Over the past few months I’ve seen some well-intentioned, bright folks get pretty sideways because they were using "rules of thumb" without understanding the principles underneath them. To be clear, we all use rules of thumb every day. It’s a part of our human existence. We are bombarded with information and crushed underneath our expectation to do more. The kids need transported to their two events each this weekend. There’s a presentation on Monday that we’re not quite prepared for. We’re looking for any shortcuts and efficiency that we can get just to squeeze a few extra minutes in so that we might be able to check the sports scores or watch or favorite TV program.
Rules of Thumb allow us to simplify the problem, to make it easier. We know if we multiply a number by 10 we just move the decimal point one place to the right – even if that means adding a zero. We know that changing our oil every 3,000 miles is a good rule of thumb. Newer cars will tell us when we need to do this but without the internal monitoring every 3,000 miles is simple to remember. The problem isn’t so much that we use rules of thumb – rather it’s that we fail to recognize when those rules of thumb aren’t going to serve us.
Pushing Things Too Far
One rule of thumb that builders use is to square an angle using the Pythagorean Theorem. It is the formula A2 + B2 = C2 that we learned in school. Builders use it to square off corners. Well, more precisely they measure three feet in one direction, four in the other, and then manipulate the angle until the diagonal is 5 feet. This is a rule of thumb – a shortcut. It works because 3 squared is nine and four squared is 16 and the square root of these two values (25) is 5. It’s quick and simple, that is until you try to use this to square off walls – or the edge of a deck that are 20′ x 40′. In my case the builder did this and created my deck which varies by 6" from one end to the other because it doesn’t line up with my house.
It’s not that the rule of thumb is wrong, it’s just that it was used at a scale that it wasn’t designed for. Small errors in measurement are amplified over distance. Moving from 4 feet to 40 makes a ½ error 6 inches on the other end. If you’re thinking that your servers will be good enough because you’ve always just been OK by purchasing stock servers with standard hard drives and putting it together, you’re using a rule of thumb. That is that what’s in the general market will probably meet your needs. Detailed understanding would look at rotational speeds, RAID configuration, controller cache, and dozens of other factors, but you don’t have time for that.
Your personal rule of thumb probably isn’t the same, but what about more expensive = better? Have you ever bought an expensive product just to find out that it broke, or didn’t work the way that you want it to? This happens to me all the time. While I’ve gotten better about evaluating items, trying to make sure that I’m not assuming that more expensive is better – I still use the rule of thumb sometimes.
I remember a story about the day that suspension bridge making changed. It was November 7th, 1940 when the Tacoma Narrows Bridge, "Galloping Gertie", collapsed under 40 mph winds. Before that point, aerodynamics weren’t considered in bridge making. Engineers had been building bridges for hundreds of years but the introduction of new stronger and lighter materials had changed a non-influential factor, aerodynamics, into a critical one. I realize that engineering as a discipline rarely uses rules of thumb, instead relying on detailed mathematical models. The point here is that even when the details are known there are times when the factors under which the models are built change and that change makes the model ineffective – or perhaps more precisely incomplete. If detailed engineering equations can be influenced by unforeseen factors so too can our rules of thumb.
I see all the time where organizations use solutions. They say "but I followed the best practices." There should be a disclaimer on "best" practices like "your mileage may vary." Why do we need a disclaimer? We need a disclaimer because a practice is only "best" given a set of circumstances. What’s the "best" car for me? Maybe it is a sedan or a SUV or a truck or a minivan? Until you know what I do in my driving there’s really no way to define what best is. I buy a minivan and suddenly decide to start hunting as a hobby and I need to start hauling the meat home.
The One Best Practice
There’s only one best practice, one universal rule of thumb. That is that you have to keep evaluating whether the rule of thumb that you’re using is still serving you – or is it at risk of causing a catastrophic event. It’s a good idea to reevaluate your rules of thumb every once in a while … and maybe call in some help to determine if your conditions and your rules of thumb are in sync.