yet (I just finished graduating high school, where I didn’t pay attention to calculus), so I can’t say what simple size would be needed to effectively rule out luck.

---

Does Hardball Times’ GB/FB/LD stuff go on hits or outs or both? I would guess both, but I could be wrong.

How did you figure out the correlation coefficient?

I had Mic. Excel do it on the graph. While I can run a linear regression through a TI-82 or Excel, I am sadly still one of those who cannot do one by hand. Maybe college math’ll learn me good.

The answer to the original question, which is a good one, is that no, two pitchers with the same G?F ratio will not necessarily have the same true HR per batted ball rate.  Another way to answer that is that, “Pitchers do not all have the same true HR per FB rate, which would be the case, if the answer to your question is in the affirmative.

If we look at two groups of pitchers with different HR per fly ball rates in one year (one low and the other high) and then we look at the next year and adjust for any park effects differences between the two groups, we find that the “trend” continues; that the high HR per fly ball group will continue to be high (it will regress toward the mean of course) and that the low HR per fly ball will continue (and regress) as well.

In fact, contrary to popular belief in some sabermetric circles, a pitcher’s HR rate per batted ball is much more predictive of his future HR rate per batted ball than his fly ball rate (if that makes any sense).  IOW, it is much more important what “types” (long, short, etc.) of fly balls a pitcher gives up than the number of fly balls, in terms of predicting his HR rate (and controlling those “types” of fly balls is definitely a skill, at least in terms of whether they become HR or not)) it is.  That is one reason why it is not that big a deal for a fly ball pitcher to pitch in a high HR park (like the old Coors Field).  That was NOT the reason why Milton was so unsuccessful his first year in CIN.

So, if thats the case, is G/F ratio a valuable stat?  Does it show the ability to limit BIPA, just not HRs?  Or is there a correlation between G/F and HR rate, but its just not a correlation of 1 (as my original question probably suggested)?

Or, is the real value the ability to have a higher WHIP translate into a lower amount of runs scored (which seemed to be the case with Silva last year, although thats just an anectdotal situation)?

Thanks for the input mgl, I plan on getting the book soon.  Can’t wait.