I did not try manual trading yet, I did spend long time watching the evolution of odds depending on the evolution of scroes (game point, break point, set point, tie breaks,...).
My guess is that elaborating a profitable strategy usually goes through the following steps:
- watch the behaviour of the markets, trying to find profitable strategies
- test manually each strategy live with small stakes
- once confirmed, increase the stakes
- if possible, automate the strategy
You can also back test manually strategies with a time machine using BF historical data. I did not test it yet.
The issue is that testing live strategies with small stakes , or better backtesting manually with time machine, takes a long time, and you often test a lot of unsucessfull strategies before finding a profitable one.
With a back testing tool (which needs to be developped), it would be:
- watch the behaviour of the markets, trying to find profitable strategies
- implement the algorythm for each strategy
- back test each strategy algorythm
- run the profitable algorythm live, with small stakes first then bigger stakes
A back testing tool would definitively help a lot. I am an IT engineer, I could write one but it will take me very long. I am not sure I want to take that path, when the final incomes are not certain, of course
Another issue about automation is to get the tennis (or other sport) score live, with a reasonable delay...