Fantasy Baseball

xFIP: Better Lucky than Good

Sometimes a player can get off to a bad start and their ERA makes them look worse than they probably are.  Other times a player can get off to a hot start and while the ERA says they are pitching well, the underlying metrics say something else.  This is one of the reasons I love xFIP when evaluating pitchers.  xFIP estimates what a pitcher should be doing based upon the raw numbers as opposed to what their surface stats show.  It gives you an idea if a pitcher has been plagued by bad luck and due to rebound or serve as a warning that they have been lucky so far and are doomed to regress. I’m a negative person by nature so today I will be looking at pitchers whose ERA’s are much better than they should be (according to their xFIP) and will likely see a huge decline going forward.   

Nick Martinez of the Rangers is my first regression candidate.  Currently his ERA stands at 2.65 and I’m sure those of you who picked him up early in the year have enjoyed the numbers thus far.  He has 9 quality starts this year (out of 12) and has allowed 2 earned runs or less in all but one of those 9 starts.  Unfortunately Martinez has an xFIP of 4.81; that is 2.16 higher than his actual ERA.  There is a lot of play between what he has done and what he should be doing. 

Martinez has increased his slider use to over 20%.  This appears to have helped decrease the number of fly balls (down 14% from last season), but only on the surface.  The FB% and LD% has been inching up the past 6 weeks while the GB% has been trending downward.  Same goes for the HR/FB%; he gave up zero home runs in April but has a HR/FB ratio of 10.5 and 9.1 respectively for May and June.  The WHIP is on the high side at 1.30 and has also increased each month starting with a 1.12 in April to a 1.36 for May and now a 1.61 in June.  The BABIP is middle of the road at .282, but that is due to a .250 for April; it has been above .300 since.  The BAA is also trending in the wrong direction; combine that with the increase in fly balls, HR/FB%, BABIP, WHIP along with the low strikeout rate (5.43) and you’ve got the makings of a perfect storm. 

Last season Martinez posted a 4.55 ERA and 1.46 WHIP over 140 innings.  While some improvements were anticipated this year, I don’t think anybody expected this.  With the weather getting warmer the balls will start flying a little more (for everyone, not just Martinez).  ZIPS and Streamer call for an ERA close to 5.0 for the remainder of the season.  If you are a Martinez owner I would be looking to trade now before the downward spiral is in effect. 

Shelby Miller has also been pitching far better than expected.  A 1.84 ERA puts him at the top of the food chain for starting pitchers and is almost 2 full points lower than last season.  Miller has allowed more than 2 earned runs in a game once in 12 starts; twice actually counting Saturdays game which I’ll mention again later since stats were compiled before that start.  Throw in a WHIP under 1.0 and things look great on the surface, but when you look at the 3.93 xFIP and league-leading .230 BABIP you can’t help but wonder…and worry. 

Some of his success is due to his pitch selection, cutting his four-seam usage in half and basically tripling the use of his cutter (5.8% to 19.4%) and two-seam fastball (10.3 to 34.2%).  The change has resulted in a 10% increase in ground balls (49.4%) and 9% decrease in fly balls (32.4%).  The lower FB% has also helped with his HR/FB% which is half of what it was last year.  The LD% is also low (deceptively so) at 18.3%.  While all these things are positives and point to future success, the numbers being produced are unsustainable.  The changes haven’t helped his paltry strikeout rate (6.48 K/9) which is slightly up from last season.  Without the K’s (which went MIA last season) he will need some luck and help on defense to maintain his current line.  Unfortunately, as good as the Braves are defensively, some of those ground balls are gonna start getting through.  Additionally, the low LD% is fueled by a 12.6% rate in May so a few more hits will get through, and a few hits can make all the difference.   

Saturday’s game against the Mets gave owners some insight of what to expect of Miller going forward.  Miller went 6.1 innings allowing 8 hits and 2 walks.  Those few extra hits where enough to raise his WHIP from .98 to 1.03.  The extra hits resulted in an extra run which raised his ERA from 1.85 to 2.02.  Miller has had two average/below par games this month and the results are a .317 BABIP for the month of June.  Before the Mets game Miller’s xFIP was 3.93, but that has also crept up to 4.02.  I don’t see Miller collapsing the way I do Martinez above, but I do see games similar to the recent one against the Mets happening more often.  Miller has a FIP of 3.34 which is close to what ZIPS and Streamer are calling for over the remainder of the season.  That still makes him a serviceable starter in fantasy, just not the elite option we saw over the first few months.  If you can sell high on Miller I would, but he is not someone I would rush to get rid of.

Speaking of rushing to get rid of someone, Mike Pelfrey exemplifies such a player.  Prior to his start on Saturday at Texas, Pelfrey had an ERA of 2.28 along with a 1.22 WHIP over 67 innings.  Of his 11 starts (excluding the Rangers game), Pelfrey allowed 2 earned runs or less in 9 starts and 7 of those were one run or less.  This is the same player with a career ERA close to 4.5 who hasn’t produced an ERA under his career average since 2010. 

Just like the first few players the surface stats paint a rosy picture, but an xFIP of 4.36 (which is 2.14 higher than his ERA) says something else.  Pelfrey has a 55% ground ball percentage; that is between 9 and 10 percent higher than what he has posted over the past 5 years.  The FB% is just below 25 which is 11 to 13% lower than it has been the past two years with the Twins.  Other than the increase in his split finger fastball, not much has changed when it comes to pitch selection to account for the additional grounders or decrease in fly balls.  Even with the decreased fly balls the HR/FB ratio has not changed, but Pelfrey has a decent track record of limiting long balls.  The velocity on his pitches remains the same as well with the exception of a 4 MPH drop in his split fingered fastball.

The two biggest issues owners should be concerned with are hits and strikeouts.  Pelfrey has a H/9 of 9.42 which is lower than his career average (10.05) but still high.  2010 was the last year he had a H/9 this low; it has been over 10 since then.  Next there is the (lack of) strikeouts.  This season his K/9 sits at 4.33, the lowest total for any full year he has played.  The lack of K’s and high hit total are a recipe for disaster because like I said earlier, some of those ground balls will start getting through.  We saw that a few days ago in Texas as Pelfrey surrendered 11 hits and 8 earned runs in 3.2 innings.  That game raised his ERA almost a full point up to 3.18 and his WHIP now stands at 1.36 (which is the lowest it has been since 2008).  The xFIP also increased up to 4.50.  Nothing has changed about Pelfrey; only the results have been different which means the end results are inevitable.  Projections give him an ERA between 4.64 and 5.11.  Even if we take the lowest estimate, that makes Pelfrey a borderline streaming option from here on out.  Given his track record, there is little chance of trading Pelfrey so be thankful for what he has given you so far.  The party will be over soon. 

Hector Santiago is next up on the regression train with a 2.59 ERA, 1.18 WHIP and 4.65 xFIP – a 2.06 differential.  He only has 7 quality starts in 13 tries, but he could have 11 quality starts if he had made it through 6 innings in 4 of those 5 non-quality starts.  Santiago has a few red flags right off the bat.  Both his FB% (52.1) and HR/FB% (8.9) are up from last season.  Last year he gave up 15 home runs in 127 innings and this year he is up to 10 in only 76.  Fortunately half of those home runs came with nobody on base which limited the damage.  Santiago lowered his H/9 over a full point from last season (7.31) and lowered his BB/9 for the third straight season (3.30).  The strikeouts are also up from last season (8.14) and are close to what we saw in Chicago before he was traded.

Just like with Shelby Miller, a portion of his success is due to pitch selection.  Santiago has basically abandoned his four-seam fastball (used 26.7% last season) in favor of an increased use of sinkers (53.1% up from 34.1%) and cutters (15% up from 8.9%).  Both the fastball and sinker are the same velocity, but the extra movement on the pitch has resulted in a decrease in contact and increase in SwStr% – again, both equal to what he did in Chicago in 2013. 

I bring up 2013 for a reason; Santiago’s xFIP that year was 4.65 (same as this year) and he ended the year with a 3.65 ERA.  Actually Santiago has done this his entire ML career with a 4.57 xFIP last season (3.75 ERA) and a 4.40 xFIP in 2012 (3.33 ERA).  Sometimes pitchers can play beyond what their numbers should be; while a decline from his current line is in order, the fall will not be great.

ZIPS and Streamer call for an ERA just above 4.0, but given the improvements he has made in walks, strikeouts, hits allowed and BAA (.220), there is a chance of that number being lower.  It will all come down to how much damage opposing teams can do to him via the home run.  This month Santiago has a FB% of 63.6% and a HR/FB ratio of 14.3%.  In half the innings pitched he has already passed his earned run total from April and is 2 shy of May.  Overall Santiago should give you more good games than bad, but you’ve gotten the best out of him already so you may want to move on.  Look to trade Santiago now; if you keep him, just be prepared for the occasional blowup. 

Speaking of blowing up, Aaron Harang is due (or already there – read on).  Prior to last nights game at Baltimore, Harang was sporting a 3.04 ERA and a 1.10 WHIP.  This is just the second time in his career with a WHIP under 1.2 and he has never finished a season with an ERA below 3.52.  That doesn’t mean that he can’t; look at what Edinson Volquez did in 2014 after years of mediocrity (putting it mildly).  Harang’s xFIP is 4.49 (a 1.45 differential).  Like Santiago above, Harang has had a few years that he has outperformed his xFIP, but only a few occasions and never consistently. 

Both his walks and strikeouts are down; while he has had years with low K’s and walks, he has never had a year where both were low.  The H/9 (7.48) is the lowest it has been in his career (9.37) and the last time Harang had a H/9 below 9.0 and a low BB/9 was back in 2007 for Cincinnati.  His FB% is up 6% from last season (47.1%) but the HR/FB ratio is only 7%.  The HR/FB ratio has been under his career mark of 10% in 4 of the past 5 seasons (including this year), but the additional fly balls and hitter friendly home park lead me to believe that could change (3 home runs – 1.98 ERA at home). 

The velocity remains the same on all of his pitchers and other than a slight decrease in curveballs almost equal to the increase in changeups, the pitch selection hasn’t changed much.  The plate discipline hasn’t changed much either.  Batters are swinging a little more and making a little more contact, but no more than some of the past years.  So he isn’t throwing any differently, there is no change in velocity and yet he is giving up fewer walks, fewer hits and has a lower home run total. 

How do we explain this?  Luck, plain and simple – and that luck is starting to run out looking at his June stats.  This month the BB/9 is above 5.0, HR/9 is 3.05 and the H/9 is above 9.0.  He has allowed 16 earned runs and 5 home runs over his past 3 games, and last night in Baltimore he allowed another homer and 4 earned runs over 6 innings.  It’s safe to say the wheels have already come off this bus.  If you are a Harang owner, dump him now.  The 3.24 ERA and 1.10 WHIP is sure to attract the attention of someone in your league.  Let them foolishly pick up Harang; better he ruin their ratios than yours.  Zips and Streamer have him with an ERA in the low 4’s for the remainder of the season.  Than means a few more blow-ups, a good number of “mehh” games and the occasional good start.  Nothing you can’t get from a streaming option.


Here are the remaining players whose ERA and FIP differ by 1.0 or more.  Some of the obvious players (Greinke) you don’t have to worry about while others (Haren) could be trouble down the line.

Player  ERA xFIP Differential Player  ERA xFIP Differential
Sonny Gray  1.60 3.23  1.63 Dan Haren 3.22 4.37 1.15
Alfredo Simon 2.58 4.06 1.48 Tim Lincecum 3.33 4.37 1.14
Zack Greinke 1.95 3.32 1.37  Anthony Desclafani 3.36 4.50 1.14
Michael Wacha 2.45 3.78 1.33 Jesse Chavez 2.64 3.73 1.09
Jake Odorizzi 2.47 3.65 1.18 Scott Kazmir 2.79 3.80
Chase Anderson 2.82 4.00 1.18 Aaron Sanchez 3.55 4.55 1.00
David Price
Kyle Gibson 3.24 4.14 0.90

Having an xFIP higher than the ERA does not necessarily mean regression, and regression is not always a bad thing depending on the player.  This is something to consider though when evaluating players on your team for the remainder of the season or potential trade targets.  Next week I will take a look at the other end of the spectrum; players whose ERA outweigh their xFIP.  Is it just a case of bad luck or are they really bad.

Fantasy Rundown Banner

Need more pitcher analysis, Streamers, prospect profiles, DFS picks and general fantasy baseball goodness, head on over to Fantasy Rundown.

By Jim Finch

The self proclaimed Grand High Exhausted Mystic Ruler of Fantasy Baseball. While I am not related to Jennie or Sidd Finch, I will attempt to uphold the integrity of the Finch family name as it relates to baseball.