Catfished! Impacts of Strategic Misrepresentation in Online Dating

Oz Kilic, Alan Tsang

Abstract: Online dating is a multibillion-dollar global industry, and is increasingly becoming the go-to method for finding partners. Intricate dynamics mark its operation, influenced by varying user preferences, strategies, and traits, as well as by the underlying matchmaking algorithm. This complexity renders it a pertinent subject for multiagent systems research. Despite its relevance, an established simulation framework for online dating is lacking. This paper introduces a multiagent simulation framework for this domain. The framework is extensible and capable of modeling agents with diverse attributes and preferences, either reported or latent. It also supports varied strategies, outcomes, and types of matchmaking logic. Using this framework, we simulate an online dating platform based on real-world demographics to examine the effects of strategic misrepresentation, a notable concern in online dating. Surprisingly, the negative effect of strategic misrepresentation on users is marginal. Moreover, it disproportionately benefits female or honest agents more, enhances the overall welfare of the user population, and benefits attractive users - whether deceitful or not - over less attractive ones.