The Magic Number Seven Plus Or Minus Two

Using GTD and OmniFocus requires us to look at lists all day long. In the 1950s, psychologists discovered chunking, the mental process of dividing items into groups small enough to be comprehended and remembered.

The landmark work on this subject is an article called The Magic Number Seven Plus Or Minus Two. In it, they explain that most people can deal with a list of 5-9 items (= 7 ± 2) before their brains refuse to cope. Everyone has their own “chunk-size.”

If you’ve ever looked at a context list in OmniFocus, and it’s just appeared like a block of letters, you’ve exceeded your chunk-size. The solution is to break the list down into smaller chunks, and your brain will process them better.

I’m bringing this up because I learned about this many years ago, and decided to test my chunk-size this week. My context lists are sliding off my brain like it was Teflon® coated, so I wondered if I could fix it with smaller chunks.

I tested my chunk size by making lists in OmniOutliner starting with nine items, then chunking them down into seven, then five items. I did this with three sets of different data, including actions, things and ideas. I determined my upper bound in chunk size in about five minutes. Let’s call that TheNumber.

Now I’m making a rule that no context list will exceed TheNumber, and I can’t view more than TheNumber of contexts at once. If I do, I use flags, focus and filters to reduce the data below TheNumber.

So far, this is making GTD much more pleasant and less overwhelming.

I hope it helps some of you, too.


This is an interesting concept.

So have you whittled your entire listing of contexts to be no more than TheNumber? Or are you just cleverly using Perspectives and such to ensure that you’re dealing with logical subsets of data that might otherwise be greater than TheNumber?

I’ve had a long-standing friction with GtD and Contexts. I try to keep everything to as few as I can get away with but as many as I need. And it’s helped. But there’s still a bit of an undefined friction there which I haven’t been able to pinpoint yet.

Cleverly using Perspectives. Also, I use a hack for creating a secondary flag. I put a predefined reserved character (like ¥) in the notes of tasks I want to use, then I put the ¥ in the search box and it filters to it. So I whittle down my action list down to TheNumber x theNumber of contexts.

I’m trying to limit any agenda to TheNumber^2 tasks.

Interesting concept indeed, will certainly give it a go.

How did you approach making your lists in OmniOutliner?
Did you start from your existing tasks or just made “random” lists?

I just made random lists. I made a list of exotic sports cars and did the chunk test, then I made a list of tasks I don’t enjoy doing and did the chunk test, and again with kinds of philosophy. I found that regardless of the content, my chunk size was about the same.

Interesting. I didn’t think about it that way. I thought you were staying focused on productivity aspects.

The point was to test my chunk number. I did it with several different kinds of content so I could make sure the content wasn’t affecting the outcome.

1 Like