Stat SPaz 9/20/09: FIP vs. ERA, pt. 1
The Starting Line
by Evan "the Censor" Dickens
evan@fantasybaseballsearch.com
First a note: The absolute most pure analysis of seasonal SP stats would require me to wait until the postseason--but I'm nothing if not anxious. So let's start dissecting 2009!
ERA is a standard fantasy statistic, and for good reason--it is theoretically the correct measurement of any pitcher's ability to limit runs that don't result from fielding errors. Of course, there are plenty of holes, such as the fact that inadequate (but not erroneous) fielding can play a big part in a pitcher's ERA suffering, as well as the fact that bad relievers can make a mess of a starter's ERA by welcoming home inherited runners.
As a result, some saber-heads have developed a metric called Fielding Independent Pitching (FIP) which discards all statistics that a pitcher does not have much control over. I won't break down the entire detail of the equation--you know at this point whether you give any credence to this stat spaz silliness, so you're either on board or you're not--but think of it as a pure distillation of a pitcher's skill peripherals, and a discarding of those pesky luck stats. Ultimately FIP serves as an ERA predictor, before the effects of luck.
Many pitchers have demonstrated the accuracy of FIP. Cliff Lee's 3.00 ERA corresponds tidily with his 3.02 FIP. James Shields' disappointing 4.09 ERA was sadly predicted by his 4.06 FIP. Tim Lincecum's exquisite 2.30 ERA is predicted, and actually surpassed, by his 2.22 FIP (which is why he's so far and away the #1 pitcher next year).
Not all pitchers have such correlation, and that's where you can get some useful information about who to target--and who to avoid--next year. In this, the first of a two-part series, I will look at the unlucky pitchers: the ones with an ERA significantly higher than their FIP. But just in case you don't think there's any use in even looking at this, let's see what you would have known if you looked at pitchers with a high (ERA-FIP) in the last two years. For the record, I'm discarding starters who are fantasy-irrelevant (after all, the difference between a 5.80 FIP and a 4.70 ERA doesn't do much for anyone's team).
Thanks as always to the amazing fangraphs.com, my source for this data.
2007
Jeremy Bonderman - 5.01 ERA vs. 4.19 FIP
Greg Maddux - 4.14 ERA vs. 3.58 FIP
Joe Blanton - 3.95 ERA vs. 3.50 FIP
Wandy Rodriguez - 4.58 ERA vs. 4.18 FIP
Bonderman dropped his ERA nearly a point in 2008 before injury scrapped his season. Maddux was overcome with age as his strikeouts plummeted in 2008, and Blanton was beset by a scary (likely injury-driven) loss of control, though he proved to be quite a prize in 2009. The real story here was Rodriguez who was a lot better in every key statistic in a shortened 2008, and then got really good in 2009.
2008
Kevin Millwood - 5.07 ERA vs. 4.02 FIP
Javier Vazquez - 4.67 ERA vs. 3.74 FIP
Andy Pettitte - 4.54 ERA vs. 3.71 FIP
Josh Beckett - 4.03 ERA vs. 3.24 FIP
Justin Verlander - 4.84 ERA vs. 4.18 FIP
See any of these names that wouldn't adequately described as a "value" pick if you were lucky enough to take them? Although Millwood has trailed off in late season, likely due to some fatigue, he was a much more solid and rosterable option for three months. Pettitte has been a nice surprise, probably as a waiver pickup, as his BABIP has normalized from last year's .339. I still hate Beckett personally but boy he was good for most of 2009. The real prizes here, though, are Verlander and Vazquez, who are both Top 10 pitchers for 2010 that were clearly undervalued at draft time. You would have wanted to take both of these guys at least two rounds higher than they were taken in your league--trust me.
Now let's see what we can learn from this year...
2009
Ricky Nolasco - 5.34 ERA vs. 3.50 FIP
Carl Pavano - 4.82 ERA vs. 3.96 FIP
Derek Lowe - 4.53 ERA vs. 3.78 FIP
Paul Maholm - 4.51 ERA vs. 3.81 FIP
Jason Hammel - 4.35 ERA vs. 3.66 FIP
Let me be clear about this: there is absolutely nothing that Nolasco has done in 2009 that makes me doubt for an instant that he is one of the 20 best starting pitchers in the majors. His K/9 has actually increased to a career-high 9.09 and his walk rate is still nice and low at 2.17. He is simply the unluckiest pitcher in years. If you doubt it, then good, join my league and take him off your draft board. If you believe in statistics, regression, and skill peripherals, you need to draft Nolasco next year. The value you're likely to get for him could be equivalent to the value you got from Chris Carpenter this year. If you don't believe in those things, you probably stopped reading a long time ago.
The other four could provide some value at the back of the draft next year. Pavano's high BABIP and very low BB/9 bodes well for a good fifth starter; he posted a career high K/BB in 2009 (and it's been a long career). Lowe is what he is, a consistent pitcher with a solid ground-ball IP-eating skillset. Maholm is actually very similar to Lowe; a solid GB pitcher with low HR rates and a really good curveball. Hammel is the real dark horse; now that his walks are under control he just keeps getting better and his curveball is devastating. He could prove to be a very solid part of the young Rockies rotation next year and should definitely not go undrafted (and should not be unowned in your league if you're still in contention).
Not all of these guys are guaranteed game-changers for 2010, but they are definitely pitchers that you don't want to underrate at draft time--the odds that they will be significantly better next year are very good, and with Nolasco, I guarantee it.
Stay tuned, at some point soon we'll take a look at the lucky ones that are deserving of multiple grains of salt...see you then!
~Evan the Censor
by Evan "the Censor" Dickens
evan@fantasybaseballsearch.com
First a note: The absolute most pure analysis of seasonal SP stats would require me to wait until the postseason--but I'm nothing if not anxious. So let's start dissecting 2009!
ERA is a standard fantasy statistic, and for good reason--it is theoretically the correct measurement of any pitcher's ability to limit runs that don't result from fielding errors. Of course, there are plenty of holes, such as the fact that inadequate (but not erroneous) fielding can play a big part in a pitcher's ERA suffering, as well as the fact that bad relievers can make a mess of a starter's ERA by welcoming home inherited runners.
As a result, some saber-heads have developed a metric called Fielding Independent Pitching (FIP) which discards all statistics that a pitcher does not have much control over. I won't break down the entire detail of the equation--you know at this point whether you give any credence to this stat spaz silliness, so you're either on board or you're not--but think of it as a pure distillation of a pitcher's skill peripherals, and a discarding of those pesky luck stats. Ultimately FIP serves as an ERA predictor, before the effects of luck.
Many pitchers have demonstrated the accuracy of FIP. Cliff Lee's 3.00 ERA corresponds tidily with his 3.02 FIP. James Shields' disappointing 4.09 ERA was sadly predicted by his 4.06 FIP. Tim Lincecum's exquisite 2.30 ERA is predicted, and actually surpassed, by his 2.22 FIP (which is why he's so far and away the #1 pitcher next year).
Not all pitchers have such correlation, and that's where you can get some useful information about who to target--and who to avoid--next year. In this, the first of a two-part series, I will look at the unlucky pitchers: the ones with an ERA significantly higher than their FIP. But just in case you don't think there's any use in even looking at this, let's see what you would have known if you looked at pitchers with a high (ERA-FIP) in the last two years. For the record, I'm discarding starters who are fantasy-irrelevant (after all, the difference between a 5.80 FIP and a 4.70 ERA doesn't do much for anyone's team).
Thanks as always to the amazing fangraphs.com, my source for this data.
2007
Jeremy Bonderman - 5.01 ERA vs. 4.19 FIP
Greg Maddux - 4.14 ERA vs. 3.58 FIP
Joe Blanton - 3.95 ERA vs. 3.50 FIP
Wandy Rodriguez - 4.58 ERA vs. 4.18 FIP
Bonderman dropped his ERA nearly a point in 2008 before injury scrapped his season. Maddux was overcome with age as his strikeouts plummeted in 2008, and Blanton was beset by a scary (likely injury-driven) loss of control, though he proved to be quite a prize in 2009. The real story here was Rodriguez who was a lot better in every key statistic in a shortened 2008, and then got really good in 2009.
2008
Kevin Millwood - 5.07 ERA vs. 4.02 FIP
Javier Vazquez - 4.67 ERA vs. 3.74 FIP
Andy Pettitte - 4.54 ERA vs. 3.71 FIP
Josh Beckett - 4.03 ERA vs. 3.24 FIP
Justin Verlander - 4.84 ERA vs. 4.18 FIP
See any of these names that wouldn't adequately described as a "value" pick if you were lucky enough to take them? Although Millwood has trailed off in late season, likely due to some fatigue, he was a much more solid and rosterable option for three months. Pettitte has been a nice surprise, probably as a waiver pickup, as his BABIP has normalized from last year's .339. I still hate Beckett personally but boy he was good for most of 2009. The real prizes here, though, are Verlander and Vazquez, who are both Top 10 pitchers for 2010 that were clearly undervalued at draft time. You would have wanted to take both of these guys at least two rounds higher than they were taken in your league--trust me.
Now let's see what we can learn from this year...
2009
Ricky Nolasco - 5.34 ERA vs. 3.50 FIP
Carl Pavano - 4.82 ERA vs. 3.96 FIP
Derek Lowe - 4.53 ERA vs. 3.78 FIP
Paul Maholm - 4.51 ERA vs. 3.81 FIP
Jason Hammel - 4.35 ERA vs. 3.66 FIP
Let me be clear about this: there is absolutely nothing that Nolasco has done in 2009 that makes me doubt for an instant that he is one of the 20 best starting pitchers in the majors. His K/9 has actually increased to a career-high 9.09 and his walk rate is still nice and low at 2.17. He is simply the unluckiest pitcher in years. If you doubt it, then good, join my league and take him off your draft board. If you believe in statistics, regression, and skill peripherals, you need to draft Nolasco next year. The value you're likely to get for him could be equivalent to the value you got from Chris Carpenter this year. If you don't believe in those things, you probably stopped reading a long time ago.
The other four could provide some value at the back of the draft next year. Pavano's high BABIP and very low BB/9 bodes well for a good fifth starter; he posted a career high K/BB in 2009 (and it's been a long career). Lowe is what he is, a consistent pitcher with a solid ground-ball IP-eating skillset. Maholm is actually very similar to Lowe; a solid GB pitcher with low HR rates and a really good curveball. Hammel is the real dark horse; now that his walks are under control he just keeps getting better and his curveball is devastating. He could prove to be a very solid part of the young Rockies rotation next year and should definitely not go undrafted (and should not be unowned in your league if you're still in contention).
Not all of these guys are guaranteed game-changers for 2010, but they are definitely pitchers that you don't want to underrate at draft time--the odds that they will be significantly better next year are very good, and with Nolasco, I guarantee it.
Stay tuned, at some point soon we'll take a look at the lucky ones that are deserving of multiple grains of salt...see you then!
~Evan the Censor
Labels: 2010, FIP, Jason Hammel, Ricky Nolasco
