We can easily find out if a color-based field is used as discrete or continuous. Blue shows a discrete field, whereas green shows a continuous field. We are not alone if our mind is blown because we always supposed those colors represented a dimension or a measure of the field. The idea that blue is measured and green is the most popular legend in TABLEAU. It is simple to comprehend why dimensions are classified by default as separate factors and have in the SHELF sizes a tiny blue ICON facing them. Measures are divided by continuous variables so that in the measuring panel they are prefaced with a green icon.

Color coding identifies discrete rather than ongoing areas rather than measures. In fact, measures can be used as discrete areas or as ongoing, and the same applies to certain aspects like dates.

There are two laws of thumb. Discrete FIELDS draw headers; continuous fields draw axes.

See the following visualizations, which examine each month's revenues. I use the date as a separate field on the first graph.

Note that for each month there is a discreet header. I am using the same precise information in the second graph, but the dates aspect has been altered from discrete to continuous:

We have a permanent axis of time, as we can see. Since the axis is continuous, we can not alter the order of the dates; from the earliest date on the left to the latest date on the right they follow the chronological order. On the other hand, we can vary the order of the dates if the dimension of the date is used as discreet as shown in the first IMAGE. We could, for instance, sort the bars down, first with the largest sales of the month and the last with the smallest sale of the month. Whether or not we should use an area as discrete or ongoing brings me to my second rule: discrete fields can be sorted, continuous fields can not be sorted.