Machine Learning Fantasy Sports

Predicting Masters Golf Fantasy Winner with Machine Learning

With the 2018 Masters Golf Tournament right around the corner in April, we are getting a jump on working on our machine learning model to predict the best Masters fantasy golf lineup for the 2018 Masters.

Our testing has gone very well and our machine learning model, for predicting the top finishers in the Masters, has performed well.

To help you understand how our machine learning model works, we have gathered individual player data from the tournaments prior to the 2016 Masters.  We then gathered how these players placed in the 2016 Masters.  We used this data to train and test various machine learning algorithms.  Based on this training and testing, we identified an algorithm that works very well.

We then gathered the same data for the players prior to the 2017 Masters, but withheld the 2017 results.  The machine learning algorithm was very accurate at predicting top finishers in the 2017 Masters.  Here are some examples:

End Train 2016 - Predict 2017 Results
                name     predict  score   place - actual
5       McIlroy-Rory  286.281815    285     7
9      Kaymer-Martin  286.365187    288     16
14      Dufner-Jason  286.858231    293     33
1        Rose-Justin  287.244477    279     2
6         Moore-Ryan  287.265230    286     9
0      Garcia-Sergio  287.319196    279     1
11         Day-Jason  287.371883    290     22
8      Fowler-Rickie  287.472613    287     11
10         Perez-Pat  287.701632    289
4         Casey-Paul  288.703321    284
2   Schwartzel-Charl  289.105086    282
12    Mickelson-Phil  289.682863    290
13   Snedeker-Brandt  291.772499    291
17          Na-Kevin  295.688431    303
7      Spieth-Jordan  296.284492    287
3        Kuchar-Matt  296.638095    283
15      Stuard-Brian  297.912767    294
18      Willet-Danny  300.038457    307
19         Els-Ernie  300.295266    308
16     Leishman-Marc  300.706434    296
DONE!

The predictions are ordered by the score predicted and you can see the actual 2017 Masters score for the players to the right. The predictions are very accurate, but we think the ordering compared to how they placed is amazing.

As you may recall Sergio Garcia won the 2017 Masters and Justin Rose placed 2nd.  The system picked both of them in the top 6 which would have landed them on our Masters fantasy roster.

We are waiting until we have more tournament data closer to the Masters to make our 2018 Masters fantasy machine learning predictions. Stay tuned.

here is a link to the 2017 Masters results

2 Comments

  1. Wayne Crawford

    March 28, 2018 at 6:46 pm

    Do you guys have a github page with data/code that you are willing to share?

    • admin

      April 1, 2018 at 5:30 pm

      We are still in the testing phase currently, but once we are satisfied with the results, we will be making predictions and code available to premium members.

Leave a Reply

Your email address will not be published. Required fields are marked *