More on the LIPS/FIP discrepancy

On Monday, I examined two ERA estimators, FIP and LIPS, and discussed the pitchers who have been most undervalued or overvalued by FIP so far this year. For my discussion on the shortcomings of FIP, be sure to check out that article. For an explanation of LIPS, check out David Gassko’s primer from yesterday.

As a couple of readers pointed out, while almost all of the players on my list from Monday had abnormally high or low HR/FB rates (which is to be expected since FIP doesn’t normalize HR/FB), there were a few pitchers who saw a big disparity between FIP and LIPS yet had completely normal HR/FB rates. How could this be?

The answer is that while the biggest difference between LIPS and FIP is the fact that LIPS normalizes HR/FB, LIPS also takes into account a few other things that FIP does not. So today, I’m going to look at a few of the starting pitchers with at least 40 innings pitched and at least a 0.50 LIPS/FIP difference and examine why this difference exists.

2009 – SP with largest differences between LIPS and FIP

+-------------+------------+----+----+-------+------+------+----------+----------+-------+
| LAST        | FIRST      | G  | GS | IP    | ERA  | FIP  | LIPS ERA | LIPS-FIP | HR/FB |
+-------------+------------+----+----+-------+------+------+----------+----------+-------+
| Hammel      | Jason A    | 12 |  9 |  54.0 | 4.33 | 3.99 |     4.94 |     0.95 |    11 |
| Cook        | Aaron      | 13 | 13 |  76.0 | 4.26 | 4.70 |     5.41 |     0.71 |    15 |
| Halladay    | Roy        | 14 | 14 | 103.0 | 2.53 | 2.62 |     3.25 |     0.63 |    10 |
| Beckett     | Josh       | 12 | 12 |  76.3 | 3.77 | 3.53 |     4.14 |     0.61 |    10 |
| Buehrle     | Mark       | 12 | 12 |  80.7 | 3.24 | 4.15 |     4.76 |     0.61 |    11 |
| Cabrera     | Daniel A   |  9 |  8 |  40.0 | 5.85 | 6.36 |     6.95 |     0.59 |    10 |
| Hampton     | Mike       | 12 | 12 |  67.0 | 4.70 | 4.64 |     5.22 |     0.58 |    13 |
| Floyd       | Gavin C    | 13 | 13 |  82.0 | 4.94 | 3.79 |     4.29 |     0.50 |    10 |
+-------------+------------+----+----+-------+------+------+----------+----------+-------+
| Slowey      | Kevin      | 13 | 13 |  78.7 | 4.23 | 4.09 |     3.57 |    -0.52 |    11 |
| Young       | Chris      | 13 | 13 |  73.7 | 4.76 | 5.13 |     4.16 |    -0.97 |    11 |
+-------------+------------+----+----+-------+------+------+----------+----------+-------+

Note: For continuity’s sake, these numbers haven’t been update since Monday’s article.

Jason Hammel – Worse than FIP indicates

Hammel has the largest negative difference between FIP and LIPS so far in 2009, and I can see two primary reasons for this.

1) Hammel pitches for the Rockies and, therefore, in Coors Field. As Coors inflates run scoring by 9.3 percent, this will have a large impact on Hammel’s numbers that FIP simply ignores.
2) Hammel’s infield fly ball rate (2.7 percent) is lower than league average (3.9 percent).

Aaron Cook – Worse than FIP indicates

.
Cook is interesting in that his LIPS ERA is worse than his FIP, yet his FIP is being driven by an abnormal 15 percent HR/FB. If you were to normalize the HR/FB and apply the FIP formula, the difference would actually be even larger (his xFIP is 4.31, a 1.10 difference from LIPS).

Like Jason Hammel, being a Rockie has a lot to do with this. The run-scoring in Coors obviously has a big effect as these are the top two guys on the list. In addition, Cook’s infield fly ball rate is a measly 0.4 percent compared to a league average of 3.9 percent. As he allows a ton of balls in play to begin with, the effect is amplified.

Roy Halladay – Worse than FIP indicates

Halladay’s difference is being driven by the same two factors as Cook.

1) The Rogers Centre inflates run scoring by 3.1 percent.
2) His infield fly ball rate (2.7 percent) is lower than league average (3.9 percent).

Mark Buehrle and Gavin Floyd – Worse than FIP indicates

Both induce fewer infield flies than average and both call hitter-friendly U.S. Cellular home.

Chris Young – Better than FIP indicates

Chris Young has the most extreme LIPS/FIP difference of any pitcher in baseball this year, whether under or overvalued. Three factors are driving this:

1) PETCO reduces run scoring by 7.7 percent.
2) He induces more than twice as many infield flies (7.7 percent) than league average (3.9 percent).
3) He hits fewer batters (0.24 per 9) than league average (0.35 per 9)

Kevin Slowey – Better than FIP indicates

Slowey’s difference is being driven primarily by two factors.

1) He induces a lot of infield flies (6.5 percent), and because his strikeout ability is merely average-ish, the raw number of infield flies is pretty high.
2) His line drive rate (21 percent) is higher than league average (19.1 percent). Because he’s letting up too many line drives, he isn’t inducing as many groundballs, pop-ups, and fly balls as he should be, all of which do less damage than line drives.

Mental Health and the CBA
A particular bit of language in the latest CBA could have negative consequences for some players.

Print This Post
Sort by:   newest | oldest | most voted
Tim
Guest
Tim

Why compare it to FIP and not xFIP?

Derek Carty
Guest
Derek Carty

Tim,
Because Monday’s article used FIP in an attempt to show it’s flaws.  The biggest one I pointed out was that it doesn’t normalize HR/FB and is pretty useless for analyzing anyone (for fantasy purposes) who doesn’t have a league average HR/FB.  Today, however, I picked out a few of the players with a normal HR/FB who still had large discrepancies between their LIPS and FIP.  If you check the xFIPs of everyone except Cook, they should be pretty close to their FIP, so it doesn’t make much of a difference anyway.

bpasinko
Guest
bpasinko

How long until fielding/luck independent pitching becomes more mainstream?

You see BABIP around, but it’s often not followed with the real juice behind it, stats like xFIP, LIPS, DIPS, tRA or what have you.

Derek Carty
Guest
Derek Carty

You know, bpasinko, I have no idea.  We are starting to see more advanced stats going mainstream (I posted a BOTR post a week or so ago about Braves announcers talking about UZR, and D’Backs announcers apparently mentioned HR/FB a few days ago), although even as far as BABIP, it still isn’t always used properly, even by internet analysts (and I’ve yet to hear of it being mentioned on TV).  Why it’s being used more, though, I’d only be guessing.  Also, I think the answer would really depend upon what definition we’re using for “mainstream”.

Dave Studeman
Guest
Dave Studeman

When you quote infield fly rate, what are you referring to?  Your numbers are different from THT’s numbers.  It would be nice to have some continuity in how we reference stats.

Derek Carty
Guest
Derek Carty

Sorry, Dave.  It’s the percentage of batted balls that are infield flies, not the percentage of all flies that are of the infield variety.

bpasinko
Guest
bpasinko
I guess what I’m talking about is that there seems to be no reason why component ERA stas aren’t talked about in the same conversations that mention BABIP.  “Mainstream” is a broad definition (as I’m not even sure what I meant by it) but you see BABIP talked about a lot now in fantasy circles, even on ESPN fantasy regularly (although you are right, never mentioned on TV).  I’m just confused as to why something like BABIP is used (relatively often) but it’s often not accompanied by component ERA talk, which to my knowledge is really the entire point of… Read more »
John Burnson
Guest
John Burnson

I suspect that sportscasters don’t want to dip their toe into *expected* (or projected) stats. BABIP says what is, not what should be. “Actual” stats like BABIP also benefit because they have (essentially) one, uncontroversial definition, whereas with component ERA, there are a number to choose from.

That said, it wouldn’t surprise me to hear “FIP” one of these days….

Andrew
Guest
Andrew

Hi, Derek. I know you like DIPS WHIP as another indicator. Do you know if that stat can be found on any website? If not, can you let me know how I could get about calculating it? Always looking for an edge. Thanks.

John Burnson
Guest
John Burnson

Andrew, DIPS is now included in the pitcher pages in Heater—you can check out a recent issue here:

http://www.hardballtimes.com/main/downloads/heater_intro

As David showed in his article “Explaining LIPS,” the calculation is ~involved~, to say the least!

digglahhh
Guest
digglahhh
My opinion on why these stats aren’t used in a bit more cynical, unfortunately. Frankly, professional baseball, as an institution, is Luddite and anti-intellectual. There’s a perverse sense of pride taken in its own counter-intellectual mythology. All the talk about the grit, hustle, and toughness that pervades the mainstream discourse on the sport is a manifestation of the jock-geek dichotomy. Jocks resent geeks (As much as the geeks resented the jocks during our formative years, the shoe is on the other foot for most of our adult lives). Jocks always had sports as the one thing they knew better than… Read more »
Derek Carty
Guest
Derek Carty

Wow, digglahhhh, excellent post.

Also, I forgot to mention that Chris Young is absolutely awful at holding baserunners.  This is something that LIPS doesn’t take into account, so his LIPS ERA should be a little higher than it actually is (but still well below his FIP, which doesn’t take it into account either).

wpDiscuz