This is the new front of the genomic medicine why not make use of it in sport science as well (since we're all running the experiment and collecting the data anyway)? We all have 99.9% identical genetic material, It's only by looking for trends in a larger portion of the population that we see that the .1% make a difference. e.g. One of the more recently identified gene mutation is associated with colorectal cancer is called APC I1307K. This gene mutation is found in 6 percent of the Ashkenazi Jewish population and 28 percent of Ashkenazi families with a history of colon cancer, making it one of the more common cancer susceptibility genes.
It's the same story for MTHFR, Factor V, Prothrombin (Factor II), and hypercoagulation (blood clotting and or cardiovascular disease). CYP 450 and drug metabolism. HER2nu and breast cancer. p53 and various cancers. These are just off the top of my head, I could go on if I dug in my paper archive.
True, but I'm of the mind that Chad would love to hear that ABBH resulted in (for example) 2% lbm gain during a 6 week cycle for 75 of 75 people regardless of diet, prior training, gender, age, steroid consumption, etc., etc., etc. And you're right, there are lots of variables that may not or can not be accounted for that's part of what I'm trying to overcome.
First of all, read Cressey's "28 Synergistic Factors" and Poliquin's "The Five Elements" and I think you'll see that while this would be tough, it's not unwanted and undoable (IMO). Also, I can quite easily envision doing what I'm thinking about for a database and my talents extend all the way to Microsoft Access (which is, in my understanding, the database equivalent to being able to drive a car).
Here's the part where I admittedly step on some toes (He's a big guy, I don't think he'll cry.) As for spurious correlations, How about high frequency training because gymnasts and mechanics do it? (Once again, no offense, if the program works then this is a non-issue and I'm a sack of shit, but you gotta admit that it's spurious.)
To answer your specific question, ideally, you would collect enough data on "Beach-muscle building 16-17 yr. olds" or "upper body biased teens" and "so-and-so offensive linemen" or "Professional Linemen" such that you end up with four groups, "Teens AM", "Teens PM", "Lineman AM", "Lineman PM", and your "Average T-person" control group and see how the factors play out.
I'm not necessarily looking for the "best" program out there (it would be nice and interesting if we stumbled across it though [something like a program that adds 50# and 500# to your squat in a year]). "We" would be looking for the program with the highest probability of success given a set of very clear variables or in the example above, the program that, almost without regard to who you are or what you do, at least puts SOME mass on you.
BTW- It's kind of scary how many people keep telling me I'm not 'seeing the big picture.' No matter what I'm talking about on these forums. I get it, training is, for large numbers of people, very, very, complex. However, depending on your religious beliefs (How's that for big picture?) it's not absolutely irreducibly complex.