Conference scheduling is a challenge which is well known to organizers of academic conferences. Traditionally, the scheduling process is done as follows: after the excepted papers are collected, they are gathered to sessions according to the different topics of the conference. Then, the sessions are assigned to time-slots so sessions of the same track will not be scheduled parallelly. In recent years, a personalized approach has been suggested and tested in few conferences. In this scheduling approach, after all sessions have been settled, all participants are asked in advance to choose their k preferred sessions. Then, the sessions are assigned to time-slots with the objective of maximizing the opportunity of participants to attend their preferred sessions. Here we analyze the expected average attendance in preferred sessions of a participant as a function of the conference parameters: number of participants, number of sessions, number of time slots. We use tools of Extreme Combinatorics for the analysis. In addition, we conduct simulations for conferences with various parameter instances. The results can suggest schedulers who use the personalized approach to decide what number, k, of preferred sessions to let each participant to choose in advance and declare what they expect the average attendance to be.