To solve the Gettier problem, Lehrer provides the defeasibility condition. The claimant appeals to anything false (false acceptance) in a justification game move, the skeptic is allowed to disqualify the move by the claimant. By this way, the claimant's justification is defeated. She lost the game to skeptic, and didn't know what she presented. Gettier's basic contention is that justified true belief may fall short of knowledge. Maybe the claimant has some justified true belief by luck. When the claimant depends on a false acceptance, the skeptic can ask her delete or replace it in her acceptance system. By Lehrer's theory, the claimant's justification is defeated. Lehrer explains the situation of knowledge attribution. Is the defeasibility condition a sufficient condition of knowledge? The critics give some Gettier-type counterexamples. In the examples of that kind, the claimant does need hot to base on the acceptance of false claim. Can Lehrer's theory explain those cases reasonably? In this article, I try to compare Lehrer's theory with his critic's, and give some comments in the end.