Premature Optimization and the 95/40 Rule

95/40 RuleI drew this diagram on my whiteboard several months ago to explain to a friend why eating a Standard American Diet (SAD) and trying to fix it with supplements is foolish. The SAD is the column on the right, achieving, for argument’s sake, 40% of an optimal 100%. The column on the left is something closer to an ideal diet for this individual (the Paleo Diet in my case). The effects of supplementation are the small 1% bars on top of each column. Sure, they both get you 1% closer to ideal, but who gives a damn about going from 40% to 41%? The difference is negligible.

After drawing this chart, something interesting occurred: my roommate and I started to draw connections in dating, work, and life in general. For example, suppose I go on a date with a girl who doesn’t like science fiction. That’s a 1% issue in my case. Maybe I can convert her to love Firefly and Asimov (or get over it), but focusing on this 1% improvement prior to looking at the starting point is flawed. If she loves reality TV and despises exercise, these fundamental issues create a poor foundation on which to worry about moderate improvements.

There’s a saying from computer scientist Donald Knuth, “premature optimization is the root of all evil.” If you hate your boss, your commute, and your daily tasks, who cares if the coffee machine is being replaced with a better one? Focusing on minutiae when fundamentals are missing is folly.

The same conclusions can be attributed to learning new skills just as easily as one’s job or diet. With tango, if my fundamental approach to the embrace and leading is broken, who cares if I know how to do ganchos? If one’s diet is 70% grains and sugar, taking a multivitamin isn’t going to fix it.

All this is related to priority of learning and energy expenditure. Find a reasonable foundation, then find the biggest wins with the least required investment and complete them. Then find the next, and repeat. Don’t build a job or a relationship on a flawed foundation, it’s not going to get dramatically better by making incremental improvements. Keep looking until you find a solid foundation, then improve.