Abstract: The talk will focus on the paper “Auctions between Regret-Minimizing Agents” (WWW ’22, see the abstract below) but will also touch on broader topics on learning dynamics and on the strategic consideration of the users of learning agents, which extend beyond auction games.
We analyze a scenario in which software agents implemented as regret-minimizing algorithms engage in a repeated auction on behalf of their users. We study first-price and second-price auctions, as well as their generalized versions (e.g., as those used for ad auctions). Using both theoretical analysis and simulations, we show that, surprisingly, in second-price auctions the players have incentives to misreport their true valuations to their own learning agents, while in the first-price auction it is a dominant strategy for all players to truthfully report their valuations to their agents.
See the paper on arXiv: https://arxiv.org/abs/2110.11855 as well as a related companion paper: https://arxiv.org/abs/2112.07640.