Martin Eastwood has a brilliant blog and he kindly published his R code here on how to optimise your Fantasy Football team. Optimisation applications are fascinating and after languishing pretty badly in my own mini league I decided to apply this technique to see if I could get some insights.
I had to rejig some of the original code due to the fact that the website had changed structure but since Gameweek 7 I have been monitoring the team that the optimisation provides. It has been interesting. The key stats that I take from this exercise are:
Capoue is the favoured man from the optimisation, closely followed by Defoe but Capoue’s form has fallen off a cliff
42 separate players have been selected
Chelsea are the leading team in terms of selection
No one from Leicester made the team
The following visualisation tracks the number of times a player was selected with their latest stats as per Gameweek 20
Looking at the individual Gameweeks it is clear to see Capoue and Defoe are neck and neck in terms of selection but Capoue has just fallen out of the team in Gameweek 20 and unless there is a major change in form I dont see him coming back.
The Fantasy Football stats do provide a “Form” metric, as of the 18th of January Delli Alli is the main guy here so my next plan is to weight the total points (the maximisation element of the optimisation) by Form to try and capture this metric
Thanks to the following for inspiring this post:
Martin Eastwood as mentioned above for his code
Data Remixed for the template based on the NBA here
Letsgodata for helping explain how the dashboard worked