Machine Learning - Fantasy Premier League (Sneak Peak)

[bradsmith]

Good afternoon everybody!

I wanted to follow-up on the machine learning posts I had made a couple of weeks ago and share a sneak peak of a fun project I am working on to showcase real world machine learning in action around a subject many of you are very familiar with… Fantasy Sports. Specifically, Fantasy Football for the Premier League. I’ll go more in depth when I finish things up, but here’s a little summary:

There’s three different methods for using machine learning within Power BI built into the report: custom data tables using live data tables and what if parameters, clustering and pricing, and player statistics. The data tables on the initial page are custom visuals that uses what-if parameters to pull player statistics directly from the fantasy premier league website and performs a simple, yet amazing optimization model. The clustering and pricing models use k-means clustering and a linear regression model. And the player statistics contains web scraping, data cleaning, and general exploratory data analysis.

[MudassirAli]

@bradsmith Many mic drops.
This looks like I am actually watching analysis on premier league…
Wow

[sam.mckay]

Wow. So intrigued about this… Let’s chat some more on a call.

[bradsmith]

Absolutely! You’ve got my Skype contact info, shoot me a message and we’ll set it up.

[bradsmith]

Thank you! I only started playing Fantasy Premier League this season and that was only to test my model haha I wasn’t entirely what to even include in this report much less where to begin with analyzing the data… so that means a lot!

[MudassirAli]

Honestly speaking, this is completely different type of report than the ones in challenges by @bradsmith.
3 to 4 visuals per page is the way to go. You can make use one color for column or bar charts or the family of colors. The rest is just mind-blowing.

[ambepat]

@bradsmith I am intrigued. I am an staunch Man united supporter here in the UK. Any info I can use to make my fantasy team better will be much appreciated. Really keen t o see what you have in your tool box.

[Rafal_J]

@bradsmith Wow, this is an awesome way to get some practical insights into ML Definitely will bookmark this for future reference when I will know more

[EnterpriseDNA]

Hello @bradsmith, due to the length by which this post has been active, we are tagging it as Solved. For further questions related to this post, please make a new thread. When starting a new post, I suggest linking this thread, provide as much context to a question as possible and include the masked demo pbix file, images of the entire scenario you are dealing with, screenshot of the data model, details of how you want to visualize a result, and any other supporting links and details. Thanks!