I think it would be interesting to run an experiment with this model where some buyers were shown what the last buyer of the game paid and another group of buyers were shown what the average price paid for the game was. My hypothesis is that the group seeing the average price would pay very close the average and have a tight distribution of payers. I think the group being shown what the last buyer paid would have a very wide distribution of payers but I have a hunch that they would see a higher average price paid due to some people seeing a high last price paid and saying, well I should at least pay X (I could comfortably get by with paying this much less) or seeing a low last price paid and saying I'll at least pay Y (I'm a good person for paying above that). Regardless, it would be an interesting experiment.