Building a consensus rule for voting in small groups with an integer scale: A note on the case of E-Cost.

We provide the assessment of voting in small groups with integer scales and suggest to nd a solution to build a consensus report. This is specially usefull for the assessment of E-Cost processes and building consensus reports for the European Commission. The rules are that we have k projects, 3 referees, and vote for each category from a scale of 1 to 5. Meaning that 1 is lowest and 5 the highest. We simulate and nd a solution which yields that maximum of the mode and arithmetic average is robust to nd and build and draft a consensus report. This solves extreme discrepancies on small voting with extreme valuations by dierent referees and their dierent perception of reality. The result is applied to E-Cost assessments, but is robust to any small voting with small groups of integer voting of any jurees. Namely, assessment of academic proofs, convergence of highschool of grades, jurees of artistic contests, just to name a few.