This season you will see countless Fantasy Baseball articles comparing a Pitcher's xFIP or FIP to their ERA. They're going to tell you who has been "lucky" or "unlucky", they're going to tell you who is going to regress or bounce back. The majority of them will be wrong because FIP and xFIP are flawed statistics. I'm a fan of Socrates' question everything philosophy and I have one question which brings down the entire ERA forecasting formula, how are you going to calculate ERA without using Base Hits, Team Defense or Ball Park Factors? Make the jump to see exactly what I am talking about and why I'm able project ERA much better than FIP or xFIP.
FIP and xFIP are calculated using four statistics: Home Runs Allowed, Walks, Strike Outs and Innings Pitched. They ignore Base Hits, Team Defense and Ball Park Factors. xFIP differs from FIP in that it tries to correct extreme numbers in HR Allowed. xFIP has the right idea, but does not follow through. Instead of correcting HR Allowed to the Pitcher's career average, they use the league average. Huge mistake. If you think Clayton Kershaw or Matt Cain will post a 10%-11% HR/FB%, you're crazy. There are plenty of Pitcher's who consistently outperform the league average for HR/FB%.
The Flawed Logic
The idea that Pitchers do not have the ability to limit Base Hits is absurd. Sure, there are fluctuations in BABIP from year-to-year, but that doesn't mean they're unpredictable. Here is a great read by Tom Tippett that explains how Pitchers are able to limit Base Hits on Balls In Play. It's a lengthy read, but it's worth the time if you're interested in the subject. He ends it perfectly with, "Using power or control or deception or a knuckleball, pitchers can keep hitters off balance and induce more than their share of routine grounders, popups, and lazy fly balls."
Batting Average of Balls In Play is directly influenced by a Pitcher's skill, the defense behind him and the ballpark he calls home. Team defenses can play differently for some Pitchers. A Ground Ball Pitcher like Joel Pineiro will benefit more from a good Infield and a Fly Ball Pitcher like Ted Lilly will benefit more from a good Outfield. Balancing player trends, free agency moves and ball park changes is the key in projecting BABIP.
My ERA Forecasting System
I'd like to call my method mdsFIP since I'm essentially using an altered version of FIP, but I use a projected BABIP. FIP stands for "Fielding Independent Pitching" and BABIP includes Ball Park Factors and Team Defense so it wouldn't make much sense for me to use a FIP title. I'd also like to give my method an original name, but I didn't spend weeks calculating numbers like Voros McCracken or Tom Tango. I simply improved their methods. I stood on the shoulders of giants. People can call my projected ERA system whatever they want, as long as they recognize it as the best.
What I do is calculate a Pitcher's FIP using my projected numbers for Strike Outs, Walks, Home Runs Allowed and Innings Pitched. Then I use my projected BABIP and figure out how many Base Hits were prevented or allowed compared to the league average BABIP (.293). These Base Hits are then equated into Earned Runs. Now I am able to use the original FIP to calculate the number of Earned Runs Allowed, add or subtract the ER from BABIP and recalculate the ERA.
FIP and xFIP are for pure baseball fans. My ERA projection system is for fantasy baseball fans.
If you don't believe me about BABIP influencing ERA, take a look at Ted Lilly. He allows a league average HR/FB%, but his BABIP is 20 points below the league average and he always outproduces his xFIP or FIP. The same story plays out with Jeremy Guthrie or any other Pitcher that is able to prevent Base Hits on Balls In Play. If you still don't believe me, I'll give you 5-1 odds that my projected ERA numbers judo chop xFIP this season.
Clayton Kershaw is able to prevent Base Hits and Home Runs. I have a projected BABIP of .280 and a projected HR/FB% of 5.5% for him. His ERA will be 2.79 using these numbers. If we were to ignore BABIP and use the league average for HR/FB%, his xFIP would be 3.66.
James Shields is not able to prevent Base Hits and Home Runs. I have a projected BABIP of .307 and a projected HR/FB% of 12.0% for him. His ERA will be 4.31 using these numbers. If we were to ignore BAIP and use the league average for HR/FB%, his xFIP would be 3.94.
More BABIP-FIP-ERA related material
I totally agree with breaking down a pitchers BABIP through their contact percentages. It all falls back to the "stuff" concept I threw out before. Being able to see trends in a pitchers "stuff" like MDS pointed out with Cahill is huge. If this downward movement trend continues his HR totals will continue to decline like they did last year from 27 to 19. FIP doesn’t account for this, which is why I'm interested to see how the mdsFIP v. FIP projections contest works out.
Cahill had more downward movement on his pitches last season, I believe in the uptick in GB%. I don't remember the exact LD/GB percentages I projected, but I did use a 29.8% FB% and downgraded the BABIP to .277. The better-than-average BABIP lowered his FIP from 3.87 to 3.56.
Whenever I look at stats like BABIP and FIP these days, I always take a glance at other rate stats like LD/GB/FB/IFFB rates.
Example: I could say that Trevor Cahill got very lucky last season due to an incredibly low .236 BABIP. His FIP was 4.19, furthur indicating that he was beyond lucky. However, consider the balls in play he gave up. An very low 15% LD% and a very high 56% GB%. Then, consider that the A's had the 5th best team UZR in 2010 and their infield defense was awesome last season...
Look, I like FIP as a basic eyeball formula, but there are many other factors that need to be considered as well.
Trevor Cahill did a few things exceedingly well and he has the team behind him to maximize his strenghts.
There is no arguing that. But, my point is that FIP fails to take into the account a pitchers “stuff” by merely writing it off as their ability to strikeout batters.
A strikeout is the best possible outcome for a Pitcher. Nothing tops strikeouts. They are very important.
Overall, I think FIP does uses the most relevant stats when evaluating pitchers. Pitchers can be broken down from their control (walks), ability to keep the ball down (home runs allowed), but somewhat mistakenly interprets a pitchers “stuff” by measuring strikeouts. I think the most misinterpreted ability is a pitchers “stuff”. When people here “stuff” they immediately think strikeouts which isn’t necessarily true. “Stuff” in my view is how difficult it is to get hits of a pitcher once the ball is put into play. For example, I believe Arroyo has better stuff than Roy Halladay, yet Halladay is much more successful for other reasons. His control is nearly unmatched, doesn’t give up a lot of homeruns, he doesn’t allow a normal amount of his pitches to be put in play due to all his strikeouts, and has incredible stamina at his age to pitch deep into games. But, when compared with Arroyo, Halladay gives up more hits when hitters put the ball in play. In finishing I think FIP puts too much stock in a pitchers ability to strikeout batters and not enough into a pitchers ability to limit hits when their balls get put in play via their “stuff”.