According to the world's most robust Bayesian regret calculations, IRV is actually significantly better than Plurality Voting. Although it's roughly equivalent

According to the world's most robust Bayesian regret calculations, IRV is actually significantly better than Plurality Voting. Although it's roughly equivalent in quality to the top-two runoff (TTR) system. Condorcet methods are typically even better, but NOT "closest to ideal". The best system (if we exclude "exotic" varieties that are too complex to be practical) is Score Voting aka Range Voting. http://ScoreVoting.net/BayRegsFig.html http://ScoreVoting.net/BayRegsFig.html It's also plausible that Score Voting and Approval Voting are, in practice, better Condorcet methods than real Condorcet methods. http://ScoreVoting.net/AppCW.html http://ScoreVoting.net/AppCW.html Clay Shentrup Co-founder, The Center for Election Science