The emergence of new technologies and new means has brought unprecedented opportunities for college teaching, and the traditional college teaching mode can no longer meet the current needs of English teaching. This paper analyzes the current English teaching mode in colleges and universities by using the S-T teaching analysis method and constructs the S-T behavioral sequence of English teaching in colleges and universities by coding the samples in classroom teaching videos. Utilizing Markov chain to evaluate the drawbacks of traditional college English classroom teaching enables reflection and modification of English classroom teaching methods and specific teaching links. Classify English classroom teaching behaviors according to S-T teaching behavior coding, establish a link between teachers’ comprehensive control level in the classroom and the area of the teaching behavior cloud, and judge teachers’ comprehensive control level according to the area of the cloud. The proposed teaching reform strategies are based on an analysis of English teaching problems in colleges and universities. The results show that under the reformed English teaching, students’ English performance has been transformed for the better, and Class 7 has the highest English performance, with an average score of 140.73 and a score grade of A. In each teaching case extracted, the teacher’s active interaction with the students is kept above 55 times, and the number of student’s active interactions with the teacher is above 65 times.