First and foremost, I would like to say that I am thrilled to be a part of Disciples of Uecker after a couple of great years with JSOnline as Bernie’s Crew. I feel extremely fortunate to work with a staff of great writers, and I am happy that everyone puts up with my throwing spaghetti on the wall.

This post here, this is more spaghetti on the wall. But, enough with the formalities; you all are probably hungry, and if this stuff sticks, it’s time to eat!

*As always, stats notes are indented, so you can just skip ’em if you only want the logic and analysis. I’m not trying to force stats down anyone’s throat. *

FIELDING INDEPENDENT PITCHING

Last year I jumped aboard the Fielding Independent Pitching bandwagon, and I immediately found the concept useful for a number of reasons. First and foremost, I suspect that so many baseball fans like FIP because it resonates with the idea that pitchers frequently cannot control the outcome of batted balls in play, and therefore pitchers don’t always receive the fate that they deserve. Really, for ages baseball fans have been arguing over which pitchers deserved what and who received the best support or who was unlucky. FIP simply updates the debate in that regard.

Furthermore,I found FIP to be useful on a team level, as a common sense, intuitive indicator of which teams might be efficient defensively, and which teams might not. One of the aspects of FIP on which few focus is the ability of FIP to serve as an index. FIP isn’t truly a stat; rather, FIP analyzes the relationship between existing stats (namely, K, HR, BB, etc.) in terms of their relationship to runs allowed. As an index, FIP can be split into two parts, and here’s where we can play around with defensive efficiency.

FIP AS AN INDEX, NOT A STAT

As an index, FIP expresses the relationship between elements of the game that pitchers can best control (such as HR, K, and BB, as well as HBP) on a scale that corresponds to ERA or runs average. FIP can be split into two parts by (1) focusing purely on the ratio between HR, K, and BB, etc., and (2) separating the index that places that ratio into the language of runs average or ERA.

For instance, in 2011 NL, the runs average was 4.17; the ratio between K, BB, and HR for the sake of calculating FIP was 0.73. In order to place that ratio of fielding independent elements into the language of runs average, we can add the number 3.44 to 0.73. Voila! 4.17!

*Stats Note: For the purpose of splitting FIP, I call the ratio between the fielding independent elements “FIPRatio,” and the index number to place FIP on a runs average or ERA scale “FIPIndex.” Therefore:*

*Runs Average: 4.17 / FIPRatio: 0.73 / FIPIndex 3.44
*

Splitting FIP is a useful way to analyze team performance because we can look at a team’s particular ratio between K, HR, and BB, and then determine the number of runs they might allow with an average defense playing behind the pitchers. This relationship is shakier on an individual level because defensive efficiency is not evenly distributed within each particular team, for a variety of reasons (just ask Zack Greinke, for instance). But, on a team level, we can isolate the ratio of strike outs, walks, and home runs to determine the effectiveness of the pitchers outside of their defense, and we can then determine how the defense might have performed, given their chances in the field.

For these purposes, here’s how the 2011 NL compared, in terms of basic Defensive Efficiency, FIPRatio, and FIPIndex:

This is a good start to the basic division of FIP into two elements, on a league scale for 2011 NL.

PARK FACTOR FIP: Pure Speculation

Throughout 2011, I thought about how to further analyze these FIP splits to adequately analyze each team’s interplay between pitchers and fielders. I asked myself specifically, “how can FIP itself be placed into park context?” I focused on the problem this way, because I figured that each team plays in a park context that influences K, BB, and HR in particular ways, and therefore defensive efficiency would be influenced according to each team’s context for pitching independent elements.

To develop specific park factors for FIP, I worked out a couple steps of reasoning:

(1) Park Factors for FIP should focus SOLELY on FIPRatio. This will allow the adjusted stat to express the relationship between K, BB, and HR as influenced by each park.

(2) Park factors for FIP should be calculated according to the park-adjusted runs average. This will allow the relationship between fielding independent elements to be expressed in a run environment that makes sense for those elements.

Once I figured out this basic reasoning, I was able to influence the FIP equations accordingly. For this, I used Bill James’ 2012 Handbook, which features three-year park factors from 2009-2011 for everything from K, BB, and HR to infield error rate, etc. I also used *Baseball-Reference*, in order to double check runs environments.

*Stats Note: I understand that this proprietary information, but the numbers posted below are my own calculations and should be no reflection of what Bill James or Baseball Info Solutions think about Fielding Independent Pitching; you see, I may be dead wrong and completely misusing these stats, and you should hold that against me, not them. If any of this should be pulled as a proprietary violation, please let me know.*

*Calculations for this type of index should be simple. Taking ((13HR+3BB-2K) / (IP)) to be the most basic formula of FIP, we can adjust the formula depending on a park’s specific impact on K, BB, and HR. So, if a park limits home runs to 80% of league average while increasing strike outs and walks 10% above league average, we might use the following formula to express FIP in that park: [((13*.8)lgHR + (3*1.1)lgBB – (2*1.1)lgK) / (lgIP)]. By adjusting the FIPratio in that way, we can potentially express that park’s specific relationship between (and influence on) fielding independent elements.*

Basically, the idea is simple: if we can figure out how each park influences K, BB, and HR, we can figure out how each park influences the interplay between fielding independent elements and defensive efficiency. Therefore, we can adjust our equations accordingly, and not only judge each pitcher according to park-adjusted runs average, but according to the very ratio of K, BB, and HR at that park.

Here’s an example of how that might look in 2011 NL:

CONCLUSION

While reflecting on this data, I believe there are a lot of places for improvement and tests for consistency, etc. Here are some basic conclusions:

(1) Some parks influence HR, K, and BB to such an extent that each team’s defense will either be required to be extremely efficient, or less efficient. This obviously depends on whether or not each park favors more batted balls in play (or less batted balls in play).

(2) FIPIndex, the number used to place fielding independent elements on a scale that reads like runs average, should not be thought of as a replica of defensive efficiency. Rather, that number can be used to suggest the relationship between (1) the extent to which the defense was required (due to a particular staff’s K/BB/HR tendencies, and (2) the extent to which the defense was actually good or bad.

So, we don’t have an actual number here that corresponds to defensive efficiency; rather this aspect of FIP corresponds to two aspects of fielding.

(3) There are park factors that may shift over time that might further influence how we consider this type of exercise.

There you have it! At the very least, I believe that we can use park-adjusted FIPratios to judge pitchers that jump from one park to another (Mat Latos’s jump from San Diego to Cincinnati will be particularly interesting, not only because PetCo Park limits home runs, but also because PetCo park increases strike outs and walks. Therefore, Latos will be pitching in an environment that could influence his fielding independent performance in a lot of extreme ways).

Furthermore, we can use these types of exercises to think about the importance of defensive efficiency in certain parks. Take Miller Park, for instance, which features a strange interaction of elements that encourages more home runs without favoring overall increases in runs scored. Miller Park is an odd breed of park, a rather average runs environment over time that happens to have strange effects on fielding independent elements. For this reason, one might ask, if the Brewers’ pitching staff does an extremely good job limiting damage in terms of their fielding independent elements, is defensive efficiency less important for the Brewers than other clubs? (I am not sure, but I believe this might explain how 2011 worked for the Brewers).

The relationship between batted balls in play and elements of the game that can limit batted balls in play defines the very basis of baseball. So, I hope that we can continue to work on these types of relationships to analyze the value of ballplayers.

SOURCES:

Baseball-Reference. Sports-Reference, LLC., 2000-2012. Accessed February 9, 2012.

James, Bill. The Bill James Handbook 2012. Chicago: Acta, 2011.

Sources intended for educational purposes.

There is so, so much I don’t know about advanced stats. Thanks for helping enlighten just a little bit.

One thing I do know a bit about, though, is law, and I’d say your post falls pretty squarely under fair use for educational and research purposes.

[…] Prevention: Randy Wolf is a Better Pitcher than Zack Greinke 2013 Pitching Rankings: Fluctuation Fielding By Fielding Independent Pitching (2011) Through 2012 Season: […]