Bernie on Haren's Late-Season 'Fade'
by bgh on Jul 23, 2010 5:10 PM EDT
Email
PrintI have an up-and-down readership with Bernie. Frankly, since the P-D inexplicably ceased the Google Reader feed of his Bytes blog (as well as the Cardinal Beat and Bird Land feeds), I have not read anywhere near as much of his writing. However, the trade deadline expands a fan's reading list and no internet stone goes unturned. I was over at stltoday.com/cards overturning stones when I came across this Byte post of Bernie's. I enjoyed the piece on the whole--especially when Bernie essentially labels McLane and Wade stupid--but I take exception to the portion addressing Danny Haren.
But here's a question that should at least be explored: just how good is Haren?
How good is Danny Haren? Um, well, he has a career FIP of 3.69 and a career xFIP of 3.61. During his time in the big-leagues, he has struck out 7.74 batters per nine innings and walked 1.97. In '08 and '09, he was an elite pitcher, FIP'ing 3.10 and 3.23 respectively and striking out over 8.5 batters per nine innings while walking less than 1.70 per nine. Yeah, Danny Haren is really good at pitching. I would love him to return to St. Louis.
Okay, okay, so Haren has a 4.60 ERA this year. If I cared about ERA, I would care about his ERA. This season, Haren is striking out 9 batters per nine innings, walking 1.85 per nine innings with a career-high HR/9 and a BABIP of .350, which is about 45 points above his career average. Because of Haren's uptick in HR's this season, lets look at his xFIP, a stat that normalizes HR rates. His xFIP 3.39. To put this in perspective, his career xFIP is 3.61, which is 22 points higher than in 2010, yet his career ERA is 3.71, a full run lower than in 2010. Clearly, this year is a bit odd statistically. I would suspect Danny Haren is putting up a far lower ERA (if you're into that sort of thing) moving forward. (For what it's worth, ZiPS agrees. Its Rest of Season projection for him is a 3.32 ERA.)
Bernie continues:
He has a history of fading in the final two months of the season. It's happening again this year. Since 2006, Haren has had a 2.98 ERA before the All-Star break and a 4.55 ERA after the ASG. From 2006-2009, he's had a 4.78 ERA in August and a $4.75 ERA in September. I just think this is worth pointing out.
Oh, I am sorry. I thought we were looking at his entire body of work as a pitcher. Apparently, Bernie is into ERA. He is probably a BWAA member, so we shouldn't be surprised by this or even really hold it against him all that much, because he probably doesn't know any better. The problem is that we aren't voting for the Cy Young here. By Bernie's own set-up to this stat dropping, we are trying to figure out how good Dan Haren is and how good he might be, in the future, for the Cardinals. To me, this lends itself to the sterile, fielding-independent stats of FIP and xFIP that we discussed above. So, let's use the Fangraps splits to analyse Haren's so-called fade.
March/April FIP/xFIP: 3.72/3.67
May FIP/xFIP: 3.71/3.71
June FIP/xFIP: 3.48/3.71
July FIP/xFIP: 3.36/3.52
August FIP/xFIP: 4.28/3.37
September/October FIP/xFIP: 3.61/3.64
I just don't see the so-called fade. This is because Haren's pitching performance is consistent very good. For whatever reason, he gives up a few less homers in June and July and a few more homers in August, but his xFIP's are amazingly constant. There really is no reason to expect Danny Haren to "fade" in July, August, or September--particularly in the home run suppressing confines of Busch III. Bernie is looking at the wrong stat (ERA) here to draw the off-base conclusion that Haren "fades." What's disappointing is that we know that Bernie visits Fangraphs and we know that Bernie is open to new methods of measuring player performance. So, why then, in a situation that is tailor-made for the use of such statistics, does Bernie revert to BBWAA form and use ERA to draw a misguided conclusion about a once and (potentially) future Cardinal?
18 comments
|
5 recs |
Do you like this story?
Comments
great post
I’m of the opinion that if we “go all in” and get a starting pitcher it should be Haren. unless the Astros soften on what they want for Oswalt of course. but Haren is a great pitcher and would be even better than Oswalt imo.
but then again, I think with either of these guys there would have to be some sort of payroll restructuring going on.
chief justice
by Cards Fan in Chitown on Jul 23, 2010 7:08 PM EDT reply actions
Bernie has a valid question
However, Bernie lacks on a valid response to the question there.
With the ERAs and FIPs and xFIPs differing so much, could it be that the D’Backs fade down the stretch, thus Haren’s numbers don’t look quite as good to the unsophisticated eye?
agree- I think it's at least interesting
that 7 months out of 8, over a 4 year period, he’s underperformed his FIP significantly in Aug/Sep. He might have been unlucky, in all of those occasions, but I think it’s enough of an interesting statistical anomaly to warrant an article about on the P-D. And it wasn’t even an article, just a blurb.
(FWIW, 4 of those 8 months were for the A’s. Not sure if that means anything, probably not, but it’s probably not just a case of the “diamondbacks fading”).
Well the girls would turn the color of the avocado when he would drive down the street in his El Dorado... -the modern lovers
This is probably quite a stretch...
but, as you probably know, it gets really dry in Oakland and all along the west coast as the summer wears on. August is the trough. Same with Arizona, but only because it gets insufferably humid out and they close the roof and turn on the AC.
Fire La Russa!
since the P-D inexplicably ceased the Google Reader feed
i don’t know exactly how google reader works because i’ve never used it, but imagine that you have to redo it since they changed the website. URLs are different and stuff
"Moneyball: It's kind of like communism."
twatter
New P-D Website
I have tried searching for the blogs on Google Reader to add them anew, but the only blogs that show up have no new postings after late June. If the new address is available on Google Reader, I have yet to be able to find it. (I haven’t looked in a couple weeks.)
"I'm gonna throw the nastiest curveball I have ever thrown...if he hits it, I'll tip my cap, but if not we're going to the Series."
--Adam Wainwright on the final pitch of the 2006 NLCS
by bgh on Jul 24, 2010 10:28 AM EDT via mobile up reply actions
It's mostly that his babip shoots way up and he's given up a lot of home runs in August...
If you look at his batted ball numbers his line drive percentage ticks up slightly as the season wears on, his groundball percentage also climbs, while his fly ball percentage goes down. I know there’s often a problem comparing these numbers across years, but even if they are accurate it all points to bad luck. His babip should not be skyrocketing and he certainly shouldn’t be giving up more homeruns on fewer fly balls. Also, he’s only pitched 200 innings per month.
It would be odd not to acquire a guy because his hr/fly doubles in August compared to all the other months.
Fire La Russa!
It is interesting that he's underperformed his xFIP in August and September throughout most of his career.
xFIP normalizes his home run rate to league average, so I don’t think it’s a good thing to use to determine if he does have a late season fade, because there might be a specific reason that he’s giving up more homers in those two months than during the rest of the season.
Look at his 2009 splits. He gave up 14 homers through July 31st, then gave up 13 in the final two months of the season. His HR/FB jump significantly, but his batted ball percentages stay pretty much the same. Now, his BABIP goes up to .352 in September, which also corresponds to a jump in GB%. That would indicate to me that he’s giving up a lot of hard hit ground balls that are going for hits, and maybe his percentage of FAT pitches is higher during these last two months.
Looking at Texas Leaguers pitch F/X database, I don’t see much difference between his pitches prior to August 1st and after August 1st, although I would note that his percentages of fly outs and singles go up by about 4% each. I’m not sure if that’s significant, but it would seem to support my hypothesis above that he’s giving up more hard hit balls in those two months for whatever reason.
Now let’s look at 2008. He significantly underperformed his xFIP and FIP with his ERA in the months of August and September in those months with BABIP’s of .426 and .347. One could surmise that he’s getting unlucky, but then look at the batted ball data. His line drive percentages go WAY up in those two months, and he gave up a ton of homers in August 2008 as well. I don’t think that BABIP is really unlucky by this measure — it looks to me like he’s just giving up way more hard hit balls in those two months than the rest of the season, which FIP and BABIP aren’t really looking at.
Clearly hitters are hitting the ball much better off of him in those two months than they are the rest of the season, so I think Bernie may have a point that he’s just not as good in those two months, at least in the last two seasons.
Can Colby round out our new MV3?
If VEP or Steve Sommer are around, perhaps they could look at the percentage of FAT pitches he's throwing?
I can’t seem to get the pitch F/X sites I use to do this the way I want them to.
I would be happy to learn how to do this on my own, as long as it doesn’t involve me classifying pitches from my own database — I’m interested, by not THAT interested, lol.
Can Colby round out our new MV3?
September
If he goes into a late season fade, I don’t see how his career September xFIP and FIP are better than April.
"I'm gonna throw the nastiest curveball I have ever thrown...if he hits it, I'll tip my cap, but if not we're going to the Series."
--Adam Wainwright on the final pitch of the 2006 NLCS
by bgh on Jul 24, 2010 1:44 PM EDT via mobile up reply actions
They are better than April, that's true, and I get what you're saying.
Think about it this way though. FIP is an attempt to measure only the things that a pitcher can control, while xFIP does the same but also controls for HR/FB by pegging that to league average before running the numbers. ERA however, is an actual measurement of earned runs given up by that pitcher. It’s not a statistic measuring what we think is valuable by controlling or eliminating certain things that we believe or don’t believe have value, it’s an actual measure of earned runs given up, on average, assuming 9 innings of work.
I’m not saying the Bernie is right for using strictly ERA, he’s not. But ERA isn’t a meaningless stat like Wins is, it does measure something valuable, and when you compare how much a pitcher underachieves or overachieves his ERA by the advanced measures, as well as looking at batted ball data (like BABIP, LD%, GB%, HR/FB%) you can discern even more information. I think taking his FIP and xFIP into account solely, without factoring in ERA and looking at batted ball data really isn’t much better than looking at ERA by itself. Now, lets look at a couple of statistics to demonstrate what I mean.
If you look at Haren’s career splits, you’ll see that August is by far his worst month of the season. His FIP is 4.28 in August, and that’s over half a run higher than his career average of 3.69. His xFIP for that month is his lowest of any month split and is a full run lower than his FIP. Why? Because his home run rate in August is 1.51/9 which is .50-.75 higher than in any other month split. That’s a HUGE difference in terms of career rate. If this was one month of a season, or a couple bad months in a row, you could chalk it up to bad luck. But when his career split over 7 full seasons is that high I would think it says something about how he pitches during the 5th month of the season. Whether it’s that he gets more of plate, or guys seem to square the ball up better, or the ball carries better in the humid air of the late summer in Northern California and Arizona, I don’t know. But I do think it’s significant and not just luck or statistical noise, and his ERA reflects it — he allows more runs in August than in any other month historically.
September is the month in which he most underachieves his FIP/xFIP the most in terms of ERA. Looking at the batted ball data, one thing really sticks out — his inflamed BABIP of .347 during that month. Generally, we accept that BABIP is a luck factor, that over the course of a full season it normalizes to around .300 or so since that’s what the league average is, and that pitchers significantly higher than that are “unlucky” and pitchers significantly lower than that are “lucky”. Pairing it with batted ball designations makes it an even better measure.
In this case, however, Haren’s batted ball classifications are nearly the same as the rest of his splits, and he’s thrown about a full season’s worth of innings in September (214 IP) over his 7 year career in the bigs. This indicates to me that although the line drive percentage isn’t up any, hitters are hitting the ball harder off of him in September and that because of this, hitters are getting more base hits on similar amounts of the same type of ball in play. Instead of hitting a weak ground to the second baseman, they’re ripping a 2 hopper through the hole or up the middle.
In summary, I don’t think you can discount his BABIP as “unlucky” any longer, since it seems to have been higher than .300 in the months of August and September for pretty much his entire career, and I think it says a lot that the only two months in which his ERA underperforms his FIP and xFIP consistently by more than half a run is in those two months. His ERA overperforms those measures in nearly every other month of the season. Clearly there’s something different about him in those two months that makes him easier to hit. I don’t think it’s just a statistical anomaly.
Can Colby round out our new MV3?
by fourstick on Jul 24, 2010 4:56 PM EDT up reply actions 1 recs
Sept. HR Rate
If the reason that Haren has a higher September BABIP is because he is giving up harder hit balls, then why is his HR rate not higher a la August? Also, giving up firmer contact typically raises one’s LD rate, which isn’t happening in September for Haren. I don’t think there is any more concrete data available to support the conclusion that Haren gets hit harder in September than the conclusion that he has been unlucky in on batted balls in that month. This is why I wish Hit Tracker kept track of hit speeds for all hits.
We could use the same type of reasoning from above to explain away every uptick in BABIP as a result of how hard a ball is hit. Where such an uptick is accompanied by tangible evidence of harder hits—such as more LDs and HRs—then I would feel better about such a conclusion. No one would argue that Miles is hitting more balls harder as the reason for his BABIP being high. His LD rate is down. His BABIP is up. This doesn’t mean he is hitting harder grounders and flyballs, though. You’re right to check BABIP against batted ball rates, but I don’t know that I would draw the conclusions you do about Haren’s September numbers where there isn’t really any basis for it in the peripherals.
"I'm gonna throw the nastiest curveball I have ever thrown...if he hits it, I'll tip my cap, but if not we're going to the Series."
--Adam Wainwright on the final pitch of the 2006 NLCS
by bgh on Jul 25, 2010 3:37 PM EDT via mobile up reply actions
There's too much information that the peripheral percentage stats don't give us though
I just don’t think you can say that giving up more runs in August and September over 200+ innings over 7 years can be deduced as being “unlucky”. You can’t make the case that the teams he’s played on just suck at defense, but only in those two months. Clearly something else is going on, because he gives up many more runs in those two months than in the first four months of the season.
Can Colby round out our new MV3?
it can, actually.
I certainly wouldn’t say “just bad luck”, but although things regress to the mean, when you flip 1000 coins you’re not always going to get between 480 and 520 heads. Sometimes the unlikely happens, and that’s how it seems with Haren. I don’t really buy the harder-hits thing because more hard-hit balls —> more line drives, and that hasn’t really been the case.
אנא טוני לא יותר ט.א.ס.ס
Prove the more hard hit balls = more line drives
You could certainly have more hard hit ground balls and fly balls that result in more base hits and home runs (which may be occurring here) without line drives going up significantly
Can Colby round out our new MV3?
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
In Association With
Big League Stew - MLB - Yahoo! Sports
- Photo: Padres fan regrets letting mother capture moment with Trevor Hoffman
- Marlins agree to unretire No. 5 for Logan Morrison, who will honor his father’s favorite player
- Weatherman Derek Holland does Dallas TV
- Ask Alex: Will Julio Teheran or Matt Moore win rookie of the year?
- Ku Klux Klan, umpires and bowling: Why the minor leagues tossed out Jim Evans academy
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
