A quote from "Practical Programming for Strength Training (3rd ed)" is apropos:
> With the peer-reviewed literature dominated by articles on exercise, forming an “evidence-based practice” – the term fashionably applied to exercise prescription based only on evidence from peer-reviewed exercise science literature – devoted to the actual training of athletes is essentially impossible. Drawing conclusions about training for athletes based on a body of literature devoted to exercise for a few small subsets of the general public cannot be and has never been productive, and all the peer-reviewed publication-worship in the universe will not make it so.
> The observations of experienced individuals – in this case, experienced coaches who have dealt with thousands of athletes over decades – are often regarded by academics in the exercise science publishing business as mere “anecdotal” reports, tantamount to hearsay and innuendo. This is a misunderstanding of the definition of “empirical,” which most definitely includes the direct, informed observations of experienced coaches. Empirical evidence gathered from an experimental study is only one type of empirical evidence, and it is dependent on observation in precisely the same way an experienced coach gathers data through observation. It is therefore precisely as valuable, especially when you consider the fact that data from a study is only as good as the methods that generated it.
> Exercise science has its problems. The populations it studies are typically small, often fewer than 20 people in the group. These people are very seldom trained athletes, and are most usually untrained college-age kids for whom any stress is adaptive. This makes for a poor way to study the effects of two different exercise methods, and completely precludes any questions regarding training. Often the methods themselves are poorly constructed, [...], completely omitting any quantification of the movement pattern being studied (precisely what is a squat? How deep is it? What is the hip angle? Does this affect muscle recruitment? How is this measured?), or display a failure on the part of the staff to standardize its interactions with the study population (“Try really really hard this time.”). Sometimes the study duration is too short to reveal anything meaningful about the question being investigated, since we are dealing with students in the study population that will only be available for one semester. Most importantly, if the study is being directed by a person without the experience to know that the study question itself is stupid (Can more weight be bench-pressed lying on a bench or balanced on a swiss ball?), and if the review staff lacks the experience to know that the PI is asking a stupid question, then stupid peer-reviewed “evidence-based” research enters the literature and adds to the problem.
> In the absence of any meaningful experimental data generated by peer-reviewed studies regarding the long-term effects of barbell training, we are forced to rely on the observations of hundreds of thousands of coaches and athletes who carefully picked their way through the mistakes made during the process of acquiring experience. This makes a rationalist out of every effective barbell training programmer. This process – if it is to be logical, effective, and productive, i.e. rational – must be guided by a thorough grounding in the sciences of physiology, chemistry, and physics, since the “exercise sciences” have proven themselves to lack the rigor and scope necessary for the task. The well-prepared coach has either a “hard” science degree or an otherwise extensive background in biology, anatomy, physiology, physics, chemistry, and probably psychology as well. Textbooks on these subjects should form the basis of the coach’s library, with practical experience under the bar and many thousands of hours coaching on the platform rounding out his abilities as a coach of barbell training.