May 20, 2021

Data Analytics


Have you ever wondered how some people keep on winning in their Fantasy football league while some even after putting in time and efforts do not do all that well? If you are playing for money, wouldn't you like to improve your odds of winning? The only differentiating factor in both cases would be the decision-making process. 

Most of the decisions when it comes to creating a fantasy team are based on intuitions. One of the rookie mistakes almost everyone makes is selecting the player based on our personal preferences and biases. While it could give you a good feeling while drafting the team, the chances of winning with such a policy would be very difficult. 

Take an example suppose a running Back dropped out of the game due to a long-term injury now you need to quickly decide on a replacement for him, you go through the list, scan the stats of the player briefly and you connect with one of the players because you like his attitude and take a snap decision and add him to your roster. And it could be just this one decision that could cost you a lot of money. 

An emotional decision may even work for some but a data science-based approach has a better chance of working and provides a reliable method for gauging the potential of the player in the long term. While gut feelings would fluctuate from time to time but Data based approach based on fantasy football stats is tried and tested and would result in a better roster most of the time. 

Fantasy Football League AnalysisMany league leaders have realized the value addition provided by employing data science in the decision-making process and the fantasy football team has relatively performed better over the course of the season and even amounted to substantial winning as well with an almost unbeatable system. 

Here one could even argue that the knowledge you have gained over time by tracking every sports event and studying the stats in depth can also result in a win, I wouldn't argue with you because after all we are not focused on winning once but to come up with a more upgraded system of selection of players that would consistently deliver results rather than having a lucky hunch. 

Now coming on to the next part of the process, which method to adopt for selecting the fantasy football roster, some track past player performance but adopting a predictive approach would be better where you would be able to rank the players according to how they would perform on the field. This method is called propensity score matching. 

Football league with Data ScienceCreate Data Points. While the performance in the current season could offer some insight about the player but it would lack depth and data points that could truly be relied upon while decision making. We need to move beyond the current season is the key, factor in the results of other NFL players who were in an identical situation. Suppose we want to know more about a veteran player who ran 1,200 yards in his 5th season we would not just factor in his stats but also of other age-appropriate running backs like him. The main aim is to create more relevant data points that could serve as a yardstick for his future performance. 

Suppose the player is a rookie and has just played six to seven matches so far, we now have a limited amount of data to work with. Even if we compare the current season no definitive predictions can be made in such a case based on his data alone. Thus, here we would compare his six to seven games to another rookie of similar position let say our main player was a quarterback. We will compare his stats with a rookie quarterback that had six games in the season and then we would see how his performance was next year. Here you need to keep in mind the rules of the league as well. 

Relying on Data science is just about having an edge over opponents who are making a decision solely based on emotions and intuitions. It would even help you make a decision independent of your biases and separate emotions from the game as well. 

While such a method should give you results but there are other methods as well that can help in selecting a better roster- Linear Regression, Random Forest and Neural Network 

There are so many factors that can affect players' performance like health, weather if you include that in your football data analysis as well, it would give you a more robust solution. The only way to get access to such comprehensive fantasy football data would be to either do it manually or to get in touch with sports data providers, who are essentially football data companies, who can give you access to historical and live data. Data Sports Group has established itself in this category by providing raw sports data services. With fast delivery of real-time data and decades of historical data combined with Artificial Intelligence (AI) and automated machine learning models improve your Football Fantasy League.

STAY IN TOUCH

Signup to our newsletter to receive updates