What's Gone Wrong With Brandon League?
by Matthew on Jul 31, 2010 11:00 AM PDT
in Seattle Mariners Stats
Email
Print
I began with David Aardsma for alphabetical reasons. Shawn Kelley followed for the same reasons. Jumping to Brandon League might make some think that Mark Lowe is up next. Those people have decent pattern recognition skills. He's not though. You should have realized that Mark Lowe didn't appear enough to warrant mentioning and got traded so screw him. Anyways, Brandon League.
The knee jerk answer if you have been following this blog or even FanGraphs for awhile is to blame the pitch selection. As I first mentioned both here and FanGraphs (and was subsequently repeated numerous times by many other people), League's splitter was the most best pitch in baseball last year as defined by the percentage of pitches resulting in a swing and miss. It was a pitch he started throwing only last season and we had high hopes that he would continue with that step forward and all those strikeouts coupled with some regression from his BABIP (.322), strand rate (just 68%) and HR/FB (15%) would reveal Brandon League as the next breakout reliever.
It hasn't worked out that way even though those who follow only ERA will think he's done fine this season. Truth is, he has been decent. We just hoped for more than decent and frankly, League is capable of delivering on those hopes.
The knee jerk answer above isn't the full answer but it isn't incorrect either. In 2009, League threw one splitter pitch for every 2.3 fastballs. That rate has dropped in 2010 to one splitter for each 5.5 fastballs. League is featuring his fastball more often and it is hurting his ability to get strikeouts as he is right back to the level he was accustomed to in Toronto pre-2009. He hasn't lost the ability to get those back though. Looking at the results for each pitch offers confirmation of this:
2009 Fastballs - 5% whiff rate, 66% strike rate, 51% ground ball rate
2010 Fastballs - 7% whiff rate, 65% strike rate, 59% ground ball rate
2009 Spliiter - 35% whiff rate, 59% strike rate, 61% ground ball rate^
2010 Splitter - 29% whiff rate, 47% strike rate, 40% ground ball rate^
^stupidly small samples
League is throwing fewer splitters and hitters are swinging at them less often, down four points from 2009. That has suppressed League's swinging strike rate and consequently eliminated the extra strikeouts he used to record.
It is not all bad news though. Increased usage of the fastball has predictably resulted in fewer walks and his net walk rate is down two points and even his grounders have increased. Neither have moved enough in the right direction to offset the loss of strikeouts however.
Returning to the old pitch mix would offer an almost immediate salve to League's strikeout woes. Under the assumption that his effectiveness with each pitch would remain the same as they have been in 2010, raising his splitter throw rate to 2009 levels would bump his swinging strike rate up to 13.5, nearly a match for 2009's 14.0%. Those extra missed bats would inevitably lead to more punch outs (and a few more walks).
The other part of the answer to what has gone wrong with Brandon League has to do with the expected regressions mentioned in the second paragraph. The BABIP has regressed as expected. The strand rate has just gotten worse, but since League has lost a small dinghy's worth of strikeouts he no longer is expected to post an above average strand rate. His current rate (67%) is still a bit too low, but that is not the chief culprit. That would be the home runs.
Let me get this out of the way again. I have looked at it a couple times with different methods and data and I, and others, have yet to find any solid and compelling evidence that ground ball pitchers suffers from higher HR/FB rates. So it is with confidence that I say that League's 18% HR/FB rate is too high. It's way too high and getting that down to a more normal range would go quite a ways to propelling League into that 'good' territory of reliever.
In conclusion, throw more splitter Brandon. You can control that part. Don't sweat the home runs though, those should take care of themselves.
34 comments
|
1 recs |
Do you like this story?
Comments
I'm not disagreeing with your conclusion
But his struggles have almost all come against lefthanded batters. He’s pretty much a ROOGY. It’s why he would be a terrible starter.
For his career, he’s .281/.345/.460 against lefties, or basically Evan Longoria.
Granted, they gave up a much better prospect and player for him, but he still has value in the right role.
I don't really see how this is at all is connected to what I said
and using career stats for someone who drastically changed his skillset just one year ago is misleading.
Every time I think of League the cliche scene from so many movies pops into my head
Old dynamite, sweating nitro. They need it so badly, and they’re so very careful with it and yet sometimes it still blows up in their face. Way to go, dynamite
hr/fb
It may not be because he is a groundballer, although maybe just some groudballers are prone. His hr/fb rate isn’t just 18.2% for this year, but also 18.2% for his career. Since the stat has been compiled, only 4 pitchers have had an 18% rate over a full season. Lowe and Webb in 2005, Maddux in 2004 and Odalis Perez in 2003. All had gb rates above 50% (likely the small fb sample means they will have higher, and also lower, rates randomly). But there were about 120 full seasons by groundball pitchers (over 50%) in that time, 17% of all full seasons, and only those 4 went over an 18% hr rate.
But League has been above average every season except his first 4 inning season, almost always in the mid to high teens. That’s fairly consistently high, and over 255 innings, 18.2% is very high. So maybe he is just an outlier, and has rolled craps every year. Even if it has nothing to do with him being a groundballer, he may be the exception and is just more prone to a meatball now and then.
Park factors.
They’re kind of important.
Toronto: HR: 113 / 112
Seattle: HR: 92 / 88
Good point
Only 5.9% this year at home, small sample. His career away percentage is 22.5% over 128 innings.
And the park factors are telling us what in relation to Brandon League?
You introduced them and told us they’re important, but not why and how they relate to Wobatus’ comment. I’m curious as to your point.
by nathaniel dawson on Jul 31, 2010 6:20 PM PDT up reply actions
His HR/FB has been high for his career because he was in a hitters park.
Now that he’s moved to Safeco it should have gone down, since Safeco is a great park for neutralizing flyballs. It has instead risen. This is bad.
I don't suppose we know if that's what Matthew was meaning...
but in reference to his HR/FB rate, it’s been high for his career because he’s given up a lot of meatballs. Pitching in Toronto hasn’t had that much to do with it. If you want to park-neutralize that 18.2% number, it becomes something like 17%* or so. That’s still really high.
*I’m not even sure it’s appropriate to do that, because I’ve never seen park factors that factor for HR/FB rates. But it’s good enough for this.
by nathaniel dawson on Jul 31, 2010 6:56 PM PDT up reply actions
That's very obviously what Matthew meant.
And your attitude is unwarranted since you tend to be wrong all the time. Please present your requests in a more productive/humble manner since everyone else seemed to know exactly what Matthew meant.
...and now I'm here
Huh?
I’m not sure what you’re reading into my comments. The perception doesn’t match the reality of my intent.
“Everyone else” meaning you and Wombat? Until your reply, Wombat and I were the only ones to reply to Matthew.
by nathaniel dawson on Jul 31, 2010 7:16 PM PDT up reply actions
Uh, what?
I’m not even sure it’s appropriate to do that, because I’ve never seen park factors that factor for HR/FB rates
That’s….that’s kinda what all HR park factors do…..I am so confused by what you are confused about.
Also I have no idea where you got the idea it would only lower his HR/FB% by 1% if you neutralized park factors
He gave up 121 fly balls in his total time in Toronto and 22 home runs. Neutralizing Toronto’s park factors reduce the number of homeruns by 13% for righties and 12% for lefties, so let’s say that’s about 3 homeruns you’d take away. 19 homeruns for 121 fly balls = 15.7%. That is 2.5% hr/fb shaved off just from the difference between Toronto and a completely neutral park for homeruns, unless I’m doing something wrong here. Then consider Safeco is not a neutral home run park either. The sample sizes are so small this is a ridiculous exercise anyways, but so are any conclusions about him being especially home run prone on his flyballs.
Oops, you'd have to park neutralize his # of flyballs too wouldn't you
I dunno, I’ll leave park factors to the smart people.
He gave up 121 flyballs while he was with Toronto, yes
Without looking at splits, only somewhere around half of those were while he was pitching in the Skydome. The other half wouldn’t be affected by Skydome’s park factor.
by nathaniel dawson on Aug 1, 2010 1:26 AM PDT up reply actions
Park factors, as I have seen them, factor for home runs per PA above or below league average.
While I don’t know that they don’t exist, I have never seen any that factor for HR/FB rate. I’d have to guess that there would be a high correlation between the two, however.
by nathaniel dawson on Jul 31, 2010 9:02 PM PDT up reply actions
Matthew derived these and put them on StatCorner
by Jeff Sullivan on Jul 31, 2010 9:04 PM PDT up reply actions
HR/FB rate, or just total flyball rate?
I see a home run factor and a flyball factor (OF, I’m assuming that’s a flyball park factor).
An odd thing I noticed. The numbers I see list the home run factor (92/88) as nearly the same as the flyball factor (92/91), which would suggest to me that the HR/FB rate for Safeco is just about league average, and Safeco depresses home runs because it depresses flyballs, not the rate at which they’re hit out.
Is that right? Without ever seeing flyball factors for Safeco field before, I would’ve assumed that the reduction in HR’s was because of a reduction in the HR/FB rate.
by nathaniel dawson on Aug 1, 2010 2:06 AM PDT up reply actions
Come to think of it, I don'think that's exactly the way it would work
A park that increases home runs by 12% would have to increase HR/FB rate by more than 12%, right?
I’m not sure how it would work out, but HR/FB% would probably go up by 20 or 30 percent……..or something like that.
Anybody, anybody?
by nathaniel dawson on Jul 31, 2010 9:00 PM PDT up reply actions
So I think his point it that while his HR/FB is close to his career average, park factors expose him as a worse pitcher.
That’s what I understood him to mean anyway.
League has allowed 154 fly balls for his career, according to Fangraphs
That is the denominator. That is a small denominator. Felix posted a HR/FB% of 16.4% in 2006, with 140 fly balls. He has since regressed, and his career rate is ~average.
Because League is such an extreme groundballer, we have to wait a lot longer for his HR/FB to stabilize.
by Jeff Sullivan on Jul 31, 2010 7:23 PM PDT up reply actions
i mention
that due to the small samples sizes, you get more volatiiity with groundballers. League also has high rates year after year. Likely just an unusually long run of bad luck, but it is interesting that his rate this year matches his career rate at a very high level, and he is consistently above average.
It's not interesting. It's trivial for exactly the reason Jeff states.
154 fly balls is not an unusually long run. You cannot be blinded that its happened over several seasons.
OK
Is there a point at which you’d start to be interested, not with respect to a streak of years, but total innings wise? 500 innings of 18.2%, which is what, 70% above average? 1,000 he will likely never reach that many)?
Not IP
Flyballs. How many flyballs?
Webb and Lowe have given up a lot more than that. Webb has a rate of 13.2% over 695 flyballs. Lowe a rate of 12.7% over 1170 flyballs. Both higher than average but not by much, with it going lower as you get higher in sample. Figures.
Well, League is an outlier, but if he keeps doing this for another 150 flyballs, I’d find it interesting.
Year after year?
Here are League’s fly ball totals by year:
26
18
8
14
53
33
Who cares what his HR/FB looks like year after year?
by Jeff Sullivan on Jul 31, 2010 7:53 PM PDT up reply actions
Understood
All tiny samples, and he could have easily given up 0% in some of those years, the samples are small. Seems pretty consistent. He hasn’t had all these homers bunched in single stretch. 18% this year, for his career, pretty regular, home runs to flyballs.
How many flyballs should we see before we might expect normalization? 300?
What about IIFB%
Fangraphs appears to count iffb% as a percentage of total fb. And hr to fb% seems to therefore include iffb. I recall a discussion in a thread there where someone else doing primary research said he didn’t include iff in calculating hr/fb, since iff by definition won’t be home runs. Which I found to be an interesting way of looking at it. Afterall, all flyballs that are not home runs are by definition not home runs.
The fangraphs author said this didn’t change the overall randomness when he checked.
However, League does have a seemingly very low IFFB%. 5 iffb in his career. More of his fly balls are leaving the infield. It makes sense that flyball % is generally random. There isn’t that much difference between a flyball that goes the 180 feet or so to get the ball beyond an infielder but not out to the 330 down the line that gets it out.
Again, League is a very small sample again of fly balls, and I don’t know how random IFFB% is. 3.2% seems low compared to the team averages I scanned. Conversely, and outlier time again, guys like Johan Santana who has given up 2000 flyballs in his career, or Rafael Betancourt, who has given up over 600, have fairly high relative IFFB% in their careers, and fairly low, fairly consistent, low hr/fb percentages. I think Santana and Betancourt have generally played in homer suppressing parks, but their road numbers are similar, if slightly higher on the hr/fb.
I don’t have a problem with hr/fb being largely random. However, I could see a few pitchers having some slight effect. It might generally be so slight that it wouldn’t deviate from the number you’d expect from pure random distribution. The chance of a ball going 300 feet versus 380 feet is a question of a given batter’s strength and usually just a fraction of an inch in where the bat hits the ball. Even if he falls within the expected random distribution, maybe Johan Santana, who has a much larger sample than League to look at, just gives up shorter flies on average, iffb or otherwise, than most pitchers. Maybe that type of “skill” is rare, small, and hard to measure. But that is not the same as saying it’s completely random.
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.
