Dr. K and Dr. G will agree with me that a suggestive study is a much different animal from a conclusive study.
Too true. 90+ % of "studies" done in the medical world are very weak, IMO. Retrospective and case control studies, the majority of studies done, at best can demonstrate correlation. Usually this is because doing a proper study is too expensive or will take too long. Naturally this doesn't prevent the press from reporting "X causes Y".
Even the gold standard randomized/controlled/blinded studies accept a (calaulated) error rate of 5% for Type I error and 80% for Type II error. Even in an ideal context, when dealing with very complez questions, it is also really hard to eliminate all sources of systemic bias even with randomization and it is often also extremely difficult to arrange things so that you're really comparing apples to apples.
...........
=== Tier III Thinking, Dept. ===
Related to Dr. K's idea, is the roto "tier" concept.
Suppose that you gave 10 historical HOF third basemen to the AL East and Central ... Schmidt to Tampa, Brett to the Indians, Boggs to the Orioles, etc.
That would raise the industry average at 3B, and therefore would raise "replacement level" which is about -20 runs below average.
But would it do the Mariners any good, if they don't have one of the ten HOF'ers, nor one of the ten good 3B's playing right now?
...
Let's say the league happens to be stocked at a position, at the moment -- suppose (as a thought experiment) there are (say) 25 great hitters in the game, all at 1B. Does that make it any easier for you to develop a 1B who hits 40 homers?
Roto owners are very familiar with this. There are Tier I first basemen like Pujols, Cabrera, AGone ... Tier II guys like Konerko, Morse, Lee, and Howard ... etc.
But every roto geek knows that if the 1B pool is 22 thumpers deep ... the industry average of .320/35/120 isn't going to help you if you get Casey Kotchman. There is an arithmetic mean, and an arithmetic median.
...
WAR fails to account for the fact that a player might be "average" at his position, and yet "far above average" in real terms. The "positional adjustment" at Fangraphs is helpful, but any given season's adjustment is not a law of thermodynamics.
That's not WAR's fault, but it's the fault of the sabertista when he scoffs at anyone who doesn't use WAR as the Last Word.
.
=== Question Authority, Dept. ===
What is replacement level? It is an assumption! It is a supposition, presumed as true, that any organization can get "Catcher X" for free, any time it wants. Tell that to the 2010 Seattle Mariners, stuck with Adam Moore batting .195/.230/.283, and no way to replace him.
GM's take care of their "stoplosses" invisibly and at great cost. There are times those stoplosses are not in place. Had the 2011 Mariners had a "replacement level" offense, we'd have had a pennant race. Very often the actual RL, at any given moment for any given team, is MUCH lower than assumed.
And all we're asking is for sabertistas to open up to real-world considerations like that. The preseason assumption blends softly casual into a simplistic model that is detached from reality.
Can you simply bring up a AAA player and get RL production? Not if the kid needs a year to adjust. So how is that "freely obtained RL production"? There is a time (not money) barrier between you and the RL performance.
***
RL is arbitrary, it is a presumption assigned -- and ASSUMED FACTS DO NOT EQUAL SCIENCE.
...........
A few sabertistas want to philosophically assume a worldview, and then treat those assumptions as scientific truth. This is where much of the debate hostility is born.
As in mainstream America generally, there are folks who can't wait to assume a thing as True, and then to get started on the Inquisition.
Senator, how soon would you remove all troops from Iraq? Governor, how much funding would you give to stem cell research? Do you accept global warming? Do you accept classical Darwinism? How about seatbelt laws? Siiiiggghhhh. Woe betide those who express an incorrect opinion.
Also in university halls in 2011, few students ever wonder about how many of science's premises are philosophical assumptions. My kingdom for a kid who will truly question authority.
Ain't it funny how some revolutionaries act, once they win the revolution and become The Man ? :- ) Heaven help those who Question Authority, after Che wins ...
.
=== Synergy on Offense ===
In this article, the zealots in the comments attacked the author angrily for "failing to realize that lineup protection is disproven."
First of all, lineup protection is not disproven, not that I've ever seen. But then, internet saberdudes have always had a very rough time discerning whether a mini-study proves something, or doesn't. Dr. K and Dr. G will agree with me that a suggestive study is a much different animal from a conclusive study.
My guess is that there have been "suggestive" studies running counter to Lineup Protection. I remember when CERA was "disproven" and PAP was "proven" and anybody suggesting otherwise got laughed out of school. Those dudes simply could not discern what proof is, and what proof isn't.
I can't even imagine how complicated it would be to conclusively prove that Lineup Protection doesn't exist.
...
Nyjer Morgan, said Hippeaux, is walking at 50% of his career rate, now that he's in front of Braun and Fielder (and Morgan is having a great year, having gone from the punchless Nats to the Brewers).
Maybe the guys in the comments thread consider it worthy of ridicule to ask whether Morgan's being pitched to because of Braun and Fielder. I know what Jack Zduriencik, Pat Gillick, and every other GM would consider it.
Obvious.
***
The funny thing is, Hippeaux didn't argue lineup protection in that article. He said, "Adrian Gonzalez hits great when the pitcher is in the stretch ... and now that he's with the Red Sox, he's getting that 52% of the time."
Without any question, pitchers are less effective when they're using the stretch, holding runners, and throwing the 31st pitch in an inning. :- )
Ballplayers will all tell you, hitting is contagious. Maybe they just mean, a tired and frazzled and un-confident pitcher is easier to hit. Maybe that's 70% of the "hitting is contagious" phenomenon; maybe it's 95%.
But the 0H 2012 Mariners are stringing some tough AB's. And once they jell together, you might find a 5 or 10 OPS+ bonus for each hitter into the bargain. Maybe, following Ackley and Carp, the 2012 Justin Smoak is going to see a few extra mistakes from those struggling SP's throwing from the stretch.
.
BABVA,
Dr D
Comments
The Mariners were randomly selected to receive a replacement-level offense as part of a double-blind experimental design.
All in the name of science, of course.
No such thing as a single stat that can tell you everything about a player's value. There never will be. WAR is a huge part in the evolution of evaluating players, but I don't know of anyone that thinks it is the end-all.
I'm trained in the quantitative physical sciences, where you claim understanding when you can construct a model based on sound accepted principles that reproduces the quantitative trends of your observations (measurements). Baseball statistical analysis is more like amateur macro-economics.
Triple-blind: the participants didn't know, the GM didn't know, and the observers never found out.
Great line, Spec!
You didn't read the comments in Specs thread that you c-pointed? You don't read Fangraphs? ;- )
I know what you mean. Nobody should think that it's the end-all.
The problem with WAR is letting the league set your standards of excellence. The problem with production dollars is letting the Yanks set your market value on star players. Look to teams that share common ground with the M's and set your values appropriately.
People who are determined to compete spend all their energy finding a way to do so, whether that be finding an as yet undiscovered niche or just plumb working harder than the other guy.
People who are determined to compete don't use competitive disadvantages as excuses to fail.
Some people are determined to try but content to fail.
Institutional thumbs on the scale of justice (the advantages of clubs like New York, Boston, Philadelphia, Chicago Pale Hose, LAAofA) are, unfortunately very, very real. The deck is indeed stacked against a team like Seattle.
That is why it was so frustrating to see the Mariners against all odds actually overcome these problems during the Lou Piniella era, literally shoving their way into a seat at the table of top teams, only to squander their newfound opportunities by front office ineptitude and the penny-wise pound-foolish miserliness of the Grand Poohbahs. Little did we know that Ichiro's rookie year of 2001, was like a sugar high, an era that would quickly come to an end and the beginning of the franchise's demise.
Things seem finally to be changing under Zduriencik. Their seems to be a combination of a plan and the expertise needed to execute it. We hope that there will also prove to be the needed support from the Grand Poohbahs to bring it to fruition AND KEEP IT THERE.
You do a much better job explaining what I was trying to represent about the relevance of lineup construction to performance. I tried to make a similar argument a few months back and also found myself struggling to represent "geometry" of rosters, either statistically or linquistically:
http://thesportinghippeaux.blogspot.com/2011/02/reading-dan-rosenheck-an...
I really appreciate the thoughtful, aphoristic work being done here.