Convergence Analysis for Weighted Joint Strategy Fictitious Play in Generalized Second Price Auction

  • Lei Yao
  • Wei Chen
  • Tie-Yan Liu
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7695)


Generalized Second Price (GSP) auction is one of the most commonly used auction mechanisms in sponsored search. As compared to conventional equilibrium analyses on GSP auctions, the convergence analysis on the dynamic behaviors of the bidders can better describe real-world sponsored search systems, and give them a more useful guideline for making improvement. However, most existing works on convergence analysis assume the bidders to be greedy in taking actions, i.e., they only utilize the bid information in the current round of auction when determining the best strategy for the next round. We argue that real-world professional advertisers are more capable and can utilize the information in a longer history to optimize their strategies. Accordingly, we propose modeling their behaviors by a weighted joint strategy fictitious play (wJSFP). In the proposed model, bidders determine their optimal strategies based on their beliefs on other bidders’ bid prices, and the beliefs are updated by considering all the information they have received so far in an iterative manner. We have obtained the following theoretical results regarding the proposed model: 1) when there are only two ad slots, the bid profile of the bidders will definitely converge; when there are multiple slots, there is a sufficient condition that guarantees the convergence of the bid profile; 2) as long as the bid profile can converge, it converges to a Nash equilibrium of GSP. To the best of our knowledge, this is the first time that the joint strategy fictitious play is adopted in such a complex game as sponsored search auctions.


Nash Equilibrium Convergence Analysis Combinatorial Auction Potential Game Joint Strategy 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bhawalkar, K., Roughgarden, T.: Welfare guarantees for combinatorial auctions with item bidding. In: Proceedings of the Twenty-Second Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 700–709. SIAM (2011)Google Scholar
  2. 2.
    Cary, M., Das, A., Edelman, B., Giotis, I., Heimerl, K., Karlin, A.R., Mathieu, C., Schwarz, M.M.: Greedy bidding strategies for keyword auctions. In: ACM Conference on Electronic Commerce (2007)Google Scholar
  3. 3.
    Christodoulou, G., Kovács, A., Schapira, M.: Bayesian Combinatorial Auctions. In: Aceto, L., Damgård, I., Goldberg, L.A., Halldórsson, M.M., Ingólfsdóttir, A., Walukiewicz, I. (eds.) ICALP 2008, Part I. LNCS, vol. 5125, pp. 820–832. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  4. 4.
    Edelman, B., Ostrovsky, M.: Strategic bidder behavior in sponsored search auctions. In: First Workshop on Sponsored Search Auctions. ACM (2005)Google Scholar
  5. 5.
    Edelman, B., Ostrovsky, M., Schwarz, M.: Internet advertising and the generalized second-price auction: Selling billions of dollars worth of keywords. The American Economic Review 97(1), 242–259 (2007)Google Scholar
  6. 6.
    Marden, J.R., Arslan, G., Shamma, J.S.: Joint strategy fictitious play with inertia for potential games. In: Proc. 44th IEEE Conf. Decision Control (2005)Google Scholar
  7. 7.
    Miyazawa, K.: On the convergence of the learning process in a 2×n non-zero-sum two-person game. Econometric Research Program 161Google Scholar
  8. 8.
    Monderer, D., Shapley, L.S.: Potential games. Games Econ. Behav. (1996)Google Scholar
  9. 9.
    Nisan, N., Schapira, M., Valiant, G., Zohar, A.: Best-response auctions. EC 2011 (2011)Google Scholar
  10. 10.
    Varian, H.: Position auctions. International Journal of Industrial Organization 25(6), 1163–1178 (2007)CrossRefGoogle Scholar
  11. 11.
    Zhou, Y., Lukose, R.: Vindictive bidding in keyword auctions. In: Second Workshop on Sponsored Search Auctions. ACM Electronic Commerce (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Lei Yao
    • 1
  • Wei Chen
    • 2
  • Tie-Yan Liu
    • 2
  1. 1.Academy of Applied Mathematics and System ScienceChinese Academy of SciencesChina
  2. 2.Microsoft Research AsiaChina

Personalised recommendations