Ode to green bar paper

Imagine trying to remember a particular point in a line of 100 points:

Even if the points were larger, you still wouldn’t be able to do it. People can’t accurately tell apart more than 9 or 10 positions on a line. In other words, people only remember about 3.25 binary bits of information about a linear position (log₂9.5).

Now let’s arrange the points in a square:

You can almost do it, can’t you? When we split the positional information into two channels – vertical and horizontal – we become more accurate. People can distinguish about 24 or 25 unique locations in a square, or about 4.6 bits of information. We don’t double the bits of accuracy, but this is still a significant improvement.

Now let’s add shading:

Piece of cake isn’t it? You can pick one of 100 dots, check your shoe size, then look back to find the same dot every time. Just by rearranging and coloring, we’ve improved accuracy by an order of magnitude. Now we can clearly distinguish 100 dots, or about 6.65 bits of accuracy.

A practical example

Imagine you are checking the spelling of the word “Mastigoteuthidae” in a printout using one of the lists below.

You’d have to switch your eves between the screen and the paper several times to check the spelling. Which of these three lists would be easier to use?

Ancistrocheiridae		Ancistrocheiridae		Ancistrocheiridae
Architeuthidae		Architeuthidae		Architeuthidae
Bathyteuthidae		Bathyteuthidae		Bathyteuthidae
Batoteuthidae		Batoteuthidae		Batoteuthidae
Brachioteuthidae		Brachioteuthidae		Brachioteuthidae
Chiroteuthidae		Chiroteuthidae		Chiroteuthidae
Chtenopterygidae		Chtenopterygidae		Chtenopterygidae
Cranchiidae		Cranchiidae		Cranchiidae
Cycloteuthidae		Cycloteuthidae		Cycloteuthidae
Enoploteuthidae		Enoploteuthidae		Enoploteuthidae
Gonatidae		Gonatidae		Gonatidae
Histioteuthidae		Histioteuthidae		Histioteuthidae
Joubiniteuthidae		Joubiniteuthidae		Joubiniteuthidae
Lepidoteuthidae		Lepidoteuthidae		Lepidoteuthidae
Lycoteuthidae		Lycoteuthidae		Lycoteuthidae
Magnapinnidae		Magnapinnidae		Magnapinnidae
Mastigoteuthidae		Mastigoteuthidae		Mastigoteuthidae
Neoteuthidae		Neoteuthidae		Neoteuthidae
Octopoteuthidae		Octopoteuthidae		Octopoteuthidae
Ommastrephidae		Ommastrephidae		Ommastrephidae
Onychoteuthidae		Onychoteuthidae		Onychoteuthidae
Pholidoteuthidae		Pholidoteuthidae		Pholidoteuthidae
Promachoteuthidae		Promachoteuthidae		Promachoteuthidae
Psychroteuthidae		Psychroteuthidae		Psychroteuthidae
Pyroteuthidae		Pyroteuthidae		Pyroteuthidae
Thysanoteuthidae		Thysanoteuthidae		Thysanoteuthidae
Walvisteuthidae		Walvisteuthidae		Walvisteuthidae

List boxes on a computer

So why don’t Microsoft and Apple shade list items in groups of three, or at least make it easier for programmers to do this?

I think they didn’t because of the cases where it wouldn’t be appropriate. For example, you would not want to shade in groups of three when individual list items can wrap for more than one line, or if the list box displays four items or less.

Still, it seems sad to have lost this simple convenience.

I recommend the article The Magical Number Seven, Plus or Minus Two: Some Limits on our Capacity for Processing Information by George A. Miller. It’s a fascinating paper.