Efficiency of Non-Truthful Auctions in Auto-bidding: The Power of Randomization
Christopher Liaw, Aranyak Mehta, Andrés Perlroth
摘要
Auto-bidding is now widely adopted as an interface between advertisers and internet advertising as it allows advertisers to specify high-level goals, such as maximizing value subject to a value-per-spend constraint. Prior research has mainly focused on auctions that are truthful (such as a second-price auction) because these auctions admit simple (uniform) bidding strategies and are thus simpler to analyze. The main contribution of this paper is to characterize the efficiency across the spectrum of all auctions, including non-truthful auctions for which optimal bidding may be complex. For deterministic auctions, we show a dominance result: any uniform bidding equilibrium of a second-price auction (SPA) can be mapped to an equilibrium of any other auction – for example, first price auction (FPA) – with identical outcomes. In this sense, SPA with uniform bidding is an instance-wise optimal deterministic auction. Consequently, the price of anarchy (PoA) of any deterministic auction is at least the PoA of SPA with uniform bidding, which is known to be 2. We complement this by showing that the PoA of FPA without uniform bidding is 2. Next, we show, surprisingly, that truthful pricing is not dominant in the randomized setting. There is a randomized version of FPA that achieves a strictly smaller price of anarchy than its truthful counterpart when there are two bidders per query. Furthermore, this randomized FPA achieves the best-known PoA for two bidders, thus showing the power of non-truthfulness when combined with randomization. Finally, we show that no prior-free auction (even randomized, non-truthful) can improve on a PoA bound of 2 when there are a large number of advertisers per auction. These results should be interpreted qualitatively as follows. When the auction pressure is low, randomization and non-truthfulness is beneficial. On the other hand, if the auction pressure is intense, the benefits diminishes and it is optimal to implement a second-price auction.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了最后一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper9
- Optimal Type-Dependent Liquid Welfare Guarantees for Autobidding Agents with BudgetsRiccardo Colini-Baldeschi, Sophie Klumper, Twan Kroll, Stefano Leonardi 等SODA 2026 · 被引用 7 次
- No-Regret Online Autobidding Algorithms in First-price AuctionsYilin Li, Yuan Deng, Wei Tang, Hanrui ZhangNeurIPS 2025 · 被引用 4 次
- Non-uniform Bid-scaling and Equilibria for Different Auctions: An Empirical StudyYuan Deng, Jieming Mao, Vahab Mirrokni, Yifeng Teng 等WWW 2024 · 被引用 4 次
- Randomized Truthful Auctions with Learning AgentsGagan Aggarwal, Anupam Gupta, Andrés Perlroth, Grigoris VelegkasNeurIPS 2024 · 被引用 3 次
- Autobidding With Interdependent ValuesMartino Banchio, Kshipra Bhawalkar, Christopher Liaw, Aranyak Mehta 等WWW 2025 · 被引用 2 次
它引用的顶会 Paper4
- Towards Efficient Auctions in an Auto-bidding WorldYuan Deng, Jieming Mao, Vahab S. Mirrokni, Song ZuoWWW 2021 · 被引用 87 次
- Auction Design for ROI-Constrained BuyersNegin Golrezaei, Ilan Lobel, Renato Paes LemeWWW 2021 · 被引用 55 次
- Auction Design in an Auto-bidding Setting: Randomization Improves Efficiency Beyond VCGAranyak MehtaWWW 2022 · 被引用 41 次
- Why Do Competitive Markets Converge to First-Price Auctions?Renato Paes Leme, Balasubramanian Sivan, Yifeng TengWWW 2020 · 被引用 36 次
相关 Paper
- Efficiency of Non-Truthful Auctions in Auto-bidding with Budget ConstraintsChristopher Liaw, Aranyak Mehta, Wennan ZhuWWW 2024 · 被引用 10 次
- Risk-Averse and Optimistic Advertiser Incentive Compatibility in Auto-biddingChristopher Liaw, Wennan ZhuICML 2026 · 被引用 1 次
- Simple Mechanisms for Welfare Maximization in Rich Advertising AuctionsGagan Aggarwal, Kshipra Bhawalkar, Aranyak Mehta, Divyarthi Mohan 等NeurIPS 2022 · 被引用 8 次
- Equilibria in Auctions with Ad TypesHadi Elzayn, Riccardo Colini-Baldeschi, Brian Lan, Okke SchrijversWWW 2022 · 被引用 6 次
- Efficiency of the First-Price Auction in the Autobidding WorldYuan Deng, Jieming Mao, Vahab Mirrokni, Hanrui Zhang 等NeurIPS 2024 · 被引用 1 次
