To put these results in perspective, in the asymmetric static (dynamic) RNNE the mean winning bid is predicted to be 103.
We calculate the probability of an inefficient allocation occurring under the asymmetric RNNE at 16.
The raw bids and the theoretical RNNE predictions for the asymmetric states in both the complete and incomplete bid revelation treatments of the one-shot, two-period (round 1), two-period (round 2), and infinite-horizon games are depicted in Figures 2-5, respectively.
To formally compare bidding behavior to theoretical predictions, we estimate an econometric model using data from the asymmetric auctions, with the dependent variable being the difference between an observed bid and the relevant RNNE prediction.
The estimated coefficients presented in Table 5 represent the difference between observed bids and the relevant RNNE in each treatment setting.