Editor’s Note: This is the second in a series of fight predictions by guest columnist Jason Noriega. Jason works as an analyst at a U.S. National Laboratory. He has a keen interest in predictive modeling, and lectures around the United States. Jason also participates in predictive modeling competitions and has placed in the top 10% in a number of them using various machine learning algorithms.
The predictive model developed by Jason Noriega for Jason’s Picks is the copyrighted intellectual property of Jason Noriega, Nery Castillo-McIntyre, and Artes Marciales Mixtas en Español.
The biggest assumption in predictive modeling is that the past is a good predictor of the future. Figure 1 displays the process followed at a high level. This involves taking historical MMA data and applying an advanced algorithm to develop a model. The model can then be applied to future MMA events to get predictions.
Figure 1: Artes Marciales Mixtas en Español Predictive Modeling Process
For UFC 142 in Rio, I used the outcomes of 42 historical UFC fights in order to develop my statistical model. I used the CHAID algorithm to predict who would win and lose (CHAID stands for Chi square – or Χ2 – Automatic Interaction Detector). The event was a good first test for the model’s performance. It predicted correctly 4 out of the 5 fights in the main card, with an accuracy of 80%*.
*I classified the Erick Silva vs. Carlo Prater fight outcome as correct due to Silva’s dominant performance and the controversial DQ by the referee.
OVERALL ACCURACY YEAR TO DATE: 80%
For the next event, UFC on FX in Nashville, TN I will use and test the performance of a new machine learning algorithm called Random Forest. If performance declines, I may revert back to CHAID. Below are my picks based on the probability of winning.
† I must confess this particular prediction makes me skeptical of the model, as I believe Barry’s knockout power combined with Morecraft’s relatively suspect chin should have yielded a prediction favorable to Barry; we will find out Friday night!
As the year progresses I will make more predictions and update the overall accuracy “rating.”