Optimized For What?

A core element of program development might be called “optimization”. Real optimality doesn’t exist, but in writing a program, what we’re trying to do is approach optimality. Much like “truth” or “correctness” (about many things, maybe not everything) we can’t ever really get all the way there, but we can do a pretty decent job of approximating optimality.

The obvious question, then, is “optimized for what?”. A program can’t be optimized generally – there must be some referent for the optimization process. Optimization may look different for two different athletes participating in the same sport. Of course, it will look wildly different for athletes competing in different sports. In writing effective programming, then, it’s important to have a clear answer to the question “optimized for what?”.

Today, I want to address a pattern I see in programming for groups, particularly in CrossFit affiliates and similar gyms. Often, coaches spend a lot of time and energy optimizing their programming for the wrong thing. Specifically, they treat the optimization process as though they were optimizing for a theoretical individual, usually a proxy-person representing the “average” needs of their gym membership.

The problem with this approach is that it treats the proxy-person as though they really existed, and the theoretical average needs as though they were real needs. Even if the coach is able to identify the collective needs of their affiliate correctly (a challenging task in itself, beyond the scope of this discussion), that average is merely representative of a set of facts about their individual members. The word is not the thing . As a result, programs often end up aimed at some purely hypothetical set of needs and goals. Undoubtedly some subsection of the membership will benefit from the program. Equally certain is that some other subsection will suffer as a result (unless a gym happens to be uniquely homogenous in the needs of it’s clients, to an extent that I’ve never seen).

The core issue is that the program is being optimized for the wrong thing. In fact, it’s being optimized for something impossible: the attempt to take an explicitly generalized model of training, and make it specific.

It’s not going to work. It’s never going to work, because there is no specific referent. You cannot reduce the membership of your gym to some set of hypothetical averages. You have to treat the membership as a cohesive whole, and accept the constraints of the training model. Those constraints include being unable to write the program to address the needs of individual clients.

The concern, of course, is that clients utilizing such a model will always be hampered in their progress. They will never improve at the rate or to the level which they would improve if they were following a program designed specifically for them. To which I say: yeah, duh. That’s the tradeoff: you can pay less money, work out with a group, get a good general program that will make and keep you fit and health. On the other hand, you can pay more money, get a great specific program that will optimize to your needs and goals.

Instead of trying to make a general program suit a set of non-existent specific needs, gyms are best served doing two things: first, making a point of optimizing the program using the amalgamated membership as the referent for which you are optimizing. There are plenty of ways to do that, which I’ll save for another day. Second, capitalize on the opportunity to provide more specific programming for clients who want it. If you have clients who need or want more than what they can get from the group program – great! That’s a fantastic chance to offer a high value service at a premium rate.

In short, when writing programming for groups: optimize for the irreducible whole, not for the imaginary individual.