🗑️ 🧵 Untitled Thread
Anonymous at Wed, 31 Jul 2024 22:03:00 UTC No. 16305052
I don't watch football but I'm interested in statistics and markets. Recently I've become aware of the Yahoo Fantasy game and its robust API. The way this game works is that you join a "fantasy league" which is like a lobby with 2-16 other human players. You take turns drafting football characters for your team, who are also football players in the real-life NFL. Every week you're pitted against another team, and when the football characters you have on your fantasy team do things in the IRL NFL such as running and scoring points, then your team gets points in the online game. It goes on like this for an entire season, until one team is the virtual super bowl winner.
Crucially, they have public fantasy leagues which anyone can join - thousands of them - including ones which cost money to join, and the winner gets money in the end of the season. So I'm thinking about how easy it would be to write a bot that enters paid fantasy leagues with a mathematically optimal draft based on each player's predicted performance, which is a variable you can get from the API, and then just sit back while it rakes in the cash. I think that the tiny size of each league (being like less than 20 people) would render each market extremely inefficient, so on average if I just go by the AI predictions and enter enough leagues I should end up overwhelmingly in the black. What do you think about this?
Anonymous at Thu, 1 Aug 2024 01:59:22 UTC No. 16305298
>>16305052
Too many variables in the NFL. You can’t just sit back. You have to monitor injuries and players who are having over/ under performing seasons for what ever reason. I could be they just don’t fit into the teams game plan that well or have a tough matchup one week. Yeah in general it would even out and if playing against random people you may come out a bit ahead but gambling of all sorts has had all its numbers crunched for a time so this is nothing new. People arbitrage baseball games and shit like that with a similar concept.
Anonymous at Thu, 1 Aug 2024 17:01:52 UTC No. 16305999
if you want to do this for the money there are some eggheads already doing that.
kevin cole, eric eager, george chahrouri from pff are all guys with some academic maths background who are deep into nfl betting and podcasting about it.
stick to best ball drafts so you don't have to manage anything and take advantage of early bets.