Percentage change from zero to anything
Something interesting I came across recently. How do we calculate a percentage change from zero to a positive number?
Something interesting I came across recently. How do we calculate a percentage change from zero to a positive number?
Context
We have a daily report running on vulnerability metrics. It’s just a quick glance for our stakeholders on any major changes from the previous day. It helps them to spot things early or query any inaccuracies.
Problem
One of the things reported is the percentage change in risk score from the previous day. For example, if Windows updates were applied to a batch of Windows laptops, we might expect to see a 10% decrease the next day in the overall risk score for Windows Laptops.
My team noticed something odd. One of the groups had a risk score of zero reported yesterday, we know why and that was expected so that’s fine. Today though, the risk score is not zero so we would expect to see a percentage change in risk score between yesterday and today. BUT, the percentage change in risk score is reported zero. That raised an interesting 5 min debate amongst us.
- It can’t be 0% change from yesterday as today there is a score where yesterday there was nothing. So that is something. Right?
- so is it 100% change because it went from nothing to something?
- Or is it infinitely more? Well it can’t be infinite because there is a definite number today…..right?
How do we get the formula to account for all of the above? Or is this actually a programming issue?
For now, this is such a rare occurrence in our reporting, we agreed to explain the anomaly rather than spend too long on debating it. There seems to be no right or wrong answer.