Aug 25, 2017 · 1 min read
Hey, it might be a bit late question but still. Lets say the average number of visitors is 5000. Yesterday you got 200, what would identified as an alarm. Today you got 200 again and that would not be identified as an alarm because you compare current day with the previous day. Imagine, something was broken on the website and therefore only users with one browser could access it. With the hack you would detect only one alarm (but lets imagine you assumed that it was due to a sunny sunday). So, the subsequent anomalies wount be alarmed that could help you to identify the problem with the website and browsers. Why just not to find the mean difference of the current day and all other days?
