Why I’m blogging about NBA DFS
Three years ago, you couldn’t walk around South Station in Boston without seeing a DraftKings ad. They were everywhere. I like to think advertising doesn’t work on me, but I gave it a try.
Three years later, I’m still playing DFS. I’m building an open-source strategy tool for DFS. DFS has been an outlet for me to enjoy two separate passions of mine — data and sports. I’ve been fortunate to have others lead the way here. Matt Swanson’s amazing DFS optimizer library formed the basis for my own.
Like everyone who has ever entered a DraftKings contest, I believe there is some optimal strategy to win on a consistent basis. Many websites encourage this belief. Behind $100 per month subscription paywalls, there are lineup recommendations that can unlock the keys to untold sums of wealth! Perhaps this is true; I’ve never paid to find out.
I’m still searching for that optimal strategy. And I believe I’ve learned a lot in that search, and continue to learn more. Importantly, what I’m learning isn’t confined to sports. DFS has taught me a tremendous amount about prediction markets, human psychology and exploring data.
I want to share those lessons and continue learning. As a developer, I’m reminded every day that information wants to be free. All of the tools I have been able to build for myself, friends, family and companies are products of others sharing knowledge. It’s possible that lessons that will be canonical in DFS are not known because they haven’t been publicized or even discovered.
That last point I think is important — because this is only the first decade of DFS being a public game, we may be living in the 1800s of stock market investing. This means the iconic books of investing — The Intelligent Investor, Margin of Safety and others — have yet to be published. Benjamin Graham and Seth Klarman forever changed the way we value assets. My goal — and sure, I probably won’t accomplish it — is to forever change how DFS players pick their lineups. And maybe make a little money along the way :-).
Even if this optimal strategy does not exist, I see DFS as an opportunity to learn about data science. Through this blog, I’ll explore data science techniques and tools. All techniques will be measured against real world success — if the predictions generated by a certain type of analysis are bad, this blog will forward an explanation of why. If they’re good, the why is equally important, and this blog will arrive at an explanation.
In short, DFS is an incredible testing platform for all types of interesting data experiments. How to arrive at the right answers — or wrong answers — will yield knowledge that extends beyond sports.
