The world of mobile devices which have the quality to stay connected to the web is going through a exponential growth, but in parallel with this event, appears the threats with the developing of an equal magnitude, which are segmented primarily by the characteristics of the target device, therefore, as an final user is very difficult to know how exposed it is to use or purchase one of these devices. Those, the present work aims at obtain and understand a mathematical formula to helps determine the probability that a mobile device should be infected, considering the threats that surround it by operating system and their version. This research is based on numerical data obtained through extensive investigation, the quantity of devices with Android operating system and their respective versions, to contrast with the amount of malware developed for each version, to thereby obtain a probability of recognition the device to be tested.