Only looking at the domains you sell will skew your opinion of domain name appraisals.
Automated domain name appraisals, such as those provided by Estibot and Godaddy, are extremely controversial in domainer circles. Focusing on cases in which domains sell for much more than the appraised value is an easy way to try to discredit these appraisals.
Consider this post that Mike Mann made on Facebook comparing his sales to GoDaddy’s appraisals:
He notes “All inclusive of our top sales, not selective.” Yet, by focusing only on his top sales, he is making a great example of selection bias.
Mike holds out for top dollar on his domains. So of course when he gets this top dollar, the number is going to be way off from an automated appraisal.
If someone appraises my domain at $10,000, and I refuse to sell it for less than $20,000, then the appraisal (whether automated or by a human) is going to seem wrong.
What about the domains that didn’t sell? Did they not sell because the price expectation was too high?
I recall a conversation at DomainFest many years ago. I was riding home from a party with a couple of other people, including a domain owner who makes lots of five and six-figure sales. That evening he had asked several people how much some of his domains were worth. These were all domains he had sold already. The domainers’ estimates were all over the place and didn’t match the sales prices.
This shouldn’t surprise anyone. Appraising domains is difficult.
Rather than focus on the accuracy of any one domain appraisal, I think people should use appraisals differently.
They should also not let selection bias color their impression of domain appraisals.