>>12912372I did back in 2015 for CSGO betting. Esports is a lot more profitable since all the data is usually logged and you can scrape game and player data from stat pages.
Example is pic related of what I'm talking about for 1 player. I was a teenager back then and I can spot holes in this looking back on it now. You can average the performance of a team, which is 5 players in a specific region (NA, EU, CIS, etc) and each region has different performing teams when compared to each other. The only problem is that 5 players is a hilariously small sample size and anything that effects their performance such as jet lag from traveling from Germany to California will effect the average performance and thus your bet. This doesn't even account for the shadyness of esports betting now from lower tier teams and even the top tier teams sometimes.
I bet you could do it for actual sports, the only problem lies in that the same logic applies to whatever real sport you're applying it to. Unless a team is performing consistently 100% of the time against a team that also performs consistently 100% of the time, there will always be anomalies because the amount of people in any given game is too small to consistently predict the outcome unless the matchup is hilariously unbalanced like the Lakers vs a college basketball team.
I haven't messed with sports betting in years now so if anyone wants to poke holes in what I'm saying then please do so.