Pirates To Hunt For Pitching This Offseason
by Charlie Wilmoth on Sep 26, 2010 11:58 PM EDT
Email
PrintI thought this article at the Trib was pretty interesting, in part because Neal Huntington pretty much calls Andy LaRoche a bust:
"We expected more," general manager Neal Huntington said. "Andy is an example of why you can't count on just that one prospect (at any position). They don't always make it."
It sounds like LaRoche will be gone this offseason.
(UPDATE: The Post-Gazette had a very similar article on this a week ago, which I'd read and forgotten about.)
There is also a note on the pitching talent the Pirates could pursue this offseason:
Most of whatever money the Pirates spend on free agents this winter likely will go toward pitchers. They might be able to get a value price on a veteran such as Todd Wellemeyer, Jake Westbrook, Jeremy Bonderman or David Bush.
Bonderman is interesting, if only because he's young, but he's reportedly thinking about retiring, and I can't imagine the prospect of moving to a 100-loss team will keep him motivated to keep playing. Westbrook may have priced himself out of the Pirates' range by performing well for the Cardinals down the stretch. Wellemeyer's peripherals have been hideous since 2008. That leaves Dave Bush, whose strikeout rates have declined to such a degree that he might not do any better in front of the Bucs' defense than Zach Duke did this year. The Trib is probably right, though, that Bush and Wellemeyer are the sorts of guys the Pirates will probably pursue. Kevin Millwood might be someone from whom the Pirates could get decent performance for a good price - he's 3-16 this year with a 5.29 ERA this year, but he had a 3.67 ERA last year, and his peripherals this year are similar.
307 comments
|
0 recs |
Do you like this story?
Comments
Rather than waste big money on mediocre pitchers
I’d rather see us go with a rotation of JMac, Ohlendorff, Maholm, Karstens and someone from the collection of Lincoln, Burres, DCutch etc until around mid season when we could see Morris or Owens or someone else ready to make that step up.
That’s not to say that if a decent pitcher can be obtained in a trade that I would not be in favor of that. I hope we are able to pull another bullpen stunt like this year and pick up maybe two or three veteran arms to put behind Hanrahan, Meek and possibly Resop and then turn them over at the trade deadline for prospects.
Brian Burres ruined my last trip to PNC Park last Thursday. There’s just no way he should be in the rotation mix or else I’ll pass out the pitchforks and torches
He’s got the worst active ERA and xFIP among those with 300+ innings the last 3 years.
Chan Ho Park, Ledezma, and Thomas were poor from what I could see, as well. The bullpen will be addressed in free agency, because there is still no depth.
by Adam Reynolds on Sep 27, 2010 3:15 AM EDT up reply actions
I have about a dozen pitchforks.
But no torches. Although I do have matches.
by IAPiratesFan on Sep 27, 2010 4:07 PM EDT up reply actions
I have pitchforks, but I’d have to purchase some torches.
by Adam Reynolds on Sep 27, 2010 12:38 PM EDT up reply actions
Anybody else go to the final home game yesterday? I, my wife, and my dad all swore that when Andy announced the location that got his jersey, it sounded like he said “the unfortunate fan sitting in section etc”.
None of those pitcing options are very appealing, but that’s what I’ve been trying to tell the Zach Duke Hate Brigade for weeks now!
I was at the game yesterday as well, with my wife and our best friends. I told them I wanted to win LaRoche’s or Duke’s jersey so when I went to pick it up, I could snidely say, “Thanks for nothing this year.”
Kind of a dick move, man. I’m glad you didn’t win.
by Vlad on Sep 27, 2010 7:28 AM EDT via mobile up reply actions
I dunno
I was there and the same kind of thought went through my head. (Not that I would have actually said it.) I told the guys I was with that I wanted Steven Jackson’s shirt so I could stiff him and watch him wander around, looking for the fan who had his name, and finally realizing, like 10 p.m. on prom night, no one was coming.
Beer muscles...
…nothing more.
Seriously, if I’d won even Pedro Ciriaco’s or Chris Leroux’s jersey, I’d have acted like I just won Walker’s or Tabata’s.
I can be a dick in my head, but I tend to be a lot more cordial when the opportunity arises.
yeah i was there & he did say that. It was to be funny, as in, You’re unlucky to win my jersey (instead of someone else better)
by Danatural08 on Sep 27, 2010 11:34 AM EDT up reply actions
Oh, super.
Those rotation options are grim, man.
This is why I’ve been trying to stand in front of the “non-tender Duke” train to keep it from leaving the station. We aren’t going to replace him at equal or lesser cost with some ideal pitcher made from rainbows and unicorn farts. Instead, we’re going to end up holding the bag on one variety or a other of scrub who can’t even match Duke’s ceiling.
by Vlad on Sep 27, 2010 7:34 AM EDT via mobile reply actions
Agree
I agree. Duke is at least is good as Millwood at this point even in his down year. Last week vs. the Sox Milwood couldn’t break 87mph. I’m not sure he even plays next year. It was sad to listen to given what he once was.
Yinzers uber alles
Millwood = Duke
Looking at starts and using the method I outlined in the posts listed below, they about the same in HQS/QS/Avg%:
Millwood: ~53%, Duke: ~ 50% (not counting his last start)
http://www.bucsdugout.com/2010/9/17/1695552/share-pirate-starters-quality
http://www.bucsdugout.com/2010/9/16/1693555/pirate-starters-quality-starts-and
Yeah, but Millwood does strike out a few more batters. The reason Duke isn’t able to turn that xFIP into the same actual ERA is that he allows so many balls in play with a terrible defense. Millwood wouldn’t have the same problem to that degree.
by Charlie Wilmoth on Sep 27, 2010 1:20 PM EDT up reply actions
Matt Morris Millwood strikes out 2% more total batters than Duke. I’ll leave it to others to tell if that’s significant. Probably not bad in the AL East. The problem is that he’s older.
by Adam Reynolds on Sep 27, 2010 2:28 PM EDT up reply actions
Yeah. I do agree it would be easier to just keep Duke and upgrade the defense.
by Charlie Wilmoth on Sep 27, 2010 2:44 PM EDT up reply actions
Milwood strikes out a few more batters right now...
…but he’s also older and as such at greater risk of unexpected decline.
If we were going to pick up an old guy with the intention of keeping the ball out of the hands of our defense, Harang would make more sense, IMO. [Assuming that the Reds decline his option, of course.] 6.48 K/9 this year, and it’s actually his lowest figure since 2003.
Plus
we wouldn’t have to face Harang anymore.
Actually, I don’t know – does he still shut us down the way he did a couple years ago? The guy was Cy Freaking Young against us.
A couple of years ago...
…he was shutting everybody down. He was a legitimate front-of-the-rotation guy in ’06 and ’07.
Only one game against us this season – gave up four runs in 6 1/3. Only one game against us last year, too, but it was pretty memorable – nine shutout innings, three hits (the only baserunners he allowed), 9 Ks.
I remember that game
In fact, I recall that he had fallen off a bit, and I was thinking, “Maybe we can finally get to him.” Nope.
I ate
A pie made of rainbows and unicorn farts last week. It was a bit tart but still delicious.
Wait, what was the point here?
Unicorn farts are the new OBP
Billy Beane is scouting Narnia as we speak.
by Garrett122 on Sep 27, 2010 12:12 PM EDT up reply actions 1 recs
I'd rather get rid of Duke whether or not Kevin Millwood/whoever
is better than him. I’d rather let Morton/Lincoln/Resop/Karstens/Burres battle it out for the #4-5 spots in the rotation (Behind McDonald, Dorf, and Maholm) than hand one to Zach Duke for his salary. That is to say nothing of the fact that Owens/Morris/Wilson/Locke will all be at AAA, and at least one of them will probably be ready by June or July.
It’s a waste of resources to pay a fringe #5 starter $6 million when you’re the Pirates.
Why would Burres be in the mix? He stinks.
by Adam Reynolds on Sep 27, 2010 2:16 PM EDT up reply actions
Burres being in the mix or not doesn't change my opinion on Duke
But whenever I’ve watched him, he hasn’t pitched half-bad. He might be someone worth bringing off the scrap heap again. Also, Crotta should be in the mix, as it looks like he pitched way better than his ERA indicates.
What do you think about him having the worst active career ERA in baseball (with peripherals to match)?
by Adam Reynolds on Sep 27, 2010 3:17 PM EDT up reply actions
Burres being in the mix should affect your opinion on Duke.
Insofar as Burres is terrible, and significantly worse than Duke. Our ability to bring in a tolerable replacement at a reasonable price is, by and large, the essential question when deciding whether or not to bring back Duke.
Crotta would get murdered by our defense just like Duke was this year. Duke has struck out 5.5 batters per nine in the majors. Crotta struck out 5.96… in AAA. He’d put up a Duke-type performance, only more so, insofar as he gets even more ground balls and misses even fewer bats.
I don't think Burres is significantly worse than Duke
What I’m saying is that regardless of whether we have 4 guys to fill the 4-5 slots or 5, I don’t think it’s worth it to hand one of those spots to Duke. So basically, I don’t hate Burres as much as you guys, particularly given that he’s being paid peanuts. I would rather have him in the mix than not. If he’s not in the mix, fine, but that still doessn’t change anything with regards to Duke for me.
You hit the nail on the head with regards to the stuff about a tolerable replacement for Duke. I agree, but I am pretty sure that we don’t need to “bring in” anyone. I am comfortable letting Morton/Lincoln/Resop/Karstens battle it out for the 2 spots. That is the core of my argument. The less important, almost tangential part, is that we might as well let Crotta and Burres compete as well. It is this part that you guys are turning into a straw man, implying that I am gun ho to hand over Duke’s spot to Crotta or Burres.
This year, Burres's xFIP is 5.16
Duke’s is 4.42.
Burres has a lower strikeout rate, a higher walk rate, and is averaging fewer innings per start. By just about any reasonable measure, he’s significantly worse than Duke.
Between Morton, Lincoln, Resop, and Karstens, we’re likely to come up with about one ML-caliber SP next year. As such, I don’t care for your plan – it calls for us to fill the rotation with random bodies, which didn’t work out so hot for us in early ’10.
2 points
1) “The less important, almost tangential part, is that we might as well let Crotta and Burres compete as well. It is this part that you guys are turning into a straw man, implying that I am gun ho to hand over Duke’s spot to Crotta or Burres.”
So that should address your first paragraph pretty well. This isn’t a Duke-Burres comparison.
2) As far as I’m concerned, Duke is no better than a random body himself- he either pitches at an average level, or gets lit up like a Christmas tree. He’ll just be a far-more-expensive random body next year. If I believed that Duke were on a higher tier than Morton, Lincoln, or Karstens, maybe I would reconsider my position.
I see below that they won’t let Resop start- that’s kinda silly of them, in my opinion.
If you weren't comparing Burres and Duke...
…then why did you say, “I don’t think Burres is significantly worse than Duke”? That seems, to my eyes, like a comparison.
If you see Duke as a random body, then you’re mistaken. It makes your position more logically valid, but less true.
Because it's a tangential point
Come on, Vlad, I know you’re a smart guy. You have chosen to emphasize the Duke/Burres thing, even though it’s a very small component of the argument, because you know it is easier to refute. Even though it’s tangential, you have chosen to hammer on it because you think that by emphasizing it, you can invalidate the entire argument:
http://www.youtube.com/watch?v=eW87GRmunMY
Anyway, to the actual argument (your second paragraph), I would suggest that we continue that down below. It seems a large portion of your conviction that Duke is an average pitcher is predicated on xFIP, while I believe that Duke is the archetypal example of why xFIP needs to be more complete. But we’re discussing it at the bottom, so…
I wasn't the one who chose to emphasize Burres.
Adam was the first one who pointed it out, and the discussion kind of grew naturally from there. He was right to focus on that, too, because it’s not really a rational or defensible position. In and of itself, it calls your entire analysis into question. It’s like suggesting that we have Richard Simmons compete for a rotation spot.
My opinion is the same with or without Burres
I am higher on him than most posters (certainly you and Adam), but even if I did think Burres was Richard Simmons, it wouldn’t change my position on Duke. Your implication that holding a minority opinion on Burres makes me too much of an idiot to debate this, then your minority opinion on Duke also excuses you from the debate. So that’s a non-starter.
The problem isn't that your opinion on Burres is a minority opinion.
The problem is that you haven’t provided any actual reasons for holding that minority opinion about Burres. Whether you accept the points of statistical analysis that I have made about Duke or not, I’ve done my best to “show my work” and lead you through my logic. I’ve provided numbers and links to outside analyses. But all that you’ve offered in favor of Burres is the contention that “whenever I’ve watched him, he hasn’t pitched half-bad,” which a neutral observer can hardly be expected to accept at face value.
I’m willing to listen to your idea that Burres might be better than I’m giving him credit for being. But you need to show your work if you expect people to find your argument persuasive.
Your work is simply that Duke has a relatively bad xFIP
instead of a horrible one (like Burres, I acknowledge). But I don’t even know really applies in his case, since xFIP doesn’t take into account that hitters tee off of Duke with ease and regularly. In fact, it actively tempers that by normalizing the number of meatballs he gives up to the mean.
I have no more desire to debate Burres. Really, I don’t. On the list of the issues affecting this team, I place Burres’ being on the team next year at about #348. Your insistence that we keep hammering at him shows how fragile your argument that we should keep Duke is.
My position...
…is that hitters are not, in fact, “tee[ing] off of Duke with ease and regularly”, in that any such teeing would lead to an increased line drive rate that’s not present. As such, xFIP doesn’t need to “take it into account”, because it doesn’t exist.
And what am I supposed to conclude from that?
Melchior didn’t even consider team defense as a possible source of the inflated BABIP, a notable omission given the generally low quality of our defenses in those years. (Though it makes some sense, I guess, given that he’s a fantasy baseball writer and not a proper analyst – he’s concerned more about whether Duke’s performance is likely to continue than whether it’s his “fault” or not). And if Melchior thinks that Duke’s flyball BABIP is being driven by his IFF rate, he needs to explain why Duke’s mostly-successful 2009 coincides with the lowest infield flyball rate of Duke’s career.
He doesn't say "IFF"
I guess you mean that IFF==popup (don’t know that Melchior means that), but his point is that Duke has always had a high fly ball BABIP.
If all flyballs are equal, and only IFF and LD tell us anything about the pitcher’s skill, why is Duke’s FB BABIP high?
Or is it not high? I can’t find a league-wide listing, so I picked a couple guys – the similar, but better-seeming Paul Maholm and, just for shits and giggles, Brian Burres. Sure enough, Duke’s career BABIP on FBs is .207 to Maholm’s .173 (!) and Burres’ .154 (!!). Care to explain how Duke and Maholm, playing in front of essentially identical defenses, have such divergent FB BABIPs? Has Duke simply been unlucky on 35 extra FBs out of 1018? And truly unlucky – no blaming Bay or McLouth on this one, because they seem to be able to catch Maholm’s FBs just fine.
'Care to explain how Duke and Maholm, playing in front of essentially identical defenses, have such divergent FB BABIPs?"
All pitchers will tend to have different ball-in-play distributions. Perhaps Duke’s particular balls in play happened to be distributed in a way that was particularly disadvantageous, even within the context of our acknowledged-to-be-subpar defense.
You could potentially verify this by manually going through Duke’s BIP log, though that would seem like an exercise in tedium.
Where did you find your FB BABIP numbers, if you don’t mind me asking?
But you said below
that where balls in play go is not a skill…you would think, then, that the LOLN would basically take care of any discrepancies.
Yeah, I don't buy it for a second
Pitchers have no control over balls in play, it’s all defense, everyone regresses to the same mean, but in a purely flukey happenstance, Duke has been unlucky over 1000 FBs in a way that the more effective Maholm has not been. Uh huh.
You're always going to have a few pure outliers.
For it to be conclusive, we’d want to at least try and rule the possibility out, right?
The other major potential skewing factors that I can come up with would be disparate average park dimensions on starts (if one pitcher happened to pitch a greater-than-expected number of starts in disadvantageous environments) or disparate defensive support (if one pitcher happened to pitch a greater-than-expected number of starts in front of backups on the one day a week that they get a start).
Both of those should be verifiable, if true – I’ll take a look.
"Not a skill for pitchers"...
…isn’t necessarily the same thing as “evenly distributed”. Flipping a coin isn’t a “skill”, but if two guys flipped a coin a hundred times, the lack of “skill” involved wouldn’t necessarily mean that they’d end up with the same number of heads.
And while I don’t think a hitter can necessarily shoot the gaps between defenders with any reliability, he can certainly try to pull the ball or go the opposite way with it. So if a pitcher happens to start games against a RHB-heavy, pull-happy lineup, his LF, 3B, and SS are going to see a lot more work than his 2B, 1B, and RF, and any skill disparaties between the first group and the second will be either magnified or minimized (depending on the vector of the disparity).
All of that makes sense
I just don’t see how it’s a sustained difference over a 5 year career in which the two are basically facing the same opponents in roughly similar proportions, with (obviously) the same home park.
I wouldn’t necessarily expect tight regression, as with HR/FB, but it seems to me that it’s either a skill or it’s random, and if it’s random there should be more overlap in year-to-year numbers (it would be weird for one coin-flipper to get 10-20% more heads than the other out of every 20 flips, as well as 15% more at the end).
It’s possible it’s just SSS, but it’s a bigger sample than HR/FB.
B-R
Here.
Year by year for Duke: .133, .214, .292, .244, .161, .201. So every year he hasn’t sucked, it’s been a very high number – but it’s never been as low as his rookie year. It’s a noisy number, but it doesn’t seem to be utterly random.
Thanks for the link.
Looks pretty high-variance, at a minimum. If a hitter batted .292 one year and .161 two years later, what would you assume to be his true-talent level of BA ability?
From your link, it looks like league BABIP on FB is usually in the neighborhood of .140. Interesting… I’ll have to look at this a little more closely.
This is basically my plan.
Might as well cut Duke and use his rotation spot to evaluate Resop as a starter, or to give Morton or Lincoln a second crack at holding down a rotation spot. Karstens has paid his dues, but I’d much rather see him in a long-relief role than a starting job. Unlike Adam Reynolds, my last live experience with Burres was a very good one; however, I have no desire to see him remain a Pirate. Send him to Indianapolis, Cleveland, Siberia; I don’t care which.
Hey, an out is an out - unless you're Mario, in which case it's probably two outs. -UtesFan89
Hard work always beats talent if talent doesn't work hard.
I agree that Lincoln could get a second shot, but Resop and Morton are around the same age as Duke and are worse pitchers, so I don’t see the point in going with them over ZD.
by Adam Reynolds on Sep 27, 2010 11:23 PM EDT up reply actions
I wonder
Whether it’s $2 million, $10 million or something higher than that, we’ve got to make sure we feel comfortable that this guy is, ideally, a multiple-year fit or, if it’s a shorter-term fit, he fills a hole that we can’t fill internally.
Viva Clemente!
That's what NH went on to say
Someone who’s getting too expensive for their current team might be available for prospects. Lots of people have financial issues right now.
Viva Clemente!
Shields has a problem with HR...
…I think he’s given up around 35 this year (check of Baseball Reference confirms 34).
He has a 5+ ERA, WHIP over 1.4 and a ERA+ of 78 this year. Not sure if he’s going to take too much to pry away from Tampa, especially if they are in payroll-slashing mode like they claim.
Garza, OTOH, has a sub-4 ERA, WHIP of 1.27 and a ERA+ of 101. He’s also younger than Shields, so he would cost more but would probably be worth far more in the long run to the Pirates.
Shields is better than his ERA this year
He’s been a bit unlucky with HRs and BABIP this year, but his peripherals are quite good. And I’m sure moving from the AL East to the NL Central will help keep those HRs under control.
Shields has been a very good pitcher this year. Lots of strikeouts and the same solid ground ball/fly ball rates he’s had his entire career. His BABIP is unsustainable, and he’s been wayyyy better than his ERA. He’s better than Garza.
by Suffering Buc on Sep 27, 2010 3:49 PM EDT up reply actions
Rich Harden, Ben Sheet, Brad Penny
Just sign one of them, it won’t cost more than $5M
From what I understand...
Sheets will be lucky if he pitches in 2011.
Certainly won't cost more than $5M to sign Sheets...
…given that he’s out for 2011 after tearing his flexor tendon, and may never pitch again.
He's just holding out
for $6 mil. “I have an owie, I may never pitch again!”
Yeah, yeah, Brett Favre, we’re on to that game.
May not be ready for start of 2012 IIRC
they went ahead and threw in TJ surgery while they had him opened up. Maybe its magnified by whiny fans, but supposedly he isn’t a big workout warrior and if he does make it back, that’s sort of what he needs to become.
We talked about this
on the way to the ballpark. Goldneck asked me how I’d spend $6 mil if I had it, and I said that after I paid off my mortgage and my girl’s student loans I’d spend it on the bullpen and do exactly what brakeman suggested in the first comment.
We’re just not going to get great pitching at a bargain price unless we develop it ourselves or steal it, like McDonald. And how did we steal him? By flipping an overrated reliever. Overrated by Coletti, at least. And it only takes one guy who likes your overrated reliever to make a deal.
Should we consider Meek and Hanrahan overrated relievers, if we can get a good starter by giving one of them up??
Absolutely.
Speaking of Meek, Pence hit a ball off him that may still be going. I didn’t look on that particular pitch, but on the previous pitches Pence was way choked up on the bat, I mean like six inches. If he hit that homer with 24 inches of bat, I’m very impressed.
Pence's power is unreal
Considering that he generates it with about 80% of the bat. Just insane.
We were talking about it a little on his home run at-bat.
I was wondering why he was choked up at all, then guessed that maybe with his team down 5-1 and a runner on, he felt it more important just to try to poke a hit somewhere and keep the inning going.
He poked it, all right. But I don’t think I’ve ever seen a power hitter choke up, period, much less as much as he did.
Remember, this is the same guy with the sidearm outfield throw
It’s pretty obvious that he was just taught to play the game a little differently.
Young Hank Aaron did it a little bit.
He batted cross-handed when he was a kid, and I think he choked up as an intermediate step when changing his form to the “correct” grip.
More common in the olden days as well, of course:
Six-time HR champ “Gavvy” Cravath (‘13-’15, ‘17-’19):
Two-time HR champ Sam Thompson (MLB’s active leader in career HR from 1898-1906):
Made more sense as a strategy back then, of course, given the dimensions of the bats players used. Modern bats are significantly lighter, and have much more of the bat’s mass concentrated in the barrel.
Naw, that's just the scale.
Thompson was a giant in his era. Listed at 6’2", 207 in the 1890s. That was HUGE for the times.
Also...
…for those people who think that there’s “too much offense” in modern baseball, and that it distorts the game, Thompson’s batting line in his career year, 1894:
.415/.465/.696, with 147 RBI in 102 games played. Only 13 HR (still fifth in the league), but 32 doubles and 28 triples.
His team (the Phillies) averaged 9.2 runs per game.
Ott had a very interesting swing
You can get a hint of it from that first picture. Big step from his lead leg (in height, not just distance), like he was trying to get over a short wall or a mid-sized dog.
IIRC,
wasn’t Sadaharu Oh’s stride modeled after Ott’s?
Free your ass and your mind will follow.
by cocktailsfor2 on Sep 28, 2010 10:28 PM EDT up reply actions
I hadn't heard that before...
…but visually, there are certainly some similarities there:
Nice catch. Ott’s torso wasn’t as straight up-and-down, and the bat position is of course different, but they’ve got the same high step and aggressive weight transfer.
I love NH's candor...
but it’s comments like this that lead to people saying he is not a people person:
“We expected more,” general manager Neal Huntington said. “Andy is an example of why you can’t count on just that one prospect (at any position). They don’t always make it.”
Wow, STF already, the guy is still on the team. And yeah he may be having a down year but he is still young and he may be productive somewhere else. Tim at PirateProspects has a great article up right now on the lessons we should learn from Batista. I for one would rather see Laroche on the bench than Vasquez.
http://www.piratesprospects.com/2010/09/what-can-the-pirates-learn-from-jose-bautista.html
Interesting piece.
The only thing I kinda disagree with is his suggestion that 52 homers isn’t “legit.” First off all, as I never tire of pointing out, he didn’t just starting blasting homers this year, he hit 10 in September ‘09, and eight after Sept. 15, when it looks like somebody flipped a switch. So he’s sustained close to a 60-homer pace for more than a year. Also, iirc somebody here noted that he’s not hitting lucky homers, he’s smoking them. It looks like he learned something.
If he falls off next year, and it’s hard to imagine he won’t, it could just as much be that pitchers adjusted and started working around him, as much as it would be that this season was a “fluke.” But then, it might be hard to work around anybody in that lineup, cause they can all kill you.
I know about Brady Anderson, that’s still an astonishing and apparently “fluke” season, but I don’t know that a single example (talk about SSS) is relevant here.
No doubt that is true.
How “little” is the question.
And dammit, now we have to wait six months to find out. I took a big gulp of ballpark air yesterday and I’m holding my breath until April. The only better smell, imo, is amusement park air (coaster grease, french fries, cotton candy and sunscreen).
They both smell exactly like summer.
Cut grass is a pretty good smell too.
I don't get this at all
We all know Andy is having a shit year. We all know that they are not pleased with his performance.
But this type of stuff officially devalues him in everyone’s eyes. Why not say something like “Despite Andy’s struggles, our coaching staff has been working on his approach and we believe he will turn things around in 2011.”
Sure, it’s b.s., and it’s probably not going to make a difference, but there’s no possible benefit from devaluing your own assets. Now if some GM has his eye on Andy as a utility option, we’ve got no bargaining ground.
Service time
Wow, I didn’t realize how badly that had been handled. It’s no surprise that he was traded for a handful of non-magic beans, given the disparity between service time and actual, functional development time. 2008 was a regression compared to 2007, but he was certainly worth keeping for the league minimum.
Thanks again, Dave Littlefield, you incompetent fuck.
DL
Seriously, if DL had protected him, when an normal individual would have, JB would have likely had a different career trajectory.
He bounced from team to team to team until the Pirates finally got him back in the Benson deal.
But that wasted year started his clock when he wasn’t ready.
He needed the at bats and fielding work.
I wanted LaRoche to succeed as much as anyone, but he got over 1000 PA on the Pirates, and hit more like Bobby Hill than Jose Bautista. It is what it is, really.
by Adam Reynolds on Sep 27, 2010 12:36 PM EDT up reply actions
Cheap, young, plays D. He’s still better to have than Ramon Vasquez. If he continues to suck and can’t backup 2B adequately, we can bring up Josh Harrison or Jim Negrych to be cheaper and younger although possibly just as sucky.
My biggest worry
about keeping LaRoche is that he doesn’t seem capable of hitting off the bench. SSS and all that, but there’s certainly a skill to pinch hitting, and it’s a skill he seems to lack. But hey, 2011 is a lost cause regardless, and having Andy LaRoche as a bench bat is not going to make a visible difference. But it could pay dividends – if nothing else, I like him better as a long term sub (Clemente forbid) at 2B or 3B than any plausible veteran FA.
Pinch-hitting ability is highly correlated to how one hits in a starting role. If a player can’t hit as an everyday player, it shouldn’t be a surprise if he cannot produce in a smaller role.
Who should replace LaRoche’s role if he’s let go? Literally anyone but Delwyn please.
by Adam Reynolds on Sep 27, 2010 2:26 PM EDT up reply actions
But he's even worse off the bench
is my point.
The average MLBer is worse off the bench than as a starter (it’s hard to test comprehensively, since there are few players with significant samples as both); a few are about equal, and those are the guys who can actually make a living as a PH (DY being, like it or not, a very good example – tOPS around 100 as a starter and as a PH). Andy has shown no ability whatsoever to hit off the bench – his career tOPS is 100; as a PH, it’s 40. SSS and all that, but there’s really no particular reason to believe that a guy with that much FAIL under his belt will suddenly learn the secret to pinch hitting.
Like you said, there aren’t enough examples to really see a difference between hitting off the bench and a regular role. Plus, the sample size is too small. I’d say there’s no real difference because hitting is hitting. The reason pinch hitters do worse is b/c they’re fringe major leaguers (unless you have a Jim Thome or whatever).
by Adam Reynolds on Sep 27, 2010 4:40 PM EDT up reply actions
Disagree
I actually spent some time looking at this last year – if you look at league PH totals year to year, and then look at the best PHs, you see guys who consistently outperform the league, and who hit about as well off the bench as when they start. And it’s very common to see guys like Andy, who PH much worse than they hit as starters (Nunez was another one – career PH tOPS+ – in 300 PAs, mind you – of 32 [career tOPS+ for any given player is 100 by definition]).
There’s very obvious reasons that pinch hitting would be harder than batting as a starter, and it’s not at all surprising or counterintuitive that some players would be better at doing it than others (just as some players are streakier than others, or even that guys with similar physical profiles would have very different power outputs).
PS – Pondering ex-Bucs who were used as subs, I thought of Joey Bats: a tOPS+ of 32! as a PH. That’s worse than at any other position, including his 44 at 2B. Point being, those are all SSS, but somehow he was worse coming off the bench even than as a dude being stuck at a position he had no business fielding.
Who are the best current PHs who outperform the league year after year? Obviously our best example of a PH is DY, and he has a .594 OPS in the role this season, so he can’t be one.
I can’t get past the sample size issue here. If even Andy LaRoche got another 250 PH AB, he’d be around that .650 area that he’s been as a starter for his career.
by Adam Reynolds on Sep 27, 2010 5:18 PM EDT up reply actions
Lenny Harris is regarded as one of the best PH of the modern era.
He was given spots on 40-man rosters for about five years after he stopped being able to play in the field, purely on the strength of his rep as a PH.
For his ML career as a PH, he hit .264/.317/.337 in 883 PA.
For his ML career as a whole, he hit .269/.318/.349 in 4289 PA.
I can buy the idea of players not being able to PH, for whatever reason, but in general I think the value of a player being a “good PH” is wildly overrated.
It's not a big deal
Looking strictly at hits, DY vs. a league-average PH gets you an extra hit every 20 ABs (although he’s got a nice AVG/OBP jump that’s probably above league PH performance as well).
But, as I said, my concern is that Andy is way, way worse than a league-average PH: his AVG as a PH is 75 points worse than his career 3B AVG (which, incidentally, is only a bit above league-average PH AVG). There’s “this isn’t a critical skill” and there’s “this person is utterly incapable of performing this task.” I’d say that an OPS of .451 approaches the latter.
If I thought Andy was in a situation where he’d be getting 50+ starts a year, I wouldn’t care at all about his PH abilities. But the only way he sees that many starts is if a far better player gets hurt (or, I suppose, if NW turns into a pumpkin, of which there are no signs), and that’s kind of irrelevant. You can’t efficiently build a roster around “what happens if our best player gets hurt?” because no one has 3+ WAR guys to spare.
So the question returns to, “is this guy valuable in the role for which we need him?”, and that role is 75-100 PH PAs and a dozen or two spot starts. As I said elsewhere, I’m OK with the idea of Andy over random vet guy, but I worry that those 75 PAs will equal ten hits and a couple walks, when DY (or someone else) would turn them into 20 hits and 6 walks. Also, too, if Andy continues to be a zero off the bench, JR will continue not to use him, and he becomes a waste of a roster spot.
Why would you look at one season?
His career PH OPS is .695. Which isn’t awesome, except that all MLB PHs have batted .221 this year (B-R doesn’t make it easy, if it’s even possible, to see league-wide PH OPS. But I think we can agree that it’s unlikely that the OBP and SLG exist to bring league PHs anywhere near .695 starting from .221).
MLB as PH, 2010: .221/.297/.348 (go here, then scroll down to the “Defensive Positions” tab)
2009: .225/.315/.353
2008: .232/.318/.350
2007: .226/.309/.351
Ah, thanks
I was searching by League Stats.
So DY is about 25 points of OPS over league average as a PH (weirdly big drop this year – you think it’s all Andy’s fault?).
5600 ABs is a small SSS?
Per this article, OBP and SLG should both stabilize (for a given hitter) at 500 PAs.
Or are league-wide numbers so different that they never stabilize?
But I didn't mean DY
I put the parentheses in a confusing place. But what I meant was that the league had dropped almost 25 points of OPS from ‘09 to ’10, after year-to-year variations of +8 and +0. I wonder if it’ll jump right back up next year.
No question DY’s year-to-year change is random variation, which is why I’m only looking at his career PH numbers, which are significantly better than the 4 year running league average, or the league average for any year that he’s played.
I know ...
I was just saying that I’d feel comfortable with a lower number, probably around 400.
Obviously, 500 gives you more statistical power. But it’s not a magical number.
In other words, there’s really nothing that can be learned from the Bautista situation. Out of 1,000 LaRoches, Clements, and Mosses, maybe a small handful find success after flaming out. Out of all the hitters similar to Garrett Jones’ pre-09 profile, only a small fraction will do anything in the big leagues.
Unlike those guys, Bautista was a solid hitter a bit below the MLB average. What he’s done, though, probably won’t happen for another 15-20 years.
by Adam Reynolds on Sep 27, 2010 1:03 PM EDT up reply actions
Pitchers
Getting what we already have is surely not the answer, if it takes trades for better pitching, then that is the way I would go, Huntington said yesterday that this is a very bad FA pool this year. I have no idea who is available on the trade market. But they do need starting and bullpen pitchers and they do have a lot of deadwood to move.
I may be in the minority, but i think we need to spend for an upgrade to the offense/defense more than we need on a SP.
+1
Makes no sense upgrading the pitching if we don’t improve the defense. That is unless NH is planning on bringing in some big K pitchers. Not too many of those available though. If you look at our starting lineup UZR’s, everyone is a negative defender but Tabata and Cedeno. Now I know the shortcomings of UZR and that some of you don’t like the stat but the numbers back up what I have seen with my eyes this year at well: McCutchen (good speed, bad routes, inconsistent arm), Jones (bad range, bad glove, bad arm, nice picks at 1st), Pedro (bad range, inconsistent, plus arm), Walker (athletic, makes plays right at him, currently lacks range). Woops, forgot Milledge, he is barely a positive defender but I suspect he will be a negative the next time the UZR’s are updated. He has struggled in RF.
Pedro's arm
Apropos of nothing, but I was right on line between PA and 1B yesterday, and man, do his throws have a nice trajectory – as level as physics will allow, and just a bit of (from the 1B’s POV) trail to the right. The kind of stuff that a little leaguer could catch on a regular basis.
That's good to hear.
I can’t wait to see him in person too but it looks like I have to wait til next year. Oh well, gives me something to look forward to.
Defense
The defense will develop with experience. Positioning is 1/2 of the Pirates problems defensively, if you put yourself where you should be, you don’t have to be a hall of famer to be a quality defensive player. I think it was Tanner that said, just catch the balls your supposed to catch, I’ll take that. Tabata is not a negative defender, but he has more problems to overcome than McCutchen has defensively. Gorkies is the best defensive outfielder in the system, but he won’t hit 300.
I think positioning is some of it yes...
at least with Cutch I think it has been a big factor. I too think McCutchen will improve. His tools are too good not to at least be league average, I hope for more though.
Tabata is not a negative defender, but he has more problems to overcome than McCutchen has defensively.
How so?
Problems
Tabata is slow mechanically with the ball, he also does a very poor job around the fence, he gives up on far too many balls. McCutchen gets some bad jumps, but that takes time in major league parks because of backgrounds and experience with hitters. I think Tabata is pretty much going to stay the way he is unless they work with him and I think McCuchen will continue to improve.
Positioning can also be useless
when your pitcher misses his spot by 6 inches and the hitter goes with the pitch, knocking the grounder through the hole.
Go after Jeff Niemann..
..just makes too much sense to me, which is probably why we won’t.
4.41 xFIP, makes about $1 million a year, is 27 and isn’t eligible for FA until 2015.
The Rays aren’t dumb, they aren’t going to unload Garza/Shields for the hell of it, but Niemann, he seems expendable. Getting out of the AL East could really turn Niemann into a solid option for us.
_________
Thoughts on this as a way of obtaining Niemann: Trade Ryan Doumit (Covering about 3 million of the 5.1 million he’s due in 2011). That essentially would be paying $4 million for Niemann (3 plus the 1 I believe he’s due to make), which is cheaper and probably an upgrade to what 2011 Duke would bring.
Why would Tampa want Doumit? His .OPS for the season, believe it or not is only 10 points worse than John Jaso’s. Jaso’s unproven and Doumit would, IMO, give them some peace of mind at the catcher position next season. Shoppach’s due 3 million next season, which is the only barrier I could see to this trade, but he has been a disaster at the plate. Even so, under my scenario, that’s only paying about 5.6 million to the catcher position for 2010. If they plan on contending next season, if Jaso struggles they’ll be in a world of trouble. In addition, we saw how much better Doumit played when he didn’t have to man the catcher position every single night. With Pena and Crawford most likely gone, they’ll be able to get Doumit some at-bats in various positions, much like we were planning on doing. If Doumit busts out in Tampa, he’s well worth his options in ‘12 and ’13, if not then they only lost a pitcher in Niemann they didn’t have room for anyways.
Si? No? Crazy?
i think the whole reson for Tampa considering movinf Shields or Garza is $$$.
by BuccoBrigade on Sep 27, 2010 2:11 PM EDT up reply actions
Niemann is pretty risky
He’s a guy with a history of serious injuries and he’s pitched like crap since he came back from the disabled list at the end of August. I would stay away from him.
by maguro on Sep 27, 2010 2:17 PM EDT up reply actions 1 recs
the only players we can trade for are
either underperforming former prospects or injured ones so why stop now? If we could get him with minimal damage to our minor league system I don’t see why its a something we shouldn’t ever look at doing.
by eyeofhorus777 on Sep 27, 2010 4:57 PM EDT up reply actions
If we could get him with minimal damage to our minor league system...
We can’t. That’s the point. The only way Tampa would let him go for a “minimal return” is if he’s due for arm surgery. He’s too valuable for them to just toss otherwise.
I can't describe how much I don't want
Millwood or Wellemeyer.
In free agency, I think the best option would be Westbrook, but as has been noted, he may have priced himself out of our range down the stretch. De la Rosa is the other guy I’d look at, but I think he’s ripe for huge overpayment. Huntington can’t make that mistake.
That’s why I go back to a trade. Tampa and Texas both have extra starters. I want one or two of them, and I’d be willing to part with pretty much anyone not named McCutchen, Alvarez, or Tabata to get one. Even Walker could be in play, though that would be for a real frontline starter.
Shields, Garza, Niemann, Feldman, Harrison, and Holland strike me as guys who could definitely be traded for by a reasonable price.
Texas loves Holland
They are really trying to protect him. The really need a catcher so maybe they would look at Doumit or Snyder but right now I can’t see them just offing Holland for some other prospects unless it really filled a need.
by eyeofhorus777 on Sep 27, 2010 3:52 PM EDT up reply actions
Holland
They were unlikely to deal him for Cliff Lee.
I really doubt they’ll trade him for any combination of Doumit or Snyder.
Unless, of course, we throw in Pedro Alvarez.
I’m playing a numbers game. I’m not saying he’ll come cheap.
They have Wilson, Lewis, and Hunter pretty much locked in for next year. Then there’s Scheppers simmering and ready to go, though I’m not sure if he’s going to start or relieve. I’d guess he’ll eventually start since they already have an awesome bullpen.
If they do well in the playoffs on Lee’s back, they’ll do all they can to keep him. Then there’s Martin Perez and a whole assortment of very close to top tier pitching prospects. They have young guys in their bullpen capable of being stretched out to start, particularly Ogando and (much less likely) Feliz.
Someone’s going to be at least available, I’d have to think.
by Suffering Buc on Sep 27, 2010 4:01 PM EDT up reply actions
They may have to trade a guy ...
I just don’t see them wanting any of our guys except Alvarez, Walker,Tabata.
Holland would be pretty expensive.
I think other teams probably can put together a much better offer.
Yeah its tough to admit
but we have very few pieces to trade at all without killing our minors.
by eyeofhorus777 on Sep 27, 2010 4:02 PM EDT up reply actions
Yes and no
If the seller is motivated (basically by $$$), then our AA arms would be pretty valuable. None of them are good enough to pry away a young #3 SP ceteris paribus, but if that pitcher’s about to be due high 7 figures for a club with good options, and we throw in useful and/or interesting guys like Doumit or Milledge (don’t snort), there might be some options.
The big question is why we’d trade the future for pieces in 2011. Barring a real opportunity, I just don’t see the value in even a 4 WAR SP for ’11. 91 losses instead of 95. Big deal.
Not only is Westbrook likely to be expensive...
…he’s also a guy with the exact same profile as Duke. Contact-oriented pitchers are going to get killed by our defense. We either need to go after pitchers who don’t put many balls in play or drastically improve the quality of our gloves.
How many open SP spots do we have?
From my perspective we have 3 solid starters right now in
1. Macdonald
2. Ohlendorf
3. Morton
You may scoff at Morton but he has pitched ok since being called back. Nobody is in the MiLB that will push him and he has the most upside of any pitcher we have. I see them not giving him a shot next season, baring an awful ST.
That leaves 2 guys out. I think Maholm comes back even at the high price tag to at least start the season.
Now we are down to 1 spot?
I would take a gamble on a player making a comeback or returing to form. What’s the worse that can happen? We have a player as bad as Duke was this year, which was pretty bad. Brad Penny would fit the bill.
If Penny or mystery pitcher X flames out we have the usual suspects like Karstens or Lincoln to run out there until someone from the minors is ready. So whats the real big issue with non tendering Duke?
Morton is the Key here but we are going to be terrible no matter what so why not take a few chances.
If we have 2 to 3 injuries well no amount of FA spending will fix that.
I'm assuming Morton is going to be there as well.
That’s unless he takes another huge step backwards tonight and in his last start. If he can keep turning the corner, he’ll probably get another chance.
I also lock in Maholm for next year. His value is at its lowest point right now, and he’s the only shred of consistency on the staff. He’s had a rough year, but he’s an average starting pitcher in the NL. His career ERA+ is 95. That’s not good, obviously, but it’s very close to exactly average. Don’t trade him coming off a 5 ERA year where he’s actually pitched a good bit better than his stats.
1. McDonald
2. Maholm
3. Ohlendorf
4. Morton
5. Free Agent/Trade Pickup
That’s what I expect to see next year. Hopefully, Maholm pitches well and can be traded for something decent when Owens, Morris, and Locke demand rotation spots.
by Suffering Buc on Sep 27, 2010 3:55 PM EDT up reply actions
Morton has made progress. His strikeout rate has gotten progressively lower each month…
by Adam Reynolds on Sep 27, 2010 4:15 PM EDT up reply actions
Morton ...
is going to get a scholarship because he can throw 95 mph.
He’s not as bad as his record indicates.
And I wish he’d quit toying around with his breaking balls.
But it’s not like the Bucs are overloaded with outstanding arms.
Actually, you throw in Duke and Maholm and that’s the rotation, I think, the Bucs will use until the Altoona guys hit the big leagues.
Possible Free Agent or Trade Options
Here’s a useful link to both free agent and other alternatives.
Not the world’s most impressive list.
Viva Clemente!
Duke also is going to make more than 6 mil
cannot break an egg and has already reached his ceiling. Its not that much about winning next year but finding out if Morton is going to be around when we actually can compete.
Morton has tools that Duke will never have now we need to see if he can make any use out of them.
by eyeofhorus777 on Sep 27, 2010 4:14 PM EDT up reply actions
Morton is around 8 months older than Duke, so it’s not like he has some huge comparable ceiling at this point. 95 or not, he can’t miss bats and gets hit even harder than Duke.
by Adam Reynolds on Sep 27, 2010 4:18 PM EDT up reply actions
Duke has much more ML experience
we know what we have with him. I am willing to give up on Duke who is a terrible fit for this team and has pitched terrible this year. Defense or not he seems to give up alot of hard hit balls as well.
Morton could be an Ace, Duke could be a 4th starter on a good team with good defense at this point.
by eyeofhorus777 on Sep 27, 2010 4:32 PM EDT up reply actions
Morton can’t be an “ace”, that’s insane. Morton could be a 5th starter, Duke is already a 4th starter over his general career.
by Adam Reynolds on Sep 27, 2010 4:37 PM EDT up reply actions
And his career is in decline so why hang onto him for more money
Obviously the team isn’t blind to the D issues on the team but even at that they think he isn’t worth the money either.
.
by eyeofhorus777 on Sep 27, 2010 5:02 PM EDT up reply actions
Duke isn’t worth the money if only compared to a James McDonald or Ohlendorf.
He is, however, worth the money compared to:
*sub-replacement Burres types
*Paying the FA price ($8-10 million) for someone just as good
*Getting someone better (like a a Ted Lilly or Kuroda), but having to pay around 3 years, $42 million or 4/$48, etc.
by Adam Reynolds on Sep 27, 2010 5:27 PM EDT up reply actions
This.
I was going to make this argument but you did it for me. Besides, with our payroll why is everyone getting so upset with what Duke will make. I mean $6M, big deal. Now if the plan was to take that $6M and add another $4M – $6M to it they may be able to upgrade enough that our defense wouldn’t offset the acquisition.
The multi-year deal for a mid-rotation starter doesn’t make sense given what’s close to the majors, with Bryan Morris, Jeff Locke, Rudy Owens ready to go next year, and potentially Justin Wilson and Brad Lincoln in the mix as well.
by Adam Reynolds on Sep 27, 2010 6:04 PM EDT up reply actions
Morton
Not sure if I think he can be an ace.
But the guy does have a special arm.
He’ll be given at least one more chance in Pittsburgh. And, if he fails that, chances by other teams.
It’s not like teams have a bunch of guys who can hit 95 in their minor league systems.
Duke's "ceiling" was pretty good.
4.06 ERA the one year we put a half-decent defense in front of him. And he was in the 3s that year until we unloaded Jack and made Delwyn Young an everyday 2B.
I don't think there is a defense for balls hit off walls and over them
That has been Dukes problem this year and for most of his career. Not being able to strike anyone out is also an issue, not just with teams with bad defenses but in general, a good pitcher that can strike out a hitter when runners are on is much more valuable than one that relies on contact.
Duke is a junk pitcher who needs a very good team behind him to be average at best;
by eyeofhorus777 on Sep 27, 2010 4:35 PM EDT up reply actions
Also we are supposed to think a 4.06 ERA is awesome?
Have our expectations become that low when we are arguing over whether its ok to let someone go who if everything pans out great for them and our D improves is a +4 ERA?????
by eyeofhorus777 on Sep 27, 2010 4:37 PM EDT up reply actions
What do you mean about "low expectations"?
Within the modern context, a 4.06 ERA is very good. League average ERA in 2009 was 4.19.
And of course, even a 4.06 understates his ability somewhat. Check out what the defensive downgrade did to his performance that year:
Prior to Jack and Freddy being traded (and replaced in the lineup with Young and Cedeno), Duke had a 3.42 ERA in 139 1/3 IP (with a BABIP of .285). From that time on, he had a 5.25 ERA in 73 2/3 IP (with a BABIP of .328).
It really is that simple.
That was also last year
he’s lost 1-3mph on every pitch, so it’s hard to believe he has the same upside.
ORLY?
Duke’s average FB, career: 88.1
Duke’s average FB, 2010: 87.5
Yeah, that half a MPH makes all the difference in the world.
Not from his baseline, he hasn't.
And the fastball speed is what determines the speed of other pitches – since they’re used as a change-of-pace from it. So if the fastball goes down a tiny bit, the pitcher’s other pitches SHOULD go down along with it, to maintain that separation.
What?
How do you mean “SHOULD”? Should, in the platonic world of effective pitching, or should, because pitchers have that kind of precise control? Because when I’ve looked at specific pitchers and their game-to-game speed variation (and within game as well), the FB changes but the changeup doesn’t. Specifically (because I know the numbers without looking), when McDonald goes from 93-95 in the early innings to 91-92 in the late innings, his changeup stays exactly at (iirc) 78-80. That’s part of why he’s less effective late in games – less separation.
Since FB speed is, for most pitchers, established by physical limits, but CU speed is established by a combination of grip and (ideally slight) mechanical variation, there’s no magical correlation between the two speeds. If pitchers could just will their relative speeds, then you wouldn’t see guys with changes that are too close in speed to their FBs to be effective pitches.
In the case of Duke, I have no idea whether his change has dropped along with his FB – it’s certainly possible. But there’s no automatic mechanism for ensuring it, and it would be a very plausible candidate for problems for a guy losing more significant velocity (I agree that half an MPH almost surely makes no difference in effectiveness, although it can’t possibly help).
Should, in the platonic world of effective pitching
This. You want some separation in velocity between your fastball and your secondary pitches.
Duke’s changeup this year is down fractionally from his career average: 81.2 career, 81.1 in 2010. Of course, since his FB is only down 0.6 from his career average, you wouldn’t expect the change to drop too much, either.
Its not that simple, really
these guys aren’t robots. Duke is awful sometimes because he has mediocre stuff and can’t locate it properly. YOU HAVE TO SOMETIMES SAY ITS THE PLAYER’S FAULT AS WELL.
If the choice is between keeping Duke or seeing if Morton can turn it around I take Morton in a heartbeat.
by eyeofhorus777 on Sep 27, 2010 6:04 PM EDT up reply actions
Sure, sometimes it is the player's fault.
Just not in this case.
And we aren’t choosing between Duke and Morton. We’re choosing between Duke and Dave Bush, or Duke and Todd Wellemeyer.
if they keep Duke and sign a FA pitcher
Morton will be most likely the odd man out. I believe signing Duke kicks Morton out so yes it is choosing between t he two
by eyeofhorus777 on Sep 27, 2010 9:35 PM EDT up reply actions
If they keep Duke...
…they don’t need to sign a crappy FA pitcher like Bush or Wellemeyer to replace Duke. So they probably won’t.
4.06 is not very good Vlad
come on, or else define very good as C+ from now on.
by eyeofhorus777 on Sep 27, 2010 6:11 PM EDT up reply actions
By modern standards...
…a C+ in school is probably like 25th percentile.
Context is everything. If my little nephew came home with standardized tests saying that he was solidly above average, I’d hardly view that as a disappointment.
Ah, the old "Duke gets hit hard" canard.
Duke’s line drive rate isn’t anything out of the ordinary for a pitcher. He isn’t giving up tons of undefendable balls off the wall.
And while his HR rate is a bit elevated this year, that just looks like random variation. Pitchers regress toward a 10% HR/FB, and he’s at 13.7% right now. He’s just gotten a little unlucky with balls barely going into the stands, rather than being caught at the warning track.
Heh, how many times has this argument been made already?
It’s clear that when fans are fed up with a certain player that it doesn’t matter what evidence is smacking them in the face. Duke is this year’s Adam Laroche. I do believe though that NH will release Duke and I too think it will be a mistake. It also doesn’t sound like there will be any significant changes to the IF defense either since NH plans to find a platoon partner for Garrett Jones at 1B. I fear for Maholm next year and I pray to God that whoever is brought in can strike a few players out and isn’t heavily dependant on the groundball.
Gives up more HR’s than most, gives up more line drives than most, gives up more hits than almost anyone… yep, he doesn’t get hit hard at all.
Yeah its all randomness there is nothing
at all that says Dukes stuff is getting worse or that it will keep trending that way.
by eyeofhorus777 on Sep 27, 2010 6:05 PM EDT up reply actions
No, it's not all randomness.
Duke is in charge of certain outcomes: Walks, strikeouts, and fly balls. Based on the things that Duke can control (once again: walks, strikeouts, and fly balls), he’s pretty much an average starting pitcher. Below average at Ks, better than average at the other two.
The “randomness” comes from the things that Duke can’t control: the defense behind him, the distribution of balls in play, and the distribution of borderline HR/FB drives. He’s gotten unlucky in all of those areas this year.
This should not be hard to understand.
I bet everyone understands the theory
probably most think there is some truth to it. But you adopt it completely and totally as gospel. And FWIW, Duke has shared responsibility for giving up hits.
Sure.
Any additional line drives he gives up beyond league average that turn into hits are on him.
It’s not “gospel”. Gospel is something that you believe in the absence of scientific evidence (“Jesus is Lord”, etc.). This is, if anything, the opposite of gospel. I didn’t believe it until I read the studies and saw the evidence, which then shaped my understanding of events. It’s not like Bill James descended from on high and then handed me two tablets dictated by a burning bush.
If Duke truly were getting hit unusually hard...
…he’d be giving up huge numbers of line drives. Line drives are just about the hardest contact that there is – you only get one when you barrel up on a ball.
But his line drive rate is only a little worse than league average, and better than that of other pitchers who are very successful. It’s lower than most of the other pitchers in our rotation: Morton, McDonald (as a Pirate), Ohlendorf, and Burres. So he’s not giving up tons of hard contact. He’s mostly giving up bleeders and grounders through the hole.
In a comment thread last week, I even went and checked the trajectory on his home runs allowed, since it was suggested that he was giving up lots of cannon shots that went over the wall. And his angle of release on HR allowed was pretty much league average there, too.
Who is distinguishing
between fly balls and “liners” and bleeders and hard grounders? That doesn’t sound like data that would support hard conclusions.
There's a couple stringers at every game
Stats Inc and BIM, iirc. They register separately; I can’t recall what they do in event of disagreement.
Vlad swears by this system (“you can teach a five year old to identify LDs in five minutes”), but DK states that he regularly sees their report pop up and it disagrees with what he just saw (I guess he’s getting live updates in the pressbox). It’s going to be imperfect, but most likely not in a systematic way.
There was a study a year or so ago looking at outcomes in different ballparks, and it appears that there’s a ballpark effect, in which stadia with high pressboxes (inc. PNC) have skewed LD rates because balls look flatter from up above – or something to that effect.
No
Gospel is an accepted truth with or without scientific evidence. I read some of the studies, just like I read a bunch of the studies on “clutch” hitting. They aren’t done by scientists, professional statisticians or done in a scientific manner. Yet they are accepted as truths – gospel.
Your notion about line drives reinforces your religious acceptance of the theories that make up adherence to DIPS.
I think most fans can see Duke is turning quickly into a BP pitcher
His lost velocity isn’t smoke and mirrors. Sometimes there are reasons outside of randomness that causes stats to rise.
The issue is about maybe getting someone who could do better than Duke. If they can than do it. Everyone acts like there are 0 pitchers in FA that will outperform Duke, I say take a chance on someone. It can’t be anyworse than it is right now.
by eyeofhorus777 on Sep 27, 2010 6:10 PM EDT up reply actions
Ah, Duke's "lost velocity"!
He’s down 0.6 MPH on his fastball this year (87.5) from his career average (88.1). Shocking! What will he do without it?
Sure, there are probably some pitchers in FA who would outperform Duke. The problem is that the vast majority of them will be significantly more expensive than him, and there’s no real way of telling the remaining ones from the teeming mass of FA pitchers-at-large.
Just for context:
Based on career averages, Matt Cain has “lost” almost twice as much velocity as Duke this year. His career average is 92.6 MPH, and his 2010 average is 91.5 MPH.
Is Matt Cain also “turning quickly into a BP pitcher”?
Speaking of context
Cain’s FB was 1.6 MPH above MLB average; it’s now 0.5 above. Duke’s was 2.9 MPH below average; it’s now 3.5 below. It would also be useful to see how fast BP pitching really is. But I bet that most hitters would rather see Duke’s FB after a mere 0.6 MPH drop than Chapman’s after a 2 MPH drop – wouldn’t you?
Then there’s movement and command. All of which is to say that absolute speed and small losses of velocity are (often) incidental to effectiveness. But your Cain example is pretty meaningless.
yes that was a terrible example Vlad
Cain even at the velocity drop still has above average control and velocity.
by eyeofhorus777 on Sep 27, 2010 9:46 PM EDT up reply actions
Sure, Cain's FB is still faster than Duke's.
But we were talking about the impact of a change in velocity, not the value of a given level of velocity. I don’t think anybody would argue against a faster fastball being better, all other things being equal.
Having a faster fastball at the start just means that Cain should be starting out at a higher level than Duke. If you’re measuring the impact of a decrease in velocity, it should still be present in any pitcher who experiences that decrease, regardless of how fast he is at the start. In terms of results, he’ll just go from superhuman to really, really good, instead of average to below-average.
Or maybe you meant...
…that it’s not possible for a SP to succeed with an average FB of 87.5 MPH?
Among SP with 100+ IP this year, the five guys immediately under Duke in average FB velocity this year are (in order): Shaun Marcum (87.1 MPH, 3.63 ERA), Jason Vargas (86.8 MPH, 3.91 ERA), Dallas Braden (86.7 MPH, 3.49 ERA), Ted Lilly (86.6 MPH, 3.83 ERA), and Wade LeBlanc (86.6 MPH, 4.25 ERA).
Hmm, how odd. That doesn’t seem to support your point, either.
No.
For the purposes of those numbers, a fastball is a fastball.
It would be nice to have that level of granularity, but the data collection methods aren’t there yet.
Then the data is without value
since its a mix of several pitches of different velocities. Just mixing in more 2-seamers or cut fastballs would rule out using the total as evidence of a velocity change.
It's a good thing you weren't a scientist in the olden days, Pete.
If you were trying to use 15th-century techniques to determine the circumfrence of the earth or the speed of light, you would’ve broken out in hives at the mere thought of all the potential sources of error.
Then and now, you use the best info on hand, and correct when you’re able.
Data collections methods ...
it’s amazing what we have now.
But you are correct. They aren’t there yet.
And we need to be careful when looking at an average fastball.
I don't ever recall making that my argument
strawman strawman strawman strawman
I remember mentioning location as well as sub par stuff.
Stay on focus Vlad
by eyeofhorus777 on Sep 27, 2010 9:45 PM EDT up reply actions
If Duke doesn't have the ability to locate pitches...
…then why is his walk rate still better than league average? And why is he actually throwing a slightly higher percentage of strikes this year than he was last year?
I "switched" to ERA...
…because it was listed in the same column for those guys on the stat table, and I’m tired of roaming around collecting stats for you mutts, when you never provide any actual affirmative evidence of your own.
Seriously? You’re going to make an objection based on that? It isn’t self-evident that Ted Lilly is a successful starter by pretty much whatever method you’d care to use?
Keerist.
Know what else is a "theory"? Gravity.
This year, Duke’s average home run came off the bat at 104.4 MPH, and traveled 405.8. Last year, Duke’s average home run came off the bat at 103.3 MPH, and traveled 403.3.
Not a lot of difference there. The only difference is the results.
10% HR/FB
I’d argue that it is misleading to say that a guy will regress to 10 percent.
That is the case with outliers, a guy at 3 percent or a guy at 30 percent.
It’s the simple statistical concept of regression to the mean.
But there is little evidence that a guy at 7 percent or 13 percent will “regress” to 10 percent. It just doesn’t hold nearly as well when you are one standard deviation above or below a league average.
Also, a huge problem with xFIP. It assumes an elite pitcher will regress to the standards of league average pitchers.
Admittedly extreme example: Bob Gibson in 1968. Let’s assume he was playing at a pitcher-friendly park like Forbes.
He was never going to regress to 10 percent.
Duke players at a ballpark that is reasonably friendly to LHP and still gives up more HRs than he should.
The 10% figure...
…is for the modern context. You’re correct that it probably wouldn’t do that for Bob Gibson. But in the ‘60s, it would still have regressed toward a given number below 10%, since pitchers didn’t have any more control over it then than they do now.
The HR/FB thing has been researched to death. I’m not just “assuming” anything here. You can take a group of (conventional) pitchers that are at 7% this year, and a group of pitchers that are at 13% this year, and then look at them both next year, and both groups will move toward 10%.
I mean, Duke’s career rate is an even 10.0%… and you’re trying to argue that he’s some kind of crazy skill outlier based solely on his 2010?
Question
Not trying to make a point either way in this argument, but it occurred to me after Pedro’s laser-like HR tonight: How are HRs categorized in the FB/LD thing? Point being, we look at HR/FB, but since X% of HRs come off LDs, that suggests that there’s probably something imperfect. 10% of the number of FBs a pitcher allows end up in the seats, but some of those HRs actually came from LDs.
I don’t know that it would matter, but I wonder whether it’s been looked at.
I'm not an expert on the classification ...
but it’s a problem either way.
The classifications assume that they are mutually exclusive.
And, as you show, that can be tricky.
I suspect it’s categorized as a line drive, though.
Depends on the data source.
But if you want to correct for line drive HRs, you can just go through the player’s log at hittrackeronline.com. They not only have gross classifications of HR, but also angle of escape, velocity off the bat, all that good stuff.
It does ...
but JRoth’s point, much like the one you make above, is that data collection methods and even operationalization are less than optimal.
Obviously, it’s amazing how far things have come in 10 years.
But it’s not a science.
Not a science?
I don’t think anybody would dispute the idea that you can always build on past work to make improvements to the process and produce more accurate results
But not a science? Can’t agree with that. Don’t let the perfect be the enemy of the good.
Vlad
Let me be clear on it.
I’d put baseball research closer to a social science than a physical science.
Does it have some validity? Absolutely.
But there are pretty simply concepts that it struggles to measure well, such as fielding.
I don't really see that as a valid comparison.
But to each his own, I guess. It’s not really a point that can be debated.
Vlad
I’m not commenting in any way about Duke and his career numbers.
I’m saying that statistically you are on much stronger ground the more extreme the pitchers’ statistics are.
Example: I doubt the 10 percent figure applies to Eric Gagne during his peak, modern example.
I would disagree that pitchers who are slightly above or below 10 percent in any way “regress” to 10 percent. It certainly doesn’t happen the next year. When it does happen, it’s because a star pitcher has lost velocity over time.
I’ve read the same articles you have. I don’t see the results the same way you do.
There is zero reason why a pitcher at 13 percent would regress to 10 percent. There is zero reason why a pitcher at 7 percent would move to 10 percent.
Regression to the mean really only applies to extreme outliers.
Moreover, other reasons I hate xFIP: You need to consider the ballpark. You need to remember that some guys are just better.
See, that's the question:
Was Gagne’s HR/FB low during his top two or three seasons because he was that good a pitcher, or did Gagne’s top two or three seasons look that good because his HR/FB happened to be that low?
If normalizing the HR rate toward 10% didn’t improve the accuracy of future projections, current xFIP wouldn’t do a better job of forecasting future ERA than either current FIP or current ERA. But it does.
[Since you can be as pedantic as I am, sometimes, I should probably note the exception for pitchers with unusual, sustainable IFF rates. That doesn’t apply in Duke’s case, though.]
Also, since I'm a pedant...
…I can’t let this go by without breaking out in hives:
Regression to the mean really only applies to extreme outliers.
Regression to the mean applies to all players, not just the outliers. It’s just that the difference is mostly trivial for the non-outliers.
Vlad
This is math day on here. Let’s say I can show that you are one standard deviation better than 10 percent.
Hypothetically, let’s say you are at 8 percent. Let’s also say that you’ve stayed at that number for three or four years.
I’d disagree that regression to the mean would apply to the upcoming season. I don’t think it’s appropriate to normalize at all.
I don’t think you’ll “regress” to the mean until you suffer an injury, age or show sudden velocity loss.
As an example, I’d challenge you to look at guys at 8 percent this year and tell me who’s going to be at 10 percent or higher next year. I know I can’t do it.
Another nonbaseball example: Schools rated on performance on standardized tests. I used to help newspapers with their “report cards.”
I can easily pick out which schools will regress to the mean at the extremes. But I can’t do it for schools just below or above the mean.
Why? They may have a slightly worse school system. They may have a slightly better school system.
Regression to the mean is important. But you also need to remember that sometimes people are just better than averaging and normalizing isn’t the best approach.
The problem is...
…that your standard isn’t really describing anyone in the real world.
Here’s the list of guys with 150+ IP this year and a HR/FB between 8.0% and 8.9% in the last three years:
Yep. Nobody met the standard – not one guy. I may not be able to tell you that a given guy in that range this year will be over X% – but I can tell you with a fairly high degree of confidence that Y guys in that range this year will be over Z%.
I mean, look at the size of your sample.
Over the course of each of the last two seasons, a typical top SP like King Felix allowed a total of about 200 flyballs. Which means that one fly ball catching a stray gust of wind and carrying out or not is going to shift his HR/FB by half a percent in and of itself. If a guy gets “lucky” or “unlucky” by two fly balls (one every three months), that’s going to carry him out of the range you’re talking about.
So if a guy does, for the sake of argument, have a legitimate ability that places him half a HR a month better than the average pitcher, how easy is it going to be to sort him out from the noise?
You are right ...
sample size is a problem.
But I still stand by my criticisms of using normalizing pitchers.
There are guys above average. And I just wish the Bucs had a few of them.
A phenomenon applies to all players?
The problem is that the nouveau baseball analysis has used regression to the mean as a substitute for moving toward average. That isn’t how the phrase began or is used outside of the baseball crowd.
Agreed
There is no logical reason why a person would move to 10 percent.
Just because something occurs statistically does not mean it is a phenomenon that applies to all players.
And if it happens, why? What is the theory behind it?
Okay ...
Seriously, though, give me a theory for why it would be true.
Why would it apply to great players? Why would it apply to average players? Why would it apply to poor players?
There is no underlying phenomenon to be explained. If so, James and others would have presented a good thesis to explain it.
Instead, what I see is defense of a statistic, not a reason for it.
There are lots of possible explanations.
The easiest one that I can come up with is that the hitter provides the majority of the energy involved in any given bat-ball collision, and thus the hitter is the one who has the vast majority of control over the distance travelled by fly balls (and by extension, the % of those fly balls that become HR). Thus, the difference between good pitchers and scrubby pitchers is limited, since any effect there is drowned out by the hitter-related effects. The league tends to stay near 10% year-to-year because the league talent pool tends to remain relatively constant over that time, hitters who can’t drive the ball mostly get screened out in amateur ball or the lower levels of the minors, and human physiology and biomechanics provide an upper boundary on how far any given hitter can drive the ball.
That might not be it, but it’s at least a plausible explanation, and there are plenty more testable hypotheses where that one came from.
Really? A plausible explanation
I’d love to see you test it.
I am more inclined to believe that stuff like this
meaning HR/FB, BABIP, etc….has two parameters. The first is the actual results and the second is the mean (10%, etc). What any individual regresses to is somewhere in between.
What would be really useful would be to know the standard deviation of pitchers with over, say, 500 innings in the MLB. If the standard deviation is very low, say, below 1% or .5%, then Vlad is probably right.
Vlad
xFIP is an interesting discussion (for people who love statistics and baseball too much).
Let’s use a truly elite guy: Gagne was the guy who came to mind.
Yes, I’d argue for a while he was that good. He had 100 mph heat that he could put anywhere.
I don’t think it’s in any way an accident that he was dominant for a few years.
Example: In 2003, he gave up 37 hits in 82 innings. He also had a k/9 of 15 and a WHIP of .692.
That suggests to me that he’s going to be well below average for HR. Guys weren’t hitting warning track shots. They were just striking out or making weak contact.
xFIP is, in my opinion, slightly better at predicting future ERA than other techniques. (But that’s, of course, saying that ERA has more value than I think it does.)
Part of its value is that regression to the mean means you may move from three standard deviations to two standard deviations without a change in performance.
Also, you simply get hurt, as Gagne did and then become an average pitcher.
Pure killer fastball doesn't preclude warning track shots.
Chapman just put up a 105 on the gun, the fastest pitch ever, and yet he got taken to the warning track in the same inning.
If Gagne really is getting “weak contact” of the sort you describe, he’s probably one of the exceptions as far as sustainable IFF rate, since there isn’t much contact weaker than the lazy popup.
I don’t mean to suggest that it’s an accident that Gagne was dominant. His K and BB rates were killer – he could’ve given up HR on 20% of his FB and still been one hell of a pitcher. I was just saying that a guy who’s both lucky and really good is always going to look better than a guy who’s “just” really good.
I don't recall anyone making that argument
I recall people making the argument that Dukes subpar stuff is getting worse, which you admit it is and his inability to strike people out along with having a subpar defense behind him makes him a horrible starter for this team due to his allowing alot of contact of his pitches.
We should look for another pitcher who fits our team better, they seem not wanting to upgrade the defense. If Duke stays in the rotation and they sign a FA then Morton most likely gets kicked out.
by eyeofhorus777 on Sep 27, 2010 9:52 PM EDT up reply actions
Duke is a better pitcher than Morton even with less stuff and even with declining stuff. Daniel Cabrera can also throw 96, and no one’s clamoring for him.
by Adam Reynolds on Sep 27, 2010 11:26 PM EDT up reply actions
I do agree with you that Duke's skill set...
does not line up with our defense, which from all early indications is not likely to improve dramatically next season.
You don't recall anybody making that argument about warning track shots?
Maybe you should check the post directly above mine, to which I was replying, in which Bernie said (regarding Gagne, his prototypical killer-stuff pitcher) “Guys weren’t hitting warning track shots.”
Christ, it’s like arguing with a five-year-old.
We should look for another pitcher who fits our team better, they seem not wanting to upgrade the defense.
There’s a hole in the bottom of the boat, and water’s coming in! But the captain doesn’t seem to want to fix the hole in the bottom of the boat, even though it would be fairly easy to do so! Therefore, abandoning ship is the only rational course!
You don’t want to drown, do you?
If Duke stays in the rotation and they sign a FA then Morton most likely gets kicked out.
If Duke stays in the rotation, they don’t need to pay for a crappy FA to replace him. So no, that’s not what would happen.
Just to clarify ...
I was suggesting that I don’t think Gagne was hit that hard based on his peripherals.
As a result, I don’t think luck (warning track shots) and not HRs was an explanation.
But my note was on one player and one season.
Big difference between Gagne and Chapman
Gagne had two other pitches that he used to complement the fastball.
For a short while, I believe, Gagne was clearly an exception.
When you are striking out 15 guys per nine, you are going to look pretty good, or pretty lucky depending on the perspective.
I don't quite get the Gange example...
are you claiming he was hurt in 2005 and that explains the increase in the HR/FB % being 20%. Couple of problems with that, his K’s per 9 were 14.85 which was the second highest total of his career which doesn’t exactly indicate a loss of stuff. His GB rate was also the highest of his career that season. Additionally his LD% was the lowest of his career at 11.1% which again does not indicate loss of stuff. Yet some how his HR/FB% increased from 7.5% (2002), 5.6% (2003) and 7.0% (2004) to 20% in 2005 when all of his other peripherals were the same or better than the previous three years. Seems to me like a regression to me that cannot be attributed to a decline in his skill set.
Slick
You do know that Gagne’s 2005 season ended after 13.1 IP because of a serious elbow injury, right? It was such a horrific injury that he only pitched 2 IP the following season.
That aside, I don’t claim that regression to the mean doesn’t occur. Of course, it does.
It occurs even with an elite guy like Gagne.
But there is nothing magical about the 10 percent figure. People overstate what it means.
A truly elite pitcher will do better until his stuff declines, he gets injured or he gets old.
A truly terrible pitcher will do worse because he sucks.
Moreover, the 13.1 IP is a completely unacceptable sample size. It’s not even close to being adequate.
So after skimming, your argument is that...
due to measurement imperfections in xFIP (namely, that it can’t measure how hard balls are hit), it is somewhat of a second parameter in predicting future ERA, with the first parameter being the actual ERA the pitcher held.
This makes sense to me. After skimming FIP and xFIP, it makes me incredulous that no one has created a sort of pitching UZR, indexed to ERA. What this means is that you calculate run expectancies for Ks, BBs, HRs…but also run expectancies for balls in play into certain zones. To be completely honest, it would be an easy calculation, since the run expectancy/zone data is already present in UZR.
Anyway, the fundamental question this would answer is “if the hitter Ks/BBs/HRs/hits a ball into a certain area, how does it effect the average run expectancy?” I don’t think this would be difficult for someone with average data to derive. Maybe it already has been?
Not sure I really see the value there.
You think the ability to induce “at-’em balls” is a repeatable skill?
There is a skill
In locating your pitches where the defense has lined up based on the intended location. If you’re supposed to be busting a guy inside, but it drifts out over the plate, the batted ball is a lot more likely to find a hole even if it’s not an LD.
You cited Duke’s BB and strike rates as indicative of solid control; I don’t think I buy that the correlation is that simple, since lots of guys with uncertain control simply aim for 2" onto the plate instead of aiming for the black. This isn’t specifically tied to Duke – it’s a broader argument about location. A great control pitcher can consistently locate in a 2-3" wide slot, and can therefore afford to aim for the black, since a pitch 1" off the black gets called a strike often enough to be effective. A mediocre control pitcher might still be consistent, but in a 3-4+" wide slot, and now he need to aim over the white, because a pitch 2" inside will almost never be called a strike (especially for a guy with mediocre control). A guy with poor control will simply be inconsistent – there is no slot, and so he’s relying on stuff to induce missed swings or nonswings (when he throws a hittable pitch for which the batter is unprepared).
If Duke – a guy with experience and a certain amount of confidence – is willing to aim for the white with his mediocre control*, then he won’t be giving up a lot of walks, and his strike rate will be high. A fair amount of the time, he’ll hit his spot (or better, miss on the black), and get Ks. But he’ll also miss his spot out over the plate, and the hit ball will not go where intended (a ball 4" in from the black is in the middle 50% of the plate), even if it’s not an LD.
Contrast that with a Kip Wells, who was terrified of throwing the ball where the batter might reach it, and so aimed for the black even though he didn’t have the control to hit it consistently. Now he’s throwing early count balls and having to throw meatballs later in the count. He had enough movement on the ball that he wasn’t completely ineffective, but his unwillingness to aim for the white meant that he was constantly behind in counts.
Point being, two guys with comparable control can get very divergent results, both in terms of strikes + balls and in terms of batted balls. I don’t think that any of this trumps BABIP or LD% as indicators, but I do think it suggests more variation and subtlety than you and xFIP are willing to acknowledge.
- again, this argument isn’t primarily about Duke – I’m using him here as an example. I haven’t looked at Pitchf/X to see how tight his control really is, but we all agree it’s not Maddux-like
Erm, not really...
but I really, really don’t believe that FIP and xFIP are the be-all, end-all pitching stats you think they are. Their obvious deficiency is that they completely neglect the trajectory of balls in play, and how hard they are hit. I (and probably others) had been colloquially referring to Duke as a batting practice pitcher for some time because we noticed that more than any other pitcher, hitters just seem to tee off against him. This is probably due to the fact that he has absolutely no stuff whatsoever.
Naturally, FIP and xFIP make sense for your argument here, since they don’t include trajectory at all. (actually, xFIP does insofar as it normalizes HR/FB rate…minimizing meatballs is yet another thing that helps Duke; go figure)
xFIP isn't the "end-all".
Just the best commonly-available stat out there right now.
I have no doubt that in ten years, we’ll have tools that make xFIP look like a slide rule. But until then, it’s the best we’ve got.
xFIP
Just to be clear: That’s a Vlad opinion.
Some people agree. Some people disagree.
Personally, I don’t think it’s an improvement over FIP. But that’s just my opinion.
It's demonstrably an improvement as a predictive tool.
There have been studies. Current xFIP is a better predictor of future ERA than current FIP is.
It may or may not be better for other purposes, depending on what you’re trying to do with it.
Vlad
Again, that is based on the belief that future ERA is valuable to predict.
That’s not something I accept.
So it’s a wonderful tool to predict something useless.
I think Vlad is postulating
that ERA is useful as an explanatory statistic, but not necessarily useful as a predictive statistic. Having a low ERA is something that highly correlates to wins, the end goal of, ummmm, baseball. It’s just that there are better predictors of future ERA than past ERA. This makes sense to me.
Is this correct?
I don't disagree with anything you've said there.
But I’m not sure where Bernie’s trying to go with this.
I think Bernie is confusing the predictive and explanatory
That since ERA isn’t a great predictive stat, it therefore isn’t a great explanatory stat.
I think we should consider why ERA became the preeminent statistic for so long- it has great explanatory power, it shows how many runs you give up in a full game. If you want, you can divide it by 9 and have ERA/inning. Allowing less runs gives your team the chance to win more games…it’s a simple, elegant stat that should highly correlate to wins- the teams that allow the least runs will probably win the most games.
Let me explain ...
Where I’m going with it is that I don’t think xFIP is all that useful. I prefer FIP. I don’t think normalizing is statistically appropriate. I think some guys are better than 10 percent. I think some guys are worse. I think the greater predictive value is simply the benefit of regression to the mean. Helpful, but more a statistical artifact than anything.
Second, I know it is more predictive than FIP for looking at future ERAs. I’d argue, so what? I just don’t see that as meaningful.
It’s like knowing that I can predict drowning deaths based on ice cream consumption (statistics 101 example). Helpful, but so what?
I do understand the history of ERA and the difference between predictive and explanatory power.
ERA absolutes correlates to wins.
If you could predict drowning deaths...
…then you could potentially use that ability to keep people from drowning.
You don’t see any value in that? Or, in the broader case, in optimizing projection systems in order to better forecast future performance?
Not really ...
I could tell you what days were going to be hot and have more people swimming (and thus more likely to die).
So no, I don’t think ice cream consumption is all that important.
It is a measure that highly correlates with the truly useful and causative variable, the temperature.
I’d rather just have a better weather report than ice cream consumption statistics.
Then I should be able to say
that for winning baseball games, the best “weather report” is most accurately described by not giving up runs?
If that's the only info that your ice cream scenario is providing...
…then I don’t think it’s a very good analogy for ERA and baseball analysis.
Vlad
That’s about how helpful I view ERA as.
Not useless. But not the underlying and important variable.
For it to be a proper analogy with ERA...
…I think you would need to have info that Swimmer X was Y% more likely to drown than Swimmer Z. Which would allow you to tell your beach’s lifeguard pay extra attention to Swimmer X, or to tell your salesman not sell life insurance to Swimmer X, or whatever your personal stake in the matter might be.
Vlad
I didn’t say it was a perfect comparison.
In fact, I noted that it is used in stats 101 to demonstrate a simple concept: correlation is not causation.
My point was that 10 percent is not a magic figure. People need to explain why it happens.
I have never seen a good theory, a good reason.
Until I do so, I’m skeptical. I think it’s more likely a regression to the mean factor in some cases.
But that it’s not a universal truth. And it certainly doesn’t apply to everyone.
I don't understand what you mean...
…by “a good reason”. A good reason for what part of it? Why pitchers regress to a certain number, or why that number currently happens to be 10%, or something else?
Vlad
I have to review a paper for a journal and get a test ready. So play time ends for me.
I don’t see any good theory for why all pitchers should regress to the same number.
It doesn’t matter how much you thrive. It doesn’t matter how much you stink. It doesn’t matter which ballpark you play in. It doesn’t matter how great the defense. This is what you’ll regress to. I don’t see a good explanation for it.
You offered the outline of a theory above. Okay. Just carry out a test and see how it holds.
Let me try another example.
IQ tests. And I’m not an educational psychologist so I may misstate something.
But … average IQ is 100 with a standard deviation of 15.
When you are young, your IQ probably underestimates your intelligence because of inexperience. You make simple mistakes. You don’t understand certain situations.
When you get into your teens through your 50s, it’s probably a solid measure of intelligence.
When you get older, it probably declines. You react slower. You have trouble reading because of eyesight. You just don’t react as quickly.
I suspect that a similar phenomenon applies to MLB pitching.
When you are going, you give up a decent amount of homers, even with plus stuff because of experience.
You reach your peak years, and if you are healthy, you do well.
Then, as you get older, you regress.
Anyhow, enough for playtime. Thanks for the debate.
Had fun chewing on it - hope you did, too.
I’ll try to think about how to structure a testable model.
Same here Vlad
Despite my occasional sarcasm (which you return well), I do enjoy the debates.
You argue your positions well.
Let me make clear what I was getting at as you develop your model, which I think is a great project.
My wife, for example, is smart. Real smart. Her IQ scores always range between 135 and 142.
If she scores 142, I know she’ll regress some on the next one.
If she scores 135, I know she’ll likely move closer to her true level on the next one (which is an IQ of 137, 138).
But she’s never going to regress to 100 unless she ages, has a medical condition or some sort of accident.
Her true level will always be above 100.
In the same way, I’d argue that there are pitchers whose “true level” is below 10 percent. The elite guys. I cited Gagne. He was below it during his elite years. He was near it in his early years and near it when he suffered his elbow injury.
On the other hand, I think there are guys
whose “true level” is above 15 percent.
They either improve or don’t last long in MLB. As a result, they are forced to “regress” to 10 percent.
Does that make sense? Off to class.
Bernie:
1) I partially agree with you- as I wrote down at some other point in this thread, I think xFIP is one parameter, and actual ERA the other parameter. The truth lies somewhere in between. The same applies to an individual pitchers HR/FB ratio and the 10% mean.
2) Pitchers are in the game to get through innings without giving up runs. That is absolutely the statistic to strive for. If you think FIP/xFIP should be indexed to something else, I would love to hear what it is.
I think that you'd get more value...
…by indexing to RA rather than ERA, since the assignment of errors by the scorekeeper is often fairly capricious.
That’s a pretty small tweak, though.
Pitt
saw your posting above. I wouldn’t disagree that there are several parameters.
And I wouldn’t disagree that the truth lies somewhere between them.
I think coming up with that better measure is the trick.
Let me think about it while I try to get my kids to sleep.
But I will concede that RA is better than ERA, as Vlad notes below.
But I’d add more to the equation.
After a good night of sleep and coffee ...
the biggest problem is that there isn’t a theory to explain why the statistic would be valid.
Consequently, you have people who defend it (Vlad).
You have people who criticize it (me).
Really, what you need is a good theory and a much more complex and improved predictive model that would include things like: 1. years of experience (I think young and older guys are more likely to be hit harder); 2. defense (I’m not wowed by any of the defensive measures); 3. ballpark effects.
There are probably other 10 factors I’d want in the equation. But I think this is a more interesting statistic than a valid one.
Not to agree with Vlad ...
but I’m pretty sure Bill James and others have shown that hitting balls into a certain zone is not a repeatable skill.
It’s something that I had accepted as baseball Gospel.
And when I looked at the data, I don’t think it’s true.
But what about a trajectory-based thing?
calculate run-expectancies for events: Home Runs, Flyballs, Popups, Grounders, Line Drivers, Ks and BBs. Calculate the percentages of ABs result in such events. (I’m certain that with GB/FB, LD%, this is already available…) Form a composite statistic. Index it if you so choose.
Clarification
That’s getting batters to hit balls into a certain zone.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
Related FanPosts on Bucs Dugout
FanPosts
Community blog posts and discussion.
Recent FanPosts
Baseball Nation Recent Stories
In Association With
Big League Stew - MLB - Yahoo! Sports
- Will Smith shows why making your big league debut at Yankee Stadium is less than ideal
- Steinbrenners selling the Yankees? Family denies New York Daily News report
- The Juice: Brandon Phillips flip sets up Todd Frazier walkoff for Reds
- Josh Hamilton makes great catch, destroys part of fence at Safeco Field
- Something’s amiss at the Big 12 championship
