My '90 percent theory' tries to explain why perfection is elusive

I've always been fairly messy in managing my personal space at home. I'm happy to let all sorts of trivial items - bank statements, magazines, pill packets, slightly soiled tissues - accumulate on my bedside table before feeling a need to clean them. And when I do finally clean it, I sort through most of it and leave it in a much neater state. Not spotless, but vastly improved.

My partner Melanie is different. She's fastidious in her neatness, letting only a handful of items accumulate before feeling the need to sort through them and clear up the space. Once she's cleaned up her bedside table, it sits in a very high state of tidiness, substantially cleaner than my side even immediately after I've completed my "cleaning". It is not unknown for her to assess my side after I'd clean it, and make some smart suggestions ways to deal with the handful of items I've left sitting there.

What intrigued me was not so much the difference in the states of mess to which we would allow our respective sides to reach, but the fact that both of us would consider that we'd achieved "cleanliness" at very different points.

It lead be to a theory I'll call my "90 percent theory" in the absence of a better name: basically, in any given task in which people seek to improve something, they will be able to improve it 90 percent of the way between how they found it and perfection.

So, applying it to my and my partner's bedside tables, from the point at which each of us decide to clean our sides, we will improve it 90 percent compared to how it was. Because I start at a more advanced state of messiness, 90 percent improvement will not make it as neat as my partner's side.

Furthermore, when she approaches my tidy-but-not-perfect bedside table after I've cleaned it, she will be able to improve it a further 90 percent on how she's found it, bringing it to 99 percent cleanliness compared to its initial state. Theoretically, were a third person to come along and seek to clean it again from the state that my partner had left it in, they would be able to improve 90 percent of what they see, bringing it up to 99.9 percent. But given the difference between 99 percent and 99.9 percent is so minimal when the entirety of the task is cleaning my bedside table, there is probably too little to notice.

It's worth noting, however, that by this rule perfection is impossible. Improvement is, in a mathematical sense, asymptotic. Each attempt gets us ever closer, but we can never reach it.

Human psychology is the main factor behind the theory. When we approach a given task, we generally assess the quality of the outcome relative to the quality of the starting point. It keeps us sane by preventing us from obsessing over every minor imperfection, but does also make us occasionally complacent and willing to forgo the pursuit of excellence.

I've noticed this theory playing out in another context, in my work as a copy editor at a newspaper. In the role, we take the stories filed by reporters and have responsibility for checking facts and spelling where possible, ensuring the expression is smooth and professional and making sure that the story is neat and logical. The task has an extra degree of difficulty because most of the reporters do not have English as their first language, and so the quality of the written expression is quite varied.

On the copy desk, the first person, usually but not always someone with less experience, will perform a copy edit; then they will pass it through to a check editor, who will give it further scrutiny; then finally a page editor will look at the page as a whole, keeping an eye out for anything that has been missed.

I've performed all three roles at one time or another, and can see that the quality of one's output is in part a product of the quality of one's input. So a copy editor dealing with a particularly poorly written story will work hard to improve expression, spelling, etc, but is likely to unconsciously let through some less-than-ideal parts because their focus is diverted to the more blatant imperfections. That same copy editor, however, working with "clean" copy in the first place, will notice and hopefully rectify much more minor issues that become more apparent because of the quality of the work around those issues.

Though I have the experience to work as a check editor, when in the role of copy editor I find myself overlooking errors that I know I should have picked up and suspect I would have were I to be a fresh pair of eyes reading it as a check editor. Improving beyond the 90 percent is remarkably difficult to achieve. (This also demonstrates my asymptote theory - despite three pairs of eyes picking up theoretically 99.9 percent of errors compared to the original story submitted by a reporter, errors still make it into print.)

You can even see the theory in action in a political context. Take traffic, or smog, or corruption, three issues that afflict Jakarta particularly badly. A government that commits to rectifying these ills will generally seek to rectify them by a proportion of the initial problem rather than to reach an absolute number. (Of course, few governments would be so bold as to seek to reduce any of these measures by 90 percent; instead they are seeking to reduce the discretionary component that may be responsive to change by 90 percent. An unchangeable hard-core will remain, and they are generally not the focus of public-policy efforts.) A government campaign to rectify any of these issues, if successful, might be able to alter 90 percent of the discretionary causes of the problem. But then it hits a wall, struggling to take it any further. Then only a new idea, often generated by a new cohort of politicians, can improve performance further.

So what does all this mean? I think it's significant for a few reasons. It means that the less severe a problem is when we first encounter it, the closer to perfection we can expect the end result to be. It means that we should judge our performance against the extent of the initial problem rather than some absolute measure. It means that one actor is not able to achieve the best result on their own, and is better working in coordination with others with the aim of improving upon each other's improvements. And it means that we should accept that perfection is impossible and we shouldn't be too hard on those who fail to achieve it.

This theory is borne exclusively from my own observations. It's possible there are some exceptional circumstances or people that confound it, but they would be few. It's also possible that the 90 percent figure is off the mark. I think of it as an average among people; some will be higher, others lower. It is also influenced by circumstances; time pressure or psychological pressure are likely to lower the figure. But for the most part, I think it explains plenty about why things are as they are.


Sarah Hobday-North said…
Interesting observations. And I think largely true by anecdotal evidence. In your last paragraph I think you started to knock another large nail on the head by introducing time as a factor in the achievement of perfection. When one has a deadline, there is only so much one can do, so one has to prioritise. Prioritisation is perhaps an alternative and more ambiguous synonynm for "the 90% rule".

My bulder and handyman husband will protest, for example, that our house is currently being renovated to a greater that 90% improvement on its original condition. We, however, do not have the constraint of time. In the construction industry (and others, I'm sure) there is a seminal diagram of Quality-Cost-Time: the idea being that a single project can not control all three in equal measure, but only two. So, if one wants a fast and excellent project, it will cost a lot. An excellent and budget conscious project will take at least five times at much time as you hoped.

Which sounds more like a 66% rule! Perhaps I should retrain as an editor! It sounds like the improvemnt rate is much higher!

Keep up the good work Ari.

Ari Sharp said…
Hi Sarah,

Thanks for the comment. External pressures are definitely a factor, but I fear that even with little or no pressure there is an upper limit to what we can achieve that is short of perfection.

In renovating your house, I'd admire your desire to reach beyond 90%, but I wonder whether, given the enormity of the task, you can be confident that every last nail is perfectly placed, that every brush stroke of paint is perfectly applied? Bravo if you can, but those would seem to be vulnerable to being overlooked.

Perhaps one other qualification I should have included is that it's possible for one person to achieve a 99 percent result, but it will be in at least two stages, with a gap between them such that they are separate efforts and the problem is confronted with fresh eyes.


Popular posts from this blog

Thanks for all the fish

A place to rest my head

The Real Bangkok Hilton