Today, many countries around the world, particularly in Africa, are experiencing post-election difficulties due to unexpected
election results. This sometimes provokes protests and revolt among the population. To overcome this major problem, several
voting systems have been developed in the literature, but some of them are not lacking in shortcomings. It was with this in mind
that the voting method based on the evaluation of the mean deviation was born. It's a voting system that seems to be appreciated
because it respects a certain number of fundamental properties of a ranking method. On the other hand, we note in the literature
that it is only applicable to small-scale data with an insignificant number of candidates and voters. For this reason, we set
ourselves the goal of implementing this method in order to extend its use to large-scale problems. Thus, we proposed the
computer program using python software, which takes as input the scores assigned to the candidates by each voter and displays as
output the best candidate. To do this, we built sub-programs such as median, arithmetic mean and mean-spread functions, each of
which plays an effective role in selecting the best candidate. We then studied the algorithmic time complexity theoretically, then
graphically, and ended by applying our computer program to several voting examples containing a very large number of
candidates and voters. Numerous applications enabled us to observe that, whatever the size of the data, we always obtained a
conclusive and satisfactory result with polynomial-type time complexity.

Implementation, Voting Method, Mean- Deviation, Election