Pitchers -- xFIP Leaders through May 25

liam4  5/26/2014 2:16 PM

Thought this was interesting.  Here are the season's pitching leaders thus far, ranked by xFIP.  I can put up an explanation of xFIP later on, but find it to be a great way to pull luck out of ERA and WHIP.  Doesn't correspond to z-scores of course and it has its own imperfections, but definitely points out for me who should continue to be do well (like Tanaka, Keuchel, Cueto, Kluber) and who's pitching better than their numbers (Price). Surprised to see guys like McCarthy and Kennedy in there.

Btw, Kershaw would be #1 with 2.01 xFIP, but doesn't have the minimum innings yet in this ranking.

gehrenberg   5/26/2014 3:01 PM

FIP and XFip are a decent starting place to look at pitchers, however if you want to get real indications of their performance these should be used as guide lines for how pitchers have performed thus far. It is a misconception that are a good way to determine future performance. These stats were created to determine a players past performance. Although I understand what is it that you are saying these stats need to be looked into more to really determine how we think pitchers will perform going forward. If pitchers peripheral stats are improved is it because they have improved or have they benefitted from other factors such as small samples. FIP and XFIP are very reliant on strikeouts, walks, and HR/FB rate (as they should they are the most important aspects for pitchers). A pitcher such as Dallas Kuechel has definitely improved and it is evident by him throwing more strikes and generating more swings and misses. However, Cueto has seen his swinging strike rate fall from 11.0 to 9.6, which is a farily significant drop off, however he has been striking out much more hitters this year than last. So I would project Cueto to have his strikeouts fall off going forward and for him to finish the year with an FIP around 3.5 or so. For me these are super fun discussion points I love to see them brought up in the forums. Please more people post opinions for discussion. 

gehrenberg   5/26/2014 3:05 PM

Also of note is the Kuechel has completely changed how he pitches. His pitch F/x shows he has done away with his cutter and starting using his slider more which appears to be the biggest reason for his success this year. 

liam4  5/26/2014 7:34 PM

I agree and disagree with you I think, though it might just be semantics.

It's definitely geared towards judging past performance, and it does not take the place of detailed analysis in considering future performance (as with your note on Cueto). BB/9 and K/9 do fluctuate as well.  However, I think it is directional and helps get a quick read on whether a pitcher's success has been a fluke or the real thing, and I think looking at the rankings is really interesting information.  At a macro level, all we can do when projecting future performance is look at historicals.  So although Keuchel being #3 certainly doesn't make me think he'll be #3 going forward, it makes me think he will stick, and he caught my eye a couple weeks back that his xFIP was so good and made me dig deeper. 

On Keuchel specifically, agree that his revised repertoire is a major contributor this year.  It seems to me that his performance will turn on whether he can maintain his BB/9.  He was below 2 BB/9 in the minors, but in the majors in 2012/2013 he was quite a bit higher.  If he keeps walks down, he could be a real ace... on a bad team, but top 15/20 pitcher.  His numbers last year also suffered from .340 BABIP, which was very fluky.

liam4  5/27/2014 12:37 PM

Following up with a quick intro to xFIP, as it's one of the dorkier metrics...

First, FIP measures what a player's ERA should have looked like, assuming that performance on balls in play (BABIP) is average.  Research found that BABIP does not correlate well across seasons and much of variance can be attributed to luck.  So it looks at what a pitcher can control:  strikeouts, walks, HBP, home runs.  Formula for those who care:  FIP = (13*HR)+(3*(BB+HBP))-(2*K))/IP + constant.

xFIP then takes FIP and replaces a pitcher's home run total with an estimate of how many home runs they should have allowed, using the league-average home run to flyball rate.

There are definitely limitations here, as K/9 and control can fluctuate, and some pitchers may be less homer prone than average.  But as a blunt object, I think it's an awesome view on how a pitcher's doing and who to target, who to sell.

gehrenberg   5/27/2014 1:15 PM

Also liam check out SIERA it's fangraphs 'upgraded' version of Fip and has been shown to be the most accurate when comparing to actual ERA. 

liam4  5/27/2014 3:18 PM

Thanks, yeah, I like SIERA, but I haven't found a place to pull rankings on it?  I'll spend some time looking

liam4  5/27/2014 8:19 PM

Found it.  Here's the top 50 pitchers by SIERA.  Rather similar but some corrections.