This thesis examines Ordinal One-Sided Matching Mechanisms, that allocate indivisible resources. We focus on two important mechanisms: Random Priority and Probabilistic Serial. Random Priority is safe from player manipulation (truthful), but often leads to suboptimal fair and efficient results, while Probabilistic Serial produces efficient results, but is vulnerable to manipulation.
We study several concepts such as social welfare, efficiency, truthfulness and analyze the performance of these mechanisms. We study the Approximation Ratio, which quantifies how close the outcome of a mechanism is to the optimal solution in terms of social welfare; the Price of Anarchy, which measures how the players' strategic behavior affects the efficiency of the mechanism; the Incentive Ratio, which expresses the maximum benefit a player can gain by manipulating her preferences compared to her true ones.
Finally, we present experimental results and discuss a new metric to assess how agents' strategic behavior affects the mechanism's social welfare relative to its outcome if everyone were telling the truth.