International Journal of Management Research and Social Science (IJMRSS) ISSN 2394-6407(Print) Volume 8, Issue 2, April – June 2021 ISSN 2394-6415(Online) 60 DOI: 10.30726/ijmrss/v8.i2.2021.82009 Three New Models for Ranking of Candidates In the Preferential Voting Systems Mohammad Azadfallah Researcher, Business Studies and Development Office, SAIPAYADAK (SAIPA after sales services organization), Tehran, Iran. E mail: m.azadfallah@yahoo.com Abstract — Election is the main challenge to the political and social science. In the meantime, in the literature, several methods to decide the winner of elections have been proposed; theoretically there is no reason to be limited to these models. Hence, in this paper, we assume three new approaches (1. election result prediction by pre-election preference information using Markov chain model [to identify the efficient electoral strategy for each candidate]. 2. Improved Borda's function method using the weights of decision makers [or voters]. And 3. A new interval TOPSIS-based approach applying ordinal set of preferences [so, data is ordinal form that first convert to interval value and then inject them into the conventional interval TOPSIS model]) for ranking candidates in voting systems. Ultimately, three numerical examples in social choice context are given to depict the feasibility and practability of the proposed methods. In sum, this paper suggests a mind line for decreasing the wrong choice winner risks correlated with voting systems. Keywords — Voting Systems; Markov Chain Model; Borda's Function; TOPSIS with Interval Data; Ordinal Preference; Ranking of Candidates Problems . 1. Introduction According to Alam, Mezbahuddin, and Shoma (2015), in the earth, election is very much liking word. When a group of people with individual preferences has to decide which alternative to choose from a given set of alternatives, an election is often carried out (Polykovskiy, Berghammer, & Neumann, 2016). Therefore, obtaining a group ranking or a winning candidate from individual's preferences on a set of alternatives is an important group decision problem with social choice and voting system implications (Aghayi & Tavana, 2019). In social choice theory, and more particularly in voting theory, a society needs to choose a candidate from a set of candidates (Bouyssou, Marchant, Pirlot, Tsoukias, & Vincke, 2006). Further, social choice theory is a field of scientific inquiry that studies the integration of individual preferences during a collective choice (Brandt, Couitzer, Endriss, Lang, & Procaccia, 2016). Meanwhile, a voting system uses the information provided by the voters in order to determine the elected candidate or, more generally, the decision made by the group (Bouyssou, Marchant, & Perny, 2009). On the other side, according to Kou and Sobel (2004), electoral outcomes in democratic countries have far- reaching domestic and (sometimes) international impact. Individuals, corporate actors, and governments who anticipate being affected by the outcome of a future election incorporate their expectations (forecasts) into current elections and policies. Outcome prediction of political events is an integral part of the practice and study of politics. In the social sciences, pure prediction models are used only for a limited number of problems, one of which is elections outcome (Stoltenberg, 2013). According to Alam, Mezbahuddin, and Shoma (2015), election prediction is very significant for the candidates and the society. Historically, from the 1970s onwards, wide ranges of forecasting techniques have been developed in the literature on electoral forecasting (Walther, 2015). Beck and Dassonneville (2015) believe that, scientific work on national election forecasting has become most developed for the United States case, where three dominant approaches can be identified: structuralists, aggregators, and synthesizers. For European cases, election forecasting models remain almost exclusively structuralists. These methods can be distinguished in terms of their application in theory, data, and time. The structuralists suggest a theoretical model of the election outcome. In contrast, the aggregators, aggregate vote intentions in opinion polls. Taking a different approach, the synthesizers borrow from both the structuralists and the aggregators. In addition, Payne (2001) believes that, three forecasting environments can be identified:  predicting the final result before the election takes place (the 'pre-forecast'),  immediately after the polling stations close (the 'prior forecast'),  During election night itself using the subset of actual results declared (the 'results-based forecast'). The concern here is with the first type of forecasting context (in other words, 'pre-forecast'), by using Markov chain model (as discussed latter in this paper). On the other side, in one hand, according to Ebrahimnejad (2012), in