Author name: danhmcinerney

Last event went 1/4 in best predictions and the one win was essentially guaranteed with bookies setting the odds at -1000. This is because A) I’m not doing unit testing, and B) I introduced (and reintroduced) multiple bugs into the model via the latest release. It is painful to do hundreds of hours of coding […]

The model now includes a new engineered stat which measures the highest and lowest values in each stat column and also compared their current stat to their peak and valley. This almost quadruples the model’s total stats. Accuracy increased from 68.8% to 71.1%. Return On Investment has increased from 21% to 28% for the last

I had to take down the predictor.mma-ai.net web application that let you check the odds of any fighter based on the algorithm because it costs me $150 for an EC2 instance with enough memory, CPU, and space to make the predictions. This makes me sad because it was the first Flask app I’d ever developed