Appraising one's own performance after a task, known as self-assessment, has been studied from a cognitive science perspective in domains such as humor, trivia, and logic. Previous studies have found that participants are systematically poor at judging their own performance, though sometimes self-assessment varies based on actual performance. We explored calibration of self-assessment on algebra problems, a domain where people have typically received explicit instruction. In this domain, we found that people do not behave as they do in other domains previously studied: they are generally well-calibrated in judging their algebra performance. This suggests that in the course of learning to solve algebra problems, people have also learned to accurately judge their performance, both absolutely and relative to others.