FanGraphs Fantasy Baseball


RSS feed for comments on this post.

  1. somewhat irrelevant but I have a serious man crush on cj Wilson I have owned him for his career as a starter and it makes me wonder what he may be had he began his career as a starter.

    Comment by josh — July 10, 2012 @ 8:31 am

  2. I watched his last start, and he really seems to perform better under pressure. Walks a few guys, gets into a lot of trouble, and it’s like he zones right in. Like he needs to feel the squeeze before he starts to come through. It makes for maddening watching, but it seems to work for him.

    Comment by kernel — July 10, 2012 @ 8:44 am

  3. Your predictions pretty much suck, man. No offense or anything cause I know they were just for fun. But yeah, they suck.

    Comment by supershredder — July 10, 2012 @ 9:06 am

  4. I agree that Wilson can cease getting lucky and still be an outstanding pitcher, but to say that Wilson or anyone else has demonstrated anything over a 3 year period is nonsense.

    Let’s accept the idea that some pitchers do indeed have the ability to consistently outperform FIP. Let’s also say that a typical pitcher without this particular skill will have a 0.5 chance of outperforming FIP in a given year, and a 0.5 chance of underperforming. Conditional on being allowed to pitch for three years, he will have a 0.125 probability of outperforming FIP in each of the three years by chance alone. With c. 150 starting pitchers in any given year, and maybe 50-100 starting pitchers who start in every year over a given three year period, we should expect to see about 7-15 pitchers who outperform in any three year period simply based on luck, and some small number more based on the FIP outperformance skill we have posited.

    Let’s say there are 5 pitchers in the skill bucket at any given time – I haven’t seen anyone nominate more than 3 or 4 current pitchers for this particular skill, so 5 seems a reasonable estimate. So among the FIP outperformers over a given period, we have 5 in the skill bucket and 7-15 in the luck bucket. So with a threshold of 3 years, we should expect MORE outperformers to be in the luck bucket than the skill bucket. To conclude that a given pitcher is in the skill bucket with even a 0.5 probability, we either need some piece of exogenous data besides his ERA vs. FIP performance, or we need a longer history of ERA/FIP outperformance than just 3 years.

    C.J. Wilson, or Cueto, or Vogelsong, or anyone else you want to nominate, may very well have a repeatable skill for outperforming FIP. But 3 years of data is just not enough to draw that conclusion about anyone in particular. Neither is 4 years. 5 years might be on the border, but you’re still talking about confidence levels of c. 0.8 at best… 5 people in the skill bucket, and an expectation of c. 1-3 in the luck bucket, which implies p less than 0.833.

    It feels like 3 years is a lot of data, but it just isn’t.

    Comment by mcbrown — July 10, 2012 @ 9:35 am

  5. I just traded Wilson for Bourne. I think it’ good trade, my team is hurting for SB.

    Comment by Jason — July 10, 2012 @ 9:50 am

  6. The data backs this up, at least for 2012. It is possible that he is pitching to contact more with the bases empty and then really focusing on getting the ball down with guys on. Or perhaps he just has better mechanics from the stretch. But regardless, as a Wilson and a Lincecum owner, I’ve spent all year wondering how I can get Wilson to avoid the walks with no one on and how I can Lincecum to avoid falling apart when guys get on base.

    Comment by Chad Young — July 10, 2012 @ 10:18 am

  7. It doesn’t work very well for him in the postseason.

    Comment by Lisa — July 10, 2012 @ 10:19 am

  8. No offense taken. At some point I will do a retrospective on those. Some of them were pretty spectacular failures, but I am pretty happy with the Wilson prediction, Prince is on pace for under 30 HR. Rosario has shown the power and could still be the most valuable rookie catcher from June on, although that remains to be seen. The Morneau prediction wasn’t TERRIBLE either. But when I missed…Raburn, Gio, Carp…yikes.

    Comment by Chad Young — July 10, 2012 @ 10:23 am

  9. you make a bad assumption early in your analysis. it doesn’t make sense to start with a p=0.5 under- or over- perform. what you’re saying is overperforming by 0.01 (<1%) is the same as overperforming by .31 (10%). no one should be saying that there is skill in outperforming fip/xfip by 0.01.

    thus, your resulting conclusion is wrong. someone was discussing this in terms of coin tosses in the other thread. what you've set up is potentially treating the outcome of a coin coming up heads 51 times out of 100 as the same probability as a coin that a coin comes up heads 65 times out of 100. the degree to which a player outperforms fip/xfip expectation *should* matter in determing whether there's something at work that allows him to regularily exceed expectations.

    Comment by ccoop — July 10, 2012 @ 10:31 am

  10. Agreed that 3 years isn’t a ton of data, but it also depends what your assumptions are. If you start with the assumption that FIP is NOT perfect and that there are guys who can consistently outperform it, three years does suggest that Wilson might fall into that class.

    If you want a few other guys who have consistently outperformed, Johan Santana has had an ERA .2 lower than his FIP for his last 6 seasons and 7 of 8 since becoming a full-time starter. Ted Lilly has done it his last 6. Carlos Zambrano for 4 of his last 5, Halladay for 4 of 6, Jeremy Gutherie for 4 of 5, Kershaw, Joe Saunders, Cueto, Cahill all for 3 of the last 4, and Wilson, John Lannan and Dickey for all of the last 3. None of those guys, in the past 7 years, has had an ERA .2 HIGHER than their FIP and all have consistently had an ERA at least .2 lower. That doesn’t even include the anti-FIP poster-child, Matt Cain, whose ERA has been within .2 of his FIP for 4 of the past 7 years, but is .35 lower for his career.

    Having said that, none of this is definitive – I haven’t really researched the data enough to make a strong statement one way or the other. But I don’t think FIP is the perfect stat. I like it a lot, I use it a lot, I think FIP, xFIP and SIERA are all great stats that tell us a ton, and are far more informative than ERA. But they also all have their limits. My point with Wilson is less “he clearly beats his FIP every year” and more “don’t just point to FIP and say he will regress.”

    Comment by Chad Young — July 10, 2012 @ 10:33 am

  11. If you need speed and have pitching, nothing wrong with that deal at all. My point isn’t to say that you should hold Wilson at all costs – just don’t sell low out of fear. I own him in one league where I need offense, and may very well trade him myself. I’m just not going to throw him away.

    Comment by Chad Young — July 10, 2012 @ 10:35 am

  12. Yeah, agreed Coop. That’s why for the quick analysis I did below, I broke guys into segments of .2 below or above, not just below or above. Not sure it is clear in my comment, but in that list, a guy who has outperformed FIP 3 of the last 4 years, was more than .2 better than his FIP for those 3 years and with .2 of his FIP the other year.

    Comment by Chad Young — July 10, 2012 @ 10:40 am

  13. Great article. Love thinking about both sides. I recently moved Wilson and Cuddyer for Felix Hernandez in a H2H 10 Team Keeper League. While I was never going to keep Cuddyer, I believe Felix at $10 more than Wilson is a better long term investment.

    Comment by Matt — July 10, 2012 @ 10:41 am

  14. I traded CJ for Justin Upton …. keeper league …. I feel pretty good, despite JUP sucking this year.


    Comment by KFmfC — July 10, 2012 @ 10:42 am

  15. Excellent Chad! I wrote the below after Jeff’s article.
    “If a pitcher is so good as to be able to avoid hits and home runs more than others, why then can’t he just avoid them hitting the ball all together? I know pitching to contact is used. But what about pitching to no contact with runners on or in scoring position? Or pitching to more ground balls with a runner on first?”
    I see you looked up data for 2012 only. I won’t, but maybe someone should look up how he’s done with runners on from the last three years. If he’s done this (higher K and GB rates) all three year then I’d buy into him a lot more. Otherwise, I see a higher possibility for bigger regression.

    Comment by Chicago Mark — July 10, 2012 @ 10:44 am

  16. Thanks for the research. I must point out that you have named 13 pitchers, and even if my methodology is wrong as coop has suggested, we should expect to see SOME number of pitchers outperform FIP due to chance alone. So of these 13, how many do you think are attributable to skill and how many are attributable to chance?

    Comment by mcbrown — July 10, 2012 @ 10:48 am

  17. I have C.J. Wilson for $4 in ottoneu. I agree with the points Chad and Jeff have both made. For me, I would only consider trading him if I was getting a similarly priced bargain in return, as even if his performance falls off, there really is no risk at his current price. Anyone have thoughts on what you would pay for a $4 Wilson in an ottoneu FG pts league?

    Comment by jjdouglas1 — July 10, 2012 @ 10:51 am

  18. in short, I have no idea how many to attribute each way. I didn’t do enough research or dive deep enough on this to have a clear answer. I think the important thing is to note that even when you put on a bit of a barrier and say to “outperform,” the gap has to be at least .2, you still get a decent sized list.

    For example, my dataset has only 22 pitchers who qualified in at least 6 of the past 7 years. Out of that set, if we used a 50-50 shot to be over or under FIP, we would expect 1.5% of pitchers to go 6 years beating their FIP by at least .00001. I have 2 of 22 beating it every year by at least .2. I’d say that at least for those two (Lilly, Santana) there is something going on there beyond pure chance and statistics. There is some selection bias, as pitcher’s performing worse than their FIP are less likely to keep pitching.

    But, again, my point isn’t that Wilson is definitively better than his FIP, just that I think we can’t assume he is NOT definitively better than his FIP.

    Comment by Chad Young — July 10, 2012 @ 10:56 am

  19. A fair point. Let’s be more rigorous.

    In 2011, the standard deviation of ERA-FIP among all pitchers who threw 100 innings or more was 0.59. So let’s estimate the implied probability of outperforming FIP by 0.2 runs as 0.367, assuming this metric is normally distributed. The probability of outperforming by 0.2 runs for 3 consecutive years is 0.05, and so we should expect something like 3-7 pitchers to do this by luck in a given three year period.

    While we’re on the subject of rigor, we’re not factoring in the inherent bias in the data. Starters that have outperformed will tend to be more likely to continue starting in the future, while starters that underperform will tend to be less likely to continue starting and thus disappear from the data set. Thus we should expect the number of pitchers who display this historical tendency to be overrepresented in the data.

    My point remains. This may indeed be a real skill, but to differentiate between who has the skill and who has just been lucky, with any reasonable confidence, requires more than three years of data.

    Comment by mcbrown — July 10, 2012 @ 11:02 am

  20. I agree that we can’t assume he doesn’t have this particular skill. My point is that in your article you say that he has demonstrated that he DOES have it, and I don’t think we can make that conclusion.

    As you say, survivor bias is a massive problem with the historical data in this kind of analysis. To account for it will require a much more rigorous analysis than I have the time to do, and may well be impossible as we have no reasonable way on estimating the performance of people who fell out of the data set.

    Comment by mcbrown — July 10, 2012 @ 11:07 am

  21. He did some of it last year (most the change in batted ball profile). 2010 he did not. That might be a a sign of potential increased concern. Could also be a shift in his approach over the years.

    Comment by Chad Young — July 10, 2012 @ 11:08 am

  22. That standard deviation, though, would include anyone who outperformed by skill and anyone who outperformed by chance. Let’s say Johan, Lilly, and Wilson actually have a skill that allows them to consistently outperform their FIP by at least .2 every season. Then including them in a data set to find a standard deviation for E-F and using that standard deviation to assume a normal distribution does not work. I thought about doing something similar, but I think it is going to take a more advanced approach and more time than I can spend on this right now to figure this out.

    Comment by Chad Young — July 10, 2012 @ 11:11 am

  23. Be sure to come back and let us know how that worked out.

    Comment by Mr. Thell — July 10, 2012 @ 11:13 am

  24. I think we are actually on more or less the same page here. I may have over-stated my case in the article, more than anything else. Should not have implied that he has that skill…only that he MAY have that skill IF it exists. More meant to combat the immediate reaction to a big ERA/FIP split, rather than as a defining piece of my argument.

    Comment by Chad Young — July 10, 2012 @ 11:13 am

  25. Fair enough. Johan didn’t pitch in 2011, so I went ahead and removed Wilson, Cueto and Lilly from the data. The standard deviation fell from 0.59 to… 0.587. P fell from 4.96% to 4.93%. So not much of an effect.

    We’re not even considering defense here. The entire Rangers staff outperformed by 0.24 runs in 2010 and 0.19 runs in 2011. The Angels so far have outperformed by 0.33 runs. Part of that is attributable to Wilson himself, but remove Wilson and the teams still outperformed overall. How much of the outperformance is attributable to the pitchers and how much to the defense? If we’re going to claim a given pitcher has this skill, we definitely have to factor this in.

    Comment by mcbrown — July 10, 2012 @ 11:24 am

  26. Excellent discussion. Maybe the first long winded discussion I read all the way through on this website.

    I will say I largely agree with mcbrown’s points though Chad does make good points as well. I am naturally skeptical of anyone who claims to have an ability to outperform FIP at this point. However, I will say that for fantasy purposes, team defense (such as the Rangers and now Angels here) should be factored into any decision in a positive favor for the pitcher. When playing fantasy baseball, I do not care if Wilson is getting good fortune by the strength of his teammates gloves as long as it figure to keep happening.

    Comment by Colin — July 10, 2012 @ 1:27 pm

Leave a comment

Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *

Close this window.

0.087 Powered by WordPress