- it must be built for automation.
- it must be built for scale.
As an experienced developer i've build a backend and reporting system that allows me to track everything from bankroll management, through realtime analysis.
I started my journey with prediction football scores, and then I realized that on an exchange, it is better to focus on human behavior and look for the movements of odds.
I applied a lay to back methodology on greyhound racing trying to catch odds raising movement of x% profit, and hope for the best if I didn't find the right match.
It surprisingly worked pretty well, and i'm running this for 8 days now, and I used some deep analysis to build strategies (for example, determine the minimum odd I will get in to a bet).
So few thoughts:
- How can I know that I found an edge? How many bets with a nice edge graph will prove the point and will make me feel more comfortable to increase stake value?
- the importance of bankroll management: the real cost of fees when you are running a long term strategy that both wins and loses (but better on the win size) - is just not spoken enough! You can have a "profitable" strategy with negative net. profit! That's crazy and counterintuitive, but crucial for success management!