Sorry, but this doesn't make sense to me. There has to be a boundary - a person who got all the answers wrong can't underestimate their performance, and a perso

Sorry, but this doesn't make sense to me. There has to be a boundary - a person who got all the answers wrong can't underestimate their performance, and a person who got all the answers right can't overestimate their performance. You could make the case that boundary effects are all DK is about, but that's not what the article is doing (and also I don't think such a claim is supported by the DK plot).