Lets say that we have an online auction where various known sellers and known buyers exchange product X. A seller will post product X and each seller will then bid in accordance with X and if they want it. The highest bidder would win X. Think of something like Ebay, though the auction I'm thinking of has a set number of known buyers and known sellers, and occurs in a must faster time period (1-2 minutes).

Obviously there is a lot of uncertainty which cannot be accounted for in a model that tries to predict the probability of winning as a buyer in such a scenario. However, I was wondering what were some of the theoretical and practical limitations in trying to use statistical models to predict outcomes in a online auction system as described above.

EDIT: I'm specifically thinking of a model which is trying to predict the likelihood of winning based on our bid as a buyer.

  • $\begingroup$ Do you have access to previous auction results for product X ? $\endgroup$ May 11, 2012 at 18:56
  • $\begingroup$ Yes, I have access to previous auction bid and results for product X. $\endgroup$
    – ATMathew
    May 11, 2012 at 19:51

1 Answer 1


I think you might want to take a look at An Introduction to the Structural Econometrics of Auction Data by Harry J. Paarsch and Han Hong (link to book publisher) and at Structural Econometric Methods in Auctions: A Guide to the Literature by Hickman et al. (working paper version). Depending on the details of the auction and what data you observe, this book and paper should get you pretty far.

  • $\begingroup$ Hi Dimitriy, If you could add full citations, I think that would be helpful. Also, the second looks like a google link to the document of interest rather than the direct link to the PDF. Cheers. :) $\endgroup$
    – cardinal
    May 12, 2012 at 15:59
  • 1
    $\begingroup$ Fixed! I was posting from a phone earlier, which makes formatting difficult. $\endgroup$
    – dimitriy
    May 14, 2012 at 21:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.