No Good Days, No Bad Days── The choice not to correct every wobble

Is there really such a thing as a good day or a bad day? I have come to think there isn't. There is only variation. And the more you reach out to correct that variation, the wider it gets. This is not a matter of attitude. It is set down, rather bluntly, in an old quality-control textbook.

01The habit of grading a day

For years I kept a number in the corner of my notebook: how many findings I had raised that day. Three and it was a light day. More than ten and it was a heavy one. I even drew a small circle or a cross beside the figure. No one else ever saw it. I simply could not close the day without giving myself a mark.

One evening I laid out six months of those numbers. There was no pattern in them at all. Three findings one day, twelve the next, back to four the day after. The weeks when I was sharp and the weeks when I was short of sleep produced the same scatter. What made the difference was not me. It was whether the material that happened to reach my desk that week was a pre-launch piece for a new product or a small revision to an existing one.

The grading had a second effect, and this one troubled me more. If I marked a day heavy, I came in braced the next morning. My eye changed. Wording I would normally have let stand now drew a red mark. After a light day I arrived relaxed, and my hand went soft. The memory of the grade was bending the next day's judgment.

I stopped keeping the number once I saw that. What happened next was almost disappointing: nothing. The count went on rising and falling, heavy days and light days arrived as before. The only thing that changed was that I no longer spent the evening passing sentence on myself.

02"Every day is a good day" is not an instruction to make it so

There is a phrase in Zen: every day is a good day. Three characters, thrown at his students by a Chinese monk named Yunmen at the end of the Tang dynasty, and preserved as the sixth case in the Blue Cliff Record, the collection of Zen exchanges compiled in the Song period.

For a long time I read it as encouragement. Let us make every day a good one. A slogan for keeping your chin up. The reading appears to run the other way. What Yunmen is saying is that there is no distinction between good days and bad days to begin with. A clear day is not the good one and rain is not the bad one. Take the clear day as a clear day and the rain as rain, and each day is already enough as it stands. Not making the day good, but giving up the habit of scoring it. The order was the reverse of what I had assumed.

Zhuangzi wrote something close to this a good deal earlier. In the chapter usually called On the Equality of Things, he argued that distinctions — beautiful and ugly, right and wrong — are not fixed to things themselves. The one who looks attaches them afterwards. The distinctions do not disappear. But knowing that you are the one holding the label changes how far the label can push you around.

In Japan, Kamo no Chomei opened his Hojoki with the river that never stops and whose water is never the same water. I was made to memorize that line at school, and for years I read it as a lament about how quickly things pass. I read it differently now. It is an observation: moving is the ordinary state, and standing still happens nowhere. Wobbling is not the anomaly. The times when nothing appears to wobble are the accident.

03There are two kinds of wobble

Even so, you cannot let everything through as it comes. If a figure in the material is off by a decimal place, that is not something to accept; it is something to fix. So where is the line between the variation you take as it is and the variation you act on? What helped me here was not Zen and not philosophy. It was an old quality-control textbook.

In 1920s America there was an engineer named Walter Shewhart, who spent his working life looking at the scatter in manufactured parts. He split variation in two. One kind is common cause: the ordinary width a system carries by its nature, with no particular culprit behind it. The other is special cause: a sign that something out of the ordinary has happened. Here there is a culprit.

The split matters because the right response to each is the opposite of the other. Special causes you find and remove. Common causes you can search all day without turning up a suspect, for the plain reason that there isn't one. If you want less common-cause variation, you have to rebuild the system itself. Interrogating individual events yields nothing.

Back to the numbers in my notebook: nearly all of the distance between three and twelve sat on the common-cause side. The type of material, the habits of whoever drafted it, where the promotional calendar fell that month. None of it moved with how hard I worked. I had been running a nightly investigation into a crime with no perpetrator.

04The more you correct, the more it wobbles

Deming, the statistician, showed why you must not reach for common-cause variation, and he showed it as an experiment. It is known as the funnel experiment.

Draw a target on a table, hold a funnel above it, drop a marble. The marbles scatter around the target. So far, so ordinary. The next part is where it turns. Suppose a marble lands three centimetres to the right. Most people shift the funnel three centimetres left before the next drop. You found a deviation, so you corrected it. It looks like sound judgment.

The result runs the other way. The scatter under continual correction comes out distinctly wider than the scatter you get by never moving the funnel at all. Leave it alone and the marbles cluster.

A second illusion sits on top of this one. Daniel Kahneman wrote about regression to the mean. After an extreme result, the next one tends to fall back toward the average with no one doing anything at all. It is nothing more than a statistical property.

Consider what follows from that. Scold someone after a bad day and the next day improves. Praise someone after a good day and the next day drops. People arrive with no difficulty at the conclusion that scolding works and praise does harm. In fact neither moved anything. The variation simply came back to the middle. We manufacture the evidence that our intervention worked out of the variation itself.

05Draw the band before the fact

Telling common causes from special ones matters, I said. But if you tell them apart according to how you feel that day, you are back to grading. A heavy day looks special; a light one looks ordinary. The line moves to suit you.

Shewhart's answer was almost anticlimactic. Draw the band first. The control chart he devised is not much more than that: mark the ordinary width of the variation in advance, and move only when something falls outside it. Draw the line after the event and people will draw it wherever suits them. So draw it first.

Held alone, the band still drifts with your mood. Better to hold it as a team. When the sense of the band is shared, the line stops moving with whoever happens to be most senior in the room. "That's a lot today" and "is this inside the band or outside it?" lead to very different next moves.

Back to my notebook: those circles and crosses were lines drawn after the event. Because I had not set the band in advance, each night I redrew it according to how tired I was. I was not passing judgment. I was moving the line.

06Accepting is not giving up

Write "take the variation as it comes" and it sounds like resignation. Like an announcement that you have stopped trying. But accepting and giving up are quite different things.

There was a Japanese psychiatrist named Shoma Morita. Treating people who suffered from anxiety, he did not try to remove the anxiety. He did the opposite: leave it as it is, and go on doing what you had decided to do. Fight it and you end up keeping it company all day. Leave it be and it rises and falls on its own, and in time changes shape.

The same idea runs through present-day therapy. ACT, assembled by Steven Hayes and his colleagues, also begins by dropping the tug-of-war with your feelings. Leave the feeling where it is and move your behaviour toward what matters to you. Accepting here is neither endurance nor surrender. It moves the effort you were spending on changing a feeling over to what you actually do.

Translated to the desk, it comes out like this. There are a lot of findings today. Take that as a fact. Do not go hunting for a cause. But the checks you decided to run, you run today exactly as you ran them yesterday. Let the mood rise and fall. Do not let the procedure rise and fall with it.

07Ease is not the same as indifference

Having written this far, one thing still bothers me. Give up scoring the day, used carelessly, becomes an excuse for someone who cares about nothing. Miss something and you can call it variation. That is wrong.

The dividing line is clear. The signs that something is out of the ordinary must always be picked up. An event well outside the usual width. A drift that keeps running in one direction. Wording that reaches past the approved efficacy or safety. These are not to be accepted; they are to be stopped. Miss them and you are not at ease, you are simply negligent.

What you take as it comes is the common-cause side only. Your own condition today, the rise and fall in the number of findings, the fact that someone was in a foul mood in the meeting. It amounts to nothing more than declining to invent a culprit where there is none.

Seen that way, ease turned out not to be a matter of easing off. It was a matter of narrowing where the effort goes. Straining equally at everything, I missed more than I do now, straining at the one place that has to be caught. That surprised me. Loosening did not make me careless. Deciding where to loosen left something in reserve for the part that mattered.

Try giving up, once, on the attempt to make the day good. Whether today was a good day or a bad one turned out to be a question that never needed answering. Leave it unanswered and you do not come in braced the next morning. Unbraced, you can look at the page in front of you as it is, rather than through yesterday's score.

Yunmen's three characters are probably not encouragement, and not an instruction to make the day good either. They read more like a report: stop attaching the labels good and bad, and the day is already enough. Since I rubbed out the number in the corner of my notebook, my days have been a little quieter. The variation still runs at the same width. I have only stopped reaching for it at night.