“the information is numerically redundant, but not visually redundant. And the point is, by keeping the series there, we can let the user put his huge visual cortex to work analyzing it.”

If we were to take your point that “the information is redundant” to the extreme, then we’d just end up with a table of numbers.

For sure, it could be considered redundant to show numerical values (as data labels) as well as numerical values (encoded in line lengths). So if it comes down to a choice between the two, which should we keep…the numbers or the images.

Quoting from Few’s “Now you see it” again:
* “Vision is not just one of the 5 channels through which we sense the world; it is by far the dominant and most powerful sense.

* “When we represent quantitative information in visual form, our ability to think about it is dramatically enhanced”

* “The eye and the visual cortex of the brain form a massively parallel processor that provides the highest-bandwidth channel into human cognitive centers” (Attributed to Colin Ware”

So if it comes down to a choice of one or the other, keep the graphic and lose the numbers. Fortunately we dont have to choose…we can have both.

The information is numerically redundant, but not visually redundant. And the point is, by keeping the series there, we can let the user put his huge visual cortex to work analyzing it.

“1. I think you sorted the sales data and not the commissions.”

Oops, my mistake. I have corrected the version and uploaded it here: http://chandoo.org/img/cb/chart-busters-sales-chart.png

2. You remove a visual analysis from the chart by using labels only to encode the percentages.

Hmm, While I agree that I have removed a series, you can easily calculate % given the two values. The information is redundant.

@Colin

“Chandoo, the chart doesn’t work. I think that it would be OK if we were showing only Sales and Commissions (and the numbers were formatted in $K). However, incorporating two measures on a bar results in a whole pile of confusion. It isn’t obvious at all what (7%, 2000), (4%, 1500) etc. mean as combined measures. Even less obvious is why you would merge the commission and ratio into one thing – especially when there’s no direct correlation between the two measures.”

If the confusion is due to the fact that I have sorted sales but not commissions, then It is my mistake. I have corrected it now.

There is direct correlation between % and commission.

1. I think you sorted the sales data and not the commissions.
2. You remove a visual analysis from the chart by using labels only to encode the percentages.

I was looking at the data and thinking may be we can use the good old bar. Here is a little something I have made

http://chandoo.org/wp/wp-content/uploads/2009/07/sales-data-chart-busters.png

Since the sales commission will have to come out of sales revenue, overlapping series works without confusion. As the % is just a calculation of commission / sales, I have included it in the labels.

What do you think?

As for the “red” and “green” colors on the chart, the colors used aren’t primary colors, so it’s unclear how a color blind person would perceive them.

Don’t see how. For example, in the “Difference in Total Percent” chart, it isn’t obvious to me which number (or numbers) you want to draw the reader’s attention to. Precisely because of the grading, my eyes don’t immediately focus on anything in particular. The point of the grading is even less clear in the other charts.

“In this example, I feel the coloring helps bring out Bob’s Commission value as something to be looked into.”

Even after staring at this chart for a while, that wasn’t obvious to me. On the other hand, if Bob’s commission bar was in a color different from the others, I would have immediately been drawn to it.