A few minutes ago, David Appelman announced the launch of several new stats here on the site, and since they hit on a topic of frequent discussion, I wanted to go into a bit more depth on our thought process behind their creation and what we see as their role in the evaluation of pitching.
Over the years, FanGraphs hasn’t been shy about promoting the concept of DIPS, which showed that most of the variance in a pitcher’s abilities can be viewed through the prism of walks, strikeouts, and home runs. We often cite a pitcher’s FIP — Fielding Independent Pitching, if you’re into proper names – when talking about his performance, and for most Major League pitchers, FIP works really well as an evaluator of their contribution to run prevention.
However, because FIP only focuses on walks, strikeouts, and home runs, it does not include all aspects of run prevention. Specifically, it takes no stance on two aspects of the game that do have a significant impact on a pitcher’s total number of runs allowed – the results of batted balls that are not home runs and the effects of sequencing of the various events. Because the spread in talent among Major League pitchers is not as large in these areas as the spread is in the components of FIP, ignoring these two areas doesn’t have a drastic result on the evaluation of most pitchers. However, there is certainly a subset of Major League pitchers who do accumulate (or fritter away) value through their performance in these two categories.
So, today, we’re introducing a set of metrics designed to help quantify the affects of run prevention that are not so easily isolated as the result of a pitcher’s actions. Because these metrics essentially serve to capture the value that FIP does not, we’re calling the sum of these metrics Fielding Dependent Pitching.
The idea for FDP was to quantify the remaining aspects of run prevention that are not measured by walks, strikeouts, and home runs. With a FIP-based WAR, we have a metric that tells us how many wins a pitcher added through success in those three key areas. What we did not have was a metric that gave us the wins added through either hit prevention or runner stranding. With FDP, we wanted to be able to break down the remaining aspects into those two categories, so that we could identify exactly where a team’s run prevention — with a specific pitcher on the mound — was coming from.
To do that, we simply worked backwards. First, we calculated the total WAR that a pitcher would receive credit for if he was only evaluated by his runs allowed, and we assumed that he had 100 percent responsibility for every variable that influenced run scoring. That stat is now on the site, and is called “RA9-Wins”. If you do not want to consider any impact of fielding on run prevention, and solely want to evaluate a pitcher by what actually happened when he was pitching (accounting for park and league adjustments, at least), then this is the metric for you,
However, for those of you who want to look at a pitcher’s contribution to run prevention in a more detailed way, we are also adding the two components of FDP to give you a better view of just how a pitcher is going about preventing runs.
To do this, we decided to calculate the linear weight value of singles and doubles — the difference between a double and a triple is almost certainly not due to something the pitcher did, and thus the value of advancing that extra base was not included — the same way that we calculate the value of walks, strikeouts, and home runs, so that we could quantify the wins added or lost that can be credited to a pitcher’s results on balls in play. Regardless of how you want to apportion the credit for those results, it is helpful to know what the value of those turning those hits into outs (or vice versa) actually is. Just as with WAR, these numbers are park adjusted and then converted into a number of wins added. These are on the site as BIP-wins, and can be thought of as the amount of wins a pitcher saved through his hit prevention, or lack thereof in many cases.
Once we knew the win value of a pitcher’s hit prevention, the remainder of his FDP could essentially be described as runner stranding. Now, this is not one particular skill, as there are many ways to skin this cat, but represents the value added through various skills that all essentially lead to the same result – not allowing baserunners to cross home plate. Some pitchers achieve this through effective control of the running game, picking off runners and refusing to let them advance through stolen bases. Other pitchers do this by simply altering the way they pitch with men on base, increasing the amount of only-semi-harmful walks they allow in order to reduce the amount of very-harmful home runs they allow. And still others simply seem to excel (or fail) at pitching out of the stretch relative to their peers, and demonstrating significant differences in their performance with the bases empty and with men on base.
No matter how they get there, however, the result can be measured by taking the remainder of a pitcher’s FDP that is not measured by his context-neutral hit prevention. This is called LOB-wins on the site, and serves as the value of wins added through all the miscellaneous ways a pitcher can strand runners.
BIP-wins and LOB-wins can be thought of as the components of Fielding Dependent Pitching, and represent the part of keeping runs off the board that aren’t measured by FIP. By definition, the sum of a pitcher’s FIP-wins and FDP-wins will equal his RA9-wins, so you can essentially see total run prevention through this basic formula:
RA9 = FIP + BIP + LOB
So, that’s the somewhat boring explanation part of the introduction. Now, let’s get to the fun stuff and actually play with the data.
In the last 10 years, here are the top five and bottom pitchers in total FDP.
These names are probably familiar to you if you’ve had any kind of discussion about the validity of FIP in the last few years. The four big names in the top five are the most often cited as pitchers who FIP underrates, and FDP shows just how large the gap is between their FIP-wins and their RA9-wins. Meanwhile, the guys on the bottom of the list are notorious underachievers, each of whom has been derided for failing to live up to their expected potential. As you can see, FDP returns the results you might expect if you were to look at the biggest FIP outliers of the last decade.
However, this is also an example of why breaking FDP down into BIP-wins and LOB-wins is useful, as we can present this same list, just showing where those wins added or lost came from.
Santana and Cain’s extra value has all come entirely through hit prevention. Santana’s stranded just about as many runners as you’d expect from a low-FIP/low-BABIP pitcher, while Cain has actually stranded fewer runners than you’d expect based on his context-neutral stats. Hudson and Weaver both accumulated value in both areas, but got the majority of their value through hit prevention, while Franklin actually got more of his value through runner stranding, though his RA9 is also significant impacted by hit prevention.
And now for the laggards:
Here, we see a more even split, with all five pitchers being negative in both areas. However, my suspicion is that being bad at both hit prevention and runner stranding is necessary to show up on an FDP leaderboard, because pitchers who truly awful at one or the other are likely weeded out before they ever make the Major Leagues, or at least spend significant time pitching for a big league club. That’s why the tails are higher at the positive end of the spectrum, as those pitchers success is keeping them in the big leagues longer and giving them more opportunities, while those who fail spectacularly at one of the two aspects of FDP simply don’t last long enough to show up on a list of most value lost over a ten year period.
Things get more fun if we look at even larger periods of time, however. If we expand the filters to cover the last 50 years, we find examples of guys where FDP tells quite an interesting story. For instance, Jim Palmer — with his career 2.86 ERA and 3.50 FIP — accumulated an incredible +27.8 wins through hit prevention and +15.5 wins through runner stranding. His +43.2 FDP-wins are, by far, the most of any pitcher in the modern era. Perhaps the most recent example of a similar type of pitcher is Tom Glavine, and he’s at +26.1 FDP-wins. Palmer is the biggest FIP outlier in the part of baseball history that at least resembles the game that is played today.
However, he’s not the leader in either BIP-wins or LOB-wins over the last 50 years. The pitcher who got the most value from hit prevention? Charlie Hough, which shouldn’t be surprising given the research that has been done on knuckleballers as the strongest exception to the DIPS theory.
Perhaps most interestingly, however, is the career of Nolan Ryan, who demonstrates how the two aspects of FDP don’t really go hand-in-hand in many cases. Ryan posted a career FIP- of 84 and a .265 BABIP, which should have resulted in dominating results that made him among the best run preventers in baseball. It didn’t, though, because Nolan Ryan was atrocious at stranding runners (relative to his own established levels, anyway), posting -30.2 LOB-wins over his career. Certainly, that is inflated to some degree simply through longevity, but there is no mistaking the fact that Ryan consistently posted higher ERAs than FIPs, even though his BABIP was also below average.
While Ryan is strange in the magnitude of his inability to prevent runners from scoring, he is representative of the lack of correlation between the two components of FDP. The correlation of BIP-wins and LOB-wins (scaled to IP, so as to account for differences in innings pitched) for all pitchers with at least 50 IP in 2012 is -0.008. In other words, there is no correlation. Of the top 10 pitchers in BIP-wins this year, only two (Jason Vargas and Kyle Lohse) also have positive LOB-wins. Clayton Kershaw is essentially acting as the current-day Nolan Ryan, and has been so ineffective stranding runners that it nearly cancels out all the value added through hit prevention, so that his FIP and ERA nearly match despite the fact that he has a .256 BABIP.
This lack of correlation holds up over longer periods of time, too, so this isn’t a sample size issue. The components of FDP are measuring two things that are quite different, and very few pitchers stray from the norm in both. In fact, those longer periods of time actually show just how effective FIP is as a measurement of pitching skill. In looking at all 3,951 pitchers who have thrown at least 100 innings in the majors since 1963, the correlation between the FIP-based WAR and RA9-wins is .96. For most pitchers with long careers, a WAR based on FIP and a WAR based on runs allowed is going to bring you to the same conclusion.
However, most pitchers is not all pitchers, and for pitchers like Jim Palmer — or, nowadays, pitchers like Jered Weaver, Tim Hudson, and Matt Cain — FDP helps us put a number on the mental adjustment we’ve been making to help compensate for the fact that FIP does not measure a part of run prevention that they have contributed to in a meaningful way.
Through adding FDP-wins (and its components, BIP-wins and LOB-wins) and RA-9 wins to the site, we hope that we’re now presenting a more comprehensive picture of how runs are saved when a pitcher is on the mound. It is still definitively true that runs are mostly saved by limiting walks and home runs and keeping batters from making contact, but FDP fills in the gap between FIP and runs allowed, and gives us a clearer picture of the impact of various performances in the things that aren’t captured in FIP.
In a subsequent post — this one is already over 2,100 words, straining the limits of the word introduction — that will be up in a few hours, we’ll discuss the role that FDP will play in our pitcher WAR and how we hope the addition of these new metrics will help us better reflect the value that different types of pitchers produce. I’m also hosting my regularly scheduled noon chat today, and will make it FDP-centric, answering as many questions about these stats as you guys are interested in asking.
For now, though, we hope you enjoy the new tools that are now available, and enjoy perusing the leaderboards and learning new and interesting things that you may not have known before, like how Nolan Ryan was a really good pitcher, but could have been so much better had he performed well with men on base.