Welcome to PHIT Football – Fantasy Premier League is Back
Last week I made a video called “making a 20 slide deck PowerPoint presentation in under 10 minutes.” Well, this is the deck being presented. I want to introduce PHIT Football my analysis of the Fantasy Premier League data to help a fictitious character called Berry Wand to do better at Fantasy Football. You watch the YouTube video here.
Like all good stories, we have a protagonist in this case it’s all about Berry Wand. Who wants to use data to be successful. Berry hasn’t done badly in the season’s gone by, I guess you could call it mediocre but she hasn’t really been in contention to win anything.
Berry’s Football Problem
Berry hasn’t got close to winning in the previous season perhaps because of unintended biases in here decision making. She doesn’t like Everton so she doesn’t pick any of The Everton players, she doesn’t like Manchester United either so doesn’t pick their players for the same reason. Berry might bring in players into her team that perhaps she thinks should score some excellent points, based on a hunch, but when it comes to it they don’t deliver any points at all for her team.
How can she pick players and make transfers and bring players into her team that will have a better chance of being successful?
Berry’s Data Solution
The FPL make their data available to you through their website. They also make it available through their publicly available APIs. Berry working in data wants to see if she can use data to make better decisions and make her way up the league.
PowerShell Modules
There is a great PowerShell module that will plug into the FPL API so you can retrieve all the data you need on all the players and use that data to help make decisions. Find the PowerShell modules here.
Power BI
You can feed the data into Power BI to carry out that further analysis and tell a story around the players, their data, and help with decisions about player selection.
There are numerous ways to bring the data into Power BI – I’ll likely just pop the output of the API call into a CSV and bring that data in initially. There will be other ways to bring this data into Power BI too.
When we have our model, we can create some visuals to highlight quickly transfers that should be made – shipping out players not performing well enough and bring in others who will
You can look for patterns in the data so – we can use some of the “AI” features in Power BI to see if it notices things that perhaps we miss
We will pick our team based on the data. We will try and remove biases as much as possible
What you need to win
If you need information on the scoring system, head over https://fantasy.premierleague.com/help/rules where it tells you how players can score points
So you can understand what scores you need to get to, the overall winners totals for the last seven seasons are below
- 2020/21 Michael Coone Teddy Bears United (2680)
- 2019/2020 – Anthony Bull 2577
- 2018/19 Adam Levy, #TheyAreUs (2,659 points)
- 2017/18 Yusuf Sheikh, Yusuf’s Team (2,512 points)
- 2016/17 Ben Crabtree, FC Crab Dogg (2,564 points)
- 2015/16 Dimitri Nicolaou, Dimitris gavles (2,458 points)
- 2014/15 Simon March, Atletico Marchid** (2,470 points)
Our Targets
This might make a good KPI visual!
We will look at this further when we pick out starting squad but we will aim for 2600 points for the season that would have won the league in 6/7 seasons
- This equates to a weekly points total of ~68.5 points per week.
- Or each million pound of budget bringing in 26 points across the season
- Or each million brining in 0.68 points a game
The figures above are to win the whole thing, for most of the private leagues I’ve played in over the years you might have won those leagues with 200 fewer points – so you can still be successful even if you don’t get to 2600 points in fact if I could have squeezed an extra 100 points in seasons gone by Berry would have challenging at the top
Get Involved
This is meant to be a bit of fun and hopefully, improve Berry’s performance in the fantasy league and learn some Power BI along the way. If you want to get involved and keep track and learn how to use Power BI along the way then join our league. Head over to https://fantasy.premierleague.com Sign up for an account if you don’t have one already, pick you team and join our league (League Code is apdxnm )
We’ll make a gethynellis.com t-shirt first prize – how can you refuse? Get involved and see if you do better than Berry. You never know, you might just win the league and the coveted prize.
If you want to join our league then the code is apdxnm subscribe to this channel, join the league and get involved.
Related Content
Four SQL Server problems you might be suffering from – PetchaKutcha Style
0 Comments
Trackbacks/Pingbacks