| Fighter 1: | |
| Date of Birth: | |
| Wins: | |
| Losses: | |
| Height: | |
| Weight: | |
| Reach: | |
| Stance: |
| Fighter 2: | |
| Date of Birth: | |
| Wins: | |
| Losses: | |
| Height: | |
| Weight: | |
| Reach: | |
| Stance: |
| Winner: | |
| Confidence: |
This application uses a machine learning model to predict outcomes of UFC bouts. This model was trained on UFC bouts occurring in or before 2022, while 2023 and 2024 bouts were used as a testing set.
The model takes as input the attributes of two fighters, along with their fight histories. Fighter attributes include age, weight, height, reach, etc. A fight history takes the form of a list of fights, where each fight contains statistics such as significant strikes attempted, takedown defense, knockdowns landed, etc.
After performing a series of calculations on the input, the model provides an output between 0 and 1. An output of 0 means a prediction that the first fighter wins with 100% confidence, an output of 1 means a prediction that the second fighter wins with 100% confidence, and an output of 0.5 means a toss-up.