One way that you can do something to force this is to make separate data sets, out of a single data set.
Say, for example, that you have a frequency that “should” happen 5 times a day. Sometimes it happens more often (and that’s a good thing) and sometimes it happens less often (which is a bad thing). So the more-frequent occurrences should be shown in Green, the less-frequent occurrences in Red, and all ‘normal’ occurrences (or in an “acceptable range” around the 5 per day) should be shown in yellow.
Your raw data can be arranged by “Date” and “Number of Occurrences”. Then you need to set up three additional columns, which would be the graphed information. To break it down,
Col. A = Date
Col. B = # of Occurrences
Col. C = Too Low: =if( b2 < 5, b2, null())
Col. D = Just Right: =if( b2 = 5, b2, null())
Col. E = Extra Occurences: =if( b2 > 5, b2, null())
This is very simple; you can make the data sets as complex as work for you.
Now you graph only Columns C, D and E against Col. A (date), and use the appropriate color for each data set.
And there you go: One set of data broken into three virtual series.