You tell me

 My peak blog readership, now that I have many years of experience, was two weeks ago.  For some unknown reason my daily readership rises slowly from February through the end of April then it climbs rapidly to a peak in mid-May. After mid-May there is a slow decline for the rest of the year.

Is May the time when students are writing papers?

This brings up the question of 'how to calculate  readership?'  My blog service provides daily site visit totals and weekly totals as well.  I could know the actual regular readership if I knew the frequency of visits. 

For example, if 20% of all readers visit the site one or more times a week and half of those visited four times a week or more, the total readership would be 2.5 times the weekly readership, using a normal distribution curve. 

For the time being that is the distribution I am using in my own head, absent the real data.