FIP Challenge Results Part II

Earlier today, in Part I of the series, I published a chart of 34 pitchers who had a difference of 0.50 or greater between their FIP and xFIP at the All-Star break and their 2nd half ERAs. Here I want to go into more detail rather than just giving a raw score for the two metrics

In rating the two systems, I considered the metrics to recommend keeping a pitcher if at the All-Star break they were at 3.50 or lower, to listen to a trade if they were between 3.51 and 4.00, to actively look to sell the player if they were between 4.01 and 4.50 and to either sell or cut a pitcher if they were above 4.51.

Of course, we also have to consider what the pitcher’s actual ERA was at the break, too. A pitcher could still be a sell candidate if one of the metrics was significantly higher than his ERA. For these extreme cases, I considered a difference between 50-75 points to be a “listen” candidate, while above 75 to be a “sell high” guy.

Zack Greinke – His xFIP was 101 points higher than his ERA, making Greinke a sell high guy. This was a big win for FIP.

Joel Pineiro – After allowing just three home runs in 17 first half games, Pineiro served up eight home runs in 15 games after the break. This was a big win for xFIP.

Tim Lincecum – It was a very good second half of the season for Lincecum, just not as good as the first half. He did have a slightly higher HR/FB rate in the second half, and xFIP did a better job predicting his post-break ERA. Still, those fantasy owners who kept him based on his FIP did not end up disappointed.

Dallas Braden – Made just four starts after the break due to a foot infection. Officially a win for xFIP, but one we should probably dismiss due to lack of playing time.

Paul Maholm – His second half ERA was lower than his first half one, despite more HR allowed. Still, this was a pitcher that FIP would have identified as a potential buy candidate at the break, so a win for xFIP.

Tim Wakefield – Made just four starts in the second half due to leg and back injuries. Officially a win for xFIP, but one we should probably dismiss due to the lack of playing time.

Clayton Kershaw – He had a 5.0 HR/FB rate at the break and was even better in the second half, as he finished the year with a 4.1 mark. His ERA finished two full runs below what xFIP predicted. This was a big win for FIP.

Derek Lowe – Opponents posted an .888 OPS versus Lowe in the second half of the season, including 10 HR in 331 ABs. This was a big win for xFIP.

Cliff Lee – Everyone thinks the move to the NL turned things around for Lee but he was 3-0 with a 1.44 ERA in his first three games with Cleveland after the break. His HR/FB rate has been below 11 percent the past five seasons. This was a big win for FIP.

Carlos Zambrano – This was the closest one, as Zambrano’s second half ERA of 4.14 was just barely closer to his first-half FIP than his xFIP. Zambrano pitched worse in the second half than in the first, but it had nothing to do with his HR rate, which declined slightly from the 5.8 he posted in the first half. This was a slight win for FIP.

Jair Jurrjens – Both FIP and xFIP predicted Jurrjens’ ERA to rise in the second half and instead he pitched even better after the break. If you went strictly by FIP at the break, you would have listened to offers for Jurrjens. If you went by xFIP you were in the sell/cut area. This was a win for FIP.

Jeff Niemann – As with Jurrjens, both of our metrics predicted an ERA rise from Niemann in the second half. FIP had him as a sell while xFIP had him as a sell or cut guy. This was a slight win for FIP.

Nick Blackburn – Yet another pitcher that both metrics forecasted a rise in ERA. Except this time, the actual rise was more drastic than even the more pessimistic xFIP predicted. Since you might have kept him if you used FIP, this was a big win for xFIP.

Edwin Jackson – Pretty much the same thing as with Blackburn above, except you were even more likely to keep Jackson if you used FIP. This was a big win for xFIP.

Mike Pelfrey – The spread with our two metrics was not nearly as great with Pelfrey as it was for Blackburn and Jackson, but the end results were the same. This was a big win for xFIP.

Jon Garland – FIP projected Garland to be virtually the same in the second half as he was in the first half while xFIP had him being noticeably worse. The trade to Los Angeles invigorated Garland, or perhaps it was simply leaving a bad home park, as he finished the year with a 5.29 ERA at Chase Field and a 1.67 ERA at Dodger Stadium. This was a win for FIP, but probably not a pitcher anyone was targeting at the break.

Felix Hernandez – Again, both metrics predicted an ERA rise in the second half, although xFIP was more pessimistic, making him a sell high guy with a difference of 94 points. Hernandez pitched even better after the break, making this a big win for FIP.

Justin Verlander – Both metrics predicted an ERA drop in the second half for Verlander, with FIP being the most optimistic. Verlander pitched well, but saw his ERA go up, making this a win for xFIP.

Brian Bannister – A 3.66 ERA in the first half made Bannister look like a useful pitcher. Both metrics saw an ERA increase, but xFIP was the most pessimistic. This was a big win for xFIP.

C.C. Sabathia – Our two metrics were split on how Sabathia would fare in the second half. With a predicted decrease from his first half ERA, this was a big win for FIP.

Brad Penny – Our two metrics were split again. But Penny’s ERA went up in the second half. This was a win for xFIP, but not many people who used FIP were angling to acquire Penny.

Vicente Padilla – His HR/FB rate went up significantly in the second half, yet Padilla produced a lower ERA after the break, thanks to a move to the NL. Neither metric identified Padilla as a pitcher to target, although FIP came very close to hitting his actual mark.

Jarrod Washburn – Both metrics identified Washburn as a sell candidate as his ERA was 92 points lower than his FIP and 150 points lower than his xFIP. Officially a win for xFIP, although you likely would have made the same decision regardless of which metric you used.

Jered Weaver – A win for xFIP, which had him as a sell, while FIP had him as a listen. There are also extra points for xFIP for exactly predicting his second half ERA.

Joe Blanton – The metrics were split on how Blanton would fare in the second half. This was a big win for xFIP, which forecasted him to be a useful pitcher and he ended up better than that.

Bronson Arroyo – Technically a win for xFIP but not many fantasy players were running out to acquire Arroyo based on his 4.99 first half xFIP.

Jamie Moyer – Repeat the comment from Arroyo, except sub in 5.06 xFIP.

Trevor Cahill – Same as the above two, except with a 5.18 FIP.

Chris Volstad – Our two metrics were split on Volstad. FIP saw him continuing to be a sell/cut guy while xFIP saw him being a useful pitcher with a sub-4.00 ERA. This was a win for FIP.

Rick Porcello – Both systems predicted a rise in ERA but FIP elevated him to cut status. This was a win for xFIP.

Braden Looper – The two metrics were split on Looper, with xFIP predicting a drop in ERA. Looper actually pitched worse in the second half but neither system would have advocated acquiring him at the break.

Josh Geer – Made just three starts after the break due to lousy pitching. Not one that either system would have suggested to add.

Rich Harden – While most of the players with above average HR/FB rates have been of little or no value in regards to fantasy, Harden is the exception. Both systems saw him improving on his first half ERA but xFIP was much more bullish. And Harden exceeded those expectations. This was a big win for xFIP.

Randy Johnson – Appeared in just four games after the break due to a rotator cuff strain. Officially a win for FIP, but one we should probably dismiss due to lack of playing time.


If you made your fantasy decisions this year based on xFIP, you would be feeling very good about your choices with Pineiro, Lowe, Blackburn, Jackson, Pelfrey, Bannister, Blanton, Harden and to a lesser extent Maholm, Verlander, Weaver and Porcello.

If you made your fantasy decisions based this year based on FIP, you would be feeling very good about your choices with Greinke, Kershaw, Lee, Hernandez, Sabathia and to a lesser extent Jurrjens and Volstad.

From a pure bulk standpoint, you were better off in 2009 using xFIP at the break. But those who relied on FIP were more likely to make the right call on four of the five pitchers with the lowest ERA in the second half among the 34 pitchers in our sample.

We really cannot make any inferences for the future based on this one small sample. What we can say is that judging strictly from results in 2009 it would be a mistake to ignore FIP completely and absolutely while making fantasy decisions at the All-Star break. This year if you used xFIP you would have made the wrong decisions on some of the best pitchers in the game.

Print This Post

8 Responses to “FIP Challenge Results Part II”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. mymrbig says:

    Good stuff Brian!

    I think a fair generalization is that xFIP is a better indicator if the pitcher has a very high HR/FB. Out of the high HR/FB guys, xFIP was better for Blanton, Arroyo, Moyer, Cahill, Porcello, and Harden. FIP was better for Volstad, Looper, Geer, and Randy Johnson.

    But RJ and Geer are thrown out because of sample size. Volstad was just plain worse after the break, which neither could have predicted (his BB/9 ratio almost doubled and his K/9 ratio dropped, which has nothing to do with his HR/FB).

    So maybe it is fair to say that xFIP does a better job than FIP of identifying buy-low candidates, but doesn’t do any better than FIP for identifying sell-high candidates?

    Vote -1 Vote +1

  2. Keith says:

    Maybe it is just harder for hitters to get their fly balls out of the park against elite pitchers. Harden was an exception this year as he is pretty elite when healthy. Is it possible that elite pitchers should be assumed to have a lower HR/FB rate than everyone else?

    Vote -1 Vote +1

    • Brian Joura says:

      I don’t think we can make that assumption yet but it is something to watch going forward. Perhaps an elite pitcher already displaying a significantly reduced HR/FB rate in a given year is likely to maintain that for the remainder of the season.

      Vote -1 Vote +1

      • baluga says:

        Why is this something we have to “watch going forward?” We have the data for plenty of years going back… it is something that can be easily tested to see if it held true in the past.

        No need for these silly “experiments.”

        Vote -1 Vote +1

    • opisgod says:

      But why does Roy Halladay have a league average HR/FB? It seems as if all the pitchers with unsustainably low rates are pitching in massive outfields, and vice versa. Kershaw gave up almost no home runs at dodger stadium and the rest came on the road, it’s obvious why.

      Vote -1 Vote +1

  3. Brian Joura says:

    Given a large enough sample, almost all MLB quality pitchers will revert to an 11 percent HR/FB rate. But a year isn’t a large enough sample for all pitchers. You mentioned Halladay – in ’07 and ’08 he had HR/FB rates in single digits. This past year it was 10.6 percent.

    And you just can’t say it’s all ballpark. Busch Stadium was one of the toughest parks to hit a HR in this year (0.736 park factor for HR according to ESPN). Pineiro then allowed more HR in the second half than he did in the first half. Turner Field (0.861 HR factor) and Derek Lowe the same way. Comerica Park (0.974) and Edwin Jackson, too.

    Vote -1 Vote +1

  4. Pat says:

    Can’t alot of pitchers HR/FB rates be expected to increase in the second half just due to the weather?

    In april in the air is denser and the ball travels less, in the summer the air is less dense and it travels further… therefore it would seem like a good strategy in a points league to stack pitching in the first half (and then look to make moves around the allstar break to acquire hitters)

    You can sell high on the pitchers you think will regress to the teams that are starved for pitching because you monopolized it early on in the year, and you can buy low on the hitting….. never heard anyone mention this strategy before and im seriously considering publishing this somewhere, but it seems in a points league only where stats accumulated in april count in sept, it would be a good idea to play the general upswing in offense later in the summer

    Vote -1 Vote +1

  5. Jimbo says:

    Look at the number of top pitchers in each list. Verlander is about the only early round draft pick in that bunch, while Volstad is about the only non-early round draft pick in that bunch.

    The league I play in usually has a bit of “name-lag” where you can still get some guys at a value before they are proven studs. In my prep for next season, I might bump up a guy like Jurrjens if he’s really joining that teir of pitching.

    Vote -1 Vote +1