Alright, so I wanted to make some extra cash, and I heard about FanDuel and NFL. I’m not a huge sports guy, but hey, money talks, right? So, I figured I’d try building a “lineup optimizer” thingy. I thought it would be easier but I have never done this before.
Getting Started (and Getting Confused)
First, I needed data. Touchdown! I easily found projections for players’ points. It’s all about how many points these guys are expected to score. More points, more money… hopefully.
I installed python on my windows, I chose python because I saw that a lot of people use this language on internet.
Then I copied some python code from somewhere that I really don’t remember now, and that was using something called “PuLP”. I guess that’s a tool that helps you pick the “best” combination of players, staying under the FanDuel salary cap. Each player costs money(virtually), and you only have a limited budget.
After several trying, I made my python code run, and I started to understand what was going on!
The “Optimizer” Does Its Thing
I put in the data, set the salary cap ($60,000, that’s what FanDuel uses, I think), and told it how many players of each position I needed (QB, RB, WR, etc. – I had to Google what those even meant, it is Quarter Back, Running Back, Wide Receiver…).
After a little bit I finally managed to run the script. It started chugging away, doing… math stuff, I guess. Eventually, it spit out a lineup:
- QB: Some Dude (I don’t even know who this is)
- RB: Another Guy, Another Guy
- WR: Three More Guys
- TE: You Get the Idea
- FLEX: Someone Else
- DST: A Whole Team?
It also told me the total projected points for this lineup, and how much “salary” I had left. Cool!
Did It Work? (The Big Question)
Okay, so here’s the thing. I built this thing, but I have NO idea if it’s actually any good. I mean, it did give me a lineup that followed all the rules. But will it win me any money? I don’t know!
I entered the lineup into a FanDuel contest, this is very easy, just select the players with the names that your python script returned.
I won nothing, but after a few changes, the python script finally returned a good rooster, I won something like $10, I was very excited!
This was my first try, and I’m sure I’ll tweak things, maybe try different data, or even learn what all those football terms actually mean. It was a fun little project, even if I’m still mostly clueless about football.
