Welcome gamers and players,
today I’ll be short, because I want only to give attention to something strange that happens every time we give judgement.
Do you know what is the Bayes’s Theorem? It’s the basic for all the statistic. It give you a way to re-calculate your chance while your knowledge grow (if fact, I know it very well. It was the pillar of my degree thesis).
This is the theorem in its simplest form
In this statement, P is for probability, A is for an event, and B is for another event. In other terms: how probably is that A happens, if B is happened? Obviously, B and A should have some kind of relation between them, and that relation is written in the second part of the formula.
There are many different interpretation of this theorem, but it rightness is undoubted. Mine interpretation (which is also the Bayesian interpretation) is that B is not simply and event: it’s a collection of events and ideas: it’s a background. But I don’t want to write too much about how to use Bayesian epistemology.
I want to point it out a simple fact: the theorem and the Dutch book tell us that assuming probabilities in real-time, even if done correctly, cannot guaranteed a correct statistical outcome. The only way to keep your inference correct during a project, with changing probabilities, is to assign a “first” probability and them keep them upgraded through the theorem.
So, always give your number before the start, and keep them up-to-date!
This post is, simply, to explain why I write so much about how calculating Roi for gamification before starting a project, and with a lesser need of historical data, compared to usual Roi-calculation-strategies.