An experiment analyzing behavior in English common value auctions is reported. English auctions raise more revenue than first-price auctions only when bidders do not suffer from a strong winner's curse. Agents employ other bidders' dropout prices along with their private information as Nash bidding theory predicts. However, a simple and natural signal-averaging rule, that does not require recognizing the adverse-selection effect of winning the auction, better characterizes the data than the Nash rule. Monte Carlo simulations using FIML estimates of the signal-averaging rule predict a number of data characteristics not directly employed in the estimation procedure.