Interesting analysis, but sample size is too small, you are live trading based on 112 live paper trades? Where are the historical stats? Ideally you need samples in the 10,000's to make it meaningful and you need to extract samples that cover different market conditions and once you have that you need to test the strategy against a sample you haven't previously tested against. There's a lot of work in all of that as you have many pairs to test against, a machine would be ideal to help you.
It's a fun experiment nevertheless, £1,000 isn't going to be life changing to lose 😆