This year he's been annihilated by sliders, while hanging in there vs. fastballs (despite seeing an AVERAGE FB velocity of about 92 mph).   And this year, he's done much, much better versus power pitchers than finesse, per B-Ref's definition (1.118 OPS with a BABIP under .150!!!).  
Seems like it's not a fatal flaw, just statistical noise.   Look at the fangraphs pitch type values again - he's been hopeless against ANY offspeed pitch this year (min. 5%).  

This is remarkable. 
You're saying that he has a slider-speed bat based solely on one statistic - the B-REF definition of 'power' versus 'finesse.'   If he actually had a slider-speed bat and simply could not catch up to a fastball, what would you expect to his pitch type scores to be?   Why has he been eaten alive by sliders?   Secondly, what's the context?   What's league average for these pitcher types, and what does it look like if you use the same BABIP for each sample?   Finally, wouldn't you expect a poor total RV for FBs given the fact that he's been a poor hitter - overall, against everybody - in his career?
Let's take another tack.   Player X has a career OPS vs. 'power' of .699, versus a career OPS vs. 'finesse' of .909.   This gap dwarfs the one you pointed out for Langerhans.    Should Player X sit versus power pitchers?  
The point here is that you haven't proven that this statistic is correlated - at all - with batspeed.   If this were a fatal flaw, results like Langerhans' 2009 would be pretty much impossible.   You're saying that once the league learns to throw him all FB's, he's toast... but it's certainly relevant that he's hitting FB's BETTER than off-speed stuff RIGHT NOW.   He's got 2 HRs on FBs off of *power pitchers* this year.   

So you're going to stick with your theory despite the fact that the metric you used is not actually measuring pitch velocity, and isn't actually considering pitch type.  If you wanted to prove a theory like this, you absolutely would NOT use the b-ref pitcher type stat - you'd need pitchfx data and you could do a plot with FBs categorized by speed.   Then you could see if Langerhans was far below league average vs. medium/hard FBs, and dominant against slow ones or something like that.   But using the B-ref stat as a proxy simply doesn't work, and it seems like you'd rather cling to the conclusion than examine other hyptheses.   The example in the previous post by the way - the mysterious Player X - is Russell Branyan.   Please explain why you think Branyan has a slider speed bat given his 'disastrous' splits in the same stat.   He makes Langerhans look like Ichiro.   I'm not saying that Branyan has a slider-speed bat, just as I cannot say that Langerhans doesn't.   I'm saying I don't know, and can't tell given fangraphs pitch-type LWs and B-ref.   The one thing I DO know is that you don't know either.  
The reason is that B-Ref's definition of 'power pitcher' is simply someone who gets Ks and BBs.   It isn't trying to measure speed, and thus would give equal credit to Trevor Hoffman and his 85mph heater as it would to Jonathan Broxton and his 97mph FB.   It also wouldn't account for guys who rack up Ks on slow stuff, like Hoffman (or even Lincecum), which is why Branyan does well v. FBs in the fangraphs metric, but is hopeless against 'power pitchers.'   This isn't the right tool for what you're trying to measure.  
If you use the fangraphs pitch type data, yes, you see he's got -7.5 runs on the fastball, which again is not at all surprising given the fact that he's not hit very well over his career and the FB is the most common pitch.   As a comparison, Jose Lopez is at -27.7 runs vs. fastballs.   Clearly a slider-speed bat there.    Yuni Betancourt has roughly twice the PAs of Langerhans, but his career FBv isn't ~-15, it's -36.5.   Erick Aybar's at -9.2 for his career in fewer PAs, but he's an above average hitter this year.   Jimmy Rollins is at -7.0 on the fastball in 2009 ALONE, while Robinson Cano was a stunning -16.4 just for 2008.   By the way, this year?  He's at +15.0.    Again, if you want to see if someone's batspeed is slipping, that's cool, but you'd probably want to plot their LD% or even slash line against different groupings of fastballs.   You could check it against hittracker's batspeed measures too.  

1) I use's B-Ref's proxy not as an attempt to make a foolproof case, but as an example that backs up a scouting-based opinion.  My initial reaction to Langerhans was not based on B-Ref's power split...it was based on actually watching Langerhans hit and observing that most of his"power" early in his stint either came on sliders over the plate that he pulled or fastballs that he was late on and happened to get opposite field results with.  After ten games or so with Langerhans starting, the common reaction over at MarinerCentral was "is it me or has Langerhans not gotten any of his doubles to right field?" (to which I responded that he actually had hooked a sinker or slider, I do not recall which) down the right field line in his first game against the Yankees, BTW).  That has continued to be true throughout his stint in Seattle...most of his extra base hits have been doubles to the wall in left or HRs pulled on sliders, two of which won games for the Mariners  I've been in sabermetrics for 9 years now...I am well aware of how B-Ref's splits work, but thanks for assuming I was an idiot and talking down to me like a 4 year old.  Awesome.
2) ASlthough the B-Ref power/finesse split doesn't identify velocity exactly, if you look at the pitchers who
are classified as power pitches, you will find that their average fastball
velocity is higher than the pitchers in the finesse group.  It's a eak but not an
incorrect proxy to use in the absence of hit F/X data.  Hittracker measures
only home runs...it's not going to help us identify a slider speed bat the way
you claim.
3) Yes, if I had access to the full pitch F/X database, I would approach the
question of whether Langerhans had trouble with high velocity pitchers very
much the way you described...checking to see if he did worse against harder
fastballs than softer ones, but I don't at the moment and none of my online
references make that kind of data available as of yet.  What I do have is my
own powers of observation and two halfway decent proxies which back up
said observations.  Your case seems to be "if it si not 100% proven, the
observation has no merit" followed by "in 100 at bats this year, the data
doesn't appear to back up your assertion, even though the career data does"
When you have somet kind of substantive evidence that runs counter to the
two proxies presented here, let me know.  Right now, your only data-driven
argument is based on a tiny sample that could easily be very misleading.

This is a telling statement, and probably explains why any sort of debate here is pointless: "Your case seems to be "if it si not 100% proven, theobservation has no merit."
 
You're using a metric that you don't understand as a proxy for something you don't know to back up something you think you've seen visually in an extremely small sample, and what's hilarious is that you think this process has something to do with sabermetrics.  
You think I'm trying to prove or disprove something; I'm not.   All I'm saying is that you have no idea what his batspeed is based on the evidence here.  If your case is based more on your observations of Langerhans, SAY SO.   In any event, most of his XBHs have been pulled, and BOTH of his game-winning HRs have been pulled *against 'power pitchers.'*
"thanks for assuming I was an idiot and talking down to me like a 4 year old.  Awesome."
If you KNOW how they work, and you know their limitations, we wouldn't be having this discussion.   You're not measuring what you think you're measuring, so you haven't 'proven' your case albeit shy of 100% proof, you have nothing.   It's really, really important to understand when you have data and when you have noise.   This is exactly equivalent to saying that some player looks good defensively in his first 20 games or so, and then using fielding percentage to 'confirm' this.   You don't have 'proof' in that case, you've got a tiny sample of a terrible metric appended to a small visual sample.   That's.... well, it's something, but it's clearly not sabermetrics.   The entire approach is different.   Look, there's no shame in saying that you don't know, or that it's something to watch for, or whatever.   But you're clinging to this thesis (and getting defensive about it) in the face of a ton of uncertainty and variable data.  You are searching for ways to 'prove' that your eyes are right, and that's fine, but it's not sabermetrics.
So what ARE the league average splits in your b-ref metric, and how far from average is Langerhans?  You haven't addressed the Branyan counterexample either.   I still don't think it measures what you want to measure, but you could *possibly* try to show that Langerhans is further from average after you normalize BABIP or something. 
 
Yes, hittrackeronline only measures HRs, so it is not a perfect metric either.   I have no problem admitting when I don't have all the evidence.   However, it's at least measuring actual batspeed, and not a proxy of a proxy.   As we all know, the best thing to do here would be to check his splits with pitch fx data.   Then we'd have something.  
"When you have somet kind of substantive evidence that runs counter to the two proxies presented here, let me know."
I can't prove a negative, and in fact I'm not trying to prove anything.   I'm just saying that you don't have any basis for your claim.   Your claim may even be right, but neither of us know that.   When you understand this, then we may be able to have a civil discussion.  
 

Asked above, but not answered:
League average (2009) of all AL hitters:
vs. Power:  .245/.333/.394/.726
vs. Finesse: .289/.344/.469/.813
Aggregate LEAGUE split gives EVERY hitter an 87 point edge against the bbref "finesse" pitchers.  I'm confused on Matt's numbers, though.  Langerhans career and 2009 numbers:
2009 Ryan vs. Power: .214/.389/.643/1,032
2009 Ryan vs Finesse: .147/.189/.353/.542
Career Ryan vs Power: .213/.325/.324/.667
Career Ryan vs Finesse: .267/.341/.424/.765 (98 point edge)
It's only 111 PAs, so the problems with sample size are rife.  But, WITH THE MARINERS, RL has been pounding the crap out of "power" pitchers, (in his 14 ABs/ 20 PAs). 
=========
In truth, I think Matt is probably very much on line with his eyeball assessment, but 100% wrong on his analytical conclusions.  I have no doubt that RL "appears" to have a slow bat these days.  But, I bet if you went back to the first weeks when he was on fire, the assessment of his bat speed and reactions would paint a completely different picture.  You see, I don't think Matt's observations refute my position on Langerhans, I think they CONFIRM it. 
RL *is* a streaky hitter.  When he's "locked in", the basic reality is that he picks up the ball better (quicker), and not only does hit bat speed "appear" excellent, he "appears" to be an absolutely fantastic hitter, who ought to be putting up MVP numbers, (as many suggested during their first week looking at him).
But, when he's NOT "locked in", he doesn't ID the ball as quickly.  So, he's late on the fast stuff.  But, here's the lesson ... he's NOT the same hitter every day, (no hitter is).  And RL's inconsistency is pretty extreme.  His streaks tend to be SHORT.  So, he'll pound the ball for a week, scuffle for a week, have a meh week, and then repeat the pattern.  The stud hitters will streak for 6 weeks at a time, and keep the slumps much shorter, (10-12 games). 
Oddly enough, Matt commented on Wilkerson (hello way-back machine), showing the EXACT same profile at the plate.  He started off the season late on everything.  BUT, being a hitter that doesn't swing UNTIL he identifies the pitch, his walk rate doesn't typically suffer during a down spell.  (In fact, walk rates tend to plunge for this type of hitter when they ARE seeing the ball well - because they start killing the strikes, instead of fouling them off). 
In truth, I think, a lot like Wilkerson, that Langerhans would probably fair better in an everyday role than coming off the bench.  But I think his inconsistency is bad enough that he's a weak choice as a full-time starter.  The plus of a guy like Langerhans as a #4 OF is that he provides quality defense, a decent bat off the bench, but a plus over a more stereotypical #4 OF, he likely improves his performance significantly if forced into a full-time role due to injury.
A guy like Orlando Palmeiro is going to be basically the same as a sub or as a starter, good conact, low K-rate, few walks or power.  He is what he is.  A guy like Langerhans won't carry a .280 average for you.  But, he'll draw walks, can run the bases, play good Dee, and has decent power, (exceptional power if compared to anyone coming off the Seattle bench in the past 5 years). 
This is especially important to consider when thinking in terms of backing up 3 OF positions AND the DH slot.  If you're backing up 4 slots, you want to have someone who can step into the lineup full-time, (due to injury), that can post a line only "slightly" below the regular.  But, most teams cannot AFFORD to keep such a bat in reserve.  Langerhans is perfect in this regard, because his abysmal average keeps his dollar cost very low, but his other plusses actually keep his real-time value (if needed) much higher than say ... an emergency call-up from the farm.