Abstract: When observing techniques for mitigating fatigue, it can be seen that the process of tapering is the main method. This can be seen in studies such as (Ribeiro et al., 2023) that state raw performance increase of Taper is 3.06%. However, Taper does not specifically address worsening technique through the race that increases a swimmer’s fatigue. As such, this study creates a hybrid CNN-LSTM deep learning model that helps resolve this issue by outputting specific fatigue values through a race. The resulting outputs are then graphed over time, outputting a fatigue graph. These graphs are useful for analyzing specific trends in swimmers; allowing for a visual representation of their fatigue data that can help swimmers decide when to focus on holding their technique and when to accelerate in their races. As such, my results demonstrate that a hybrid deep learning model can be used to turn footage of elite swimmers into a fatigue graph that can be either an alternative or a complement to taper.