We advance a complexity-grounded, quantitative method for uncovering temporal patterns in CSCL discussions. We focus on convergence because understanding how complex group discussions converge presents a major challenge in CSCL research. From a complex systems perspective, convergence in group discussions is an emergent behavior arising from the transactional interactions between group members. Leveraging the concepts of emergent simplicity and emergent complexity (Bar-Yam 2003), a set of theoretically-sound yet simple rules was hypothesized: Interactions between group members were conceptualized as goal-seeking adaptations that either help the group move towards or away from its goal, or maintain its status quo. Operationalizing this movement as a Markov walk, we present quantitative and qualitative findings from a study of online problem-solving groups. Findings suggest high (or low) quality contributions have a greater positive (or negative) impact on convergence when they come earlier in a discussion than later. Significantly, convergence analysis was able to predict a group´s performance based on what happened in the first 30-40% of its discussion. Findings and their implications for CSCL theory, methodology, and design are discussed.