Time and System Design

Note to self (sort of obvious).

In system design time is an assumption it’s not feedback.

The proof of this is trivial. No matter what your system does or even if it exists, time marches on. Time doesn’t depend on your system in any way.

For example, if I design some sort of whiz bang toothbrush and I build an alarm into it that goes off after five minutes indicating you can stop brushing then I’ve made and assumption that five minutes is enough time to brush. No portion of my system has actually measured how clean your teeth are. I’ve only built an assumption into the system.

It may have been an informed assumption. For example, studies may have shown that on average brushing for only three minutes is not enough for most people and brushing for six or seven minutes doesn’t substantially improve how clean your teeth may be. In the real world however there will be circumstances where three minutes was plenty of time to get your teeth clean or seven minutes wasn’t nearly long enough. Outcomes are distributed and I’m building an informed statistical assumption into the system, but it’s still an assumption.

There are a lot of assumptions we build into systems, not just time based. In the toothbrush example, what is clean enough? Is my definition of clean enough the same as yours? My assumption may be based on long term dental health. Yours may be based on how your smile looks and your breath smells. If I know about all of these needs, how do I prioritize which is most important? Your priorities may be different. When cleanliness was measured in our experiments, how was it measured? If it wasn’t comprehensive in some way then there’s room for error. System design is a bet and is very subject to judgement. The objective of system design is to create a system that works often enough. It’s not to create a system that works all of the time.

It’s good to remember this because I often forget and get frustrated when something doesn’t do what I expected it to do. To avoid some kinds of frustration it can be good to ask yourself, “Does this system seem to work most of the time for most people?” If the answer is yes, then let go of the frustration. For me I either need to change my expectations or I need to stop engaging that system.

Interestingly, we arrive yet again at the basic equation …

stress = expectation – reality

This is another reminder.

Bookmark the permalink.

Leave a Reply