During backtesting I have found that accidents have provided me some of my best results. I will be tweaking variables trying to represent some idea I have and Ill get a result back that is significantly better. But after reviewing what I adjusted I realize it doesn't match with my logic and was wrong. Ive found some of my best improvements have happened like this or I just made a typo. Sometimes I figure out why its working but other times I still have no idea why its so successful. Anybody else have this happen? It almost seems like the most illogical systems work the best..maybe because no one else is doing it?
Or maybe because you're illogical. But success is about serendipity. Trials and errors. Tinkering. IMO.
If it's an obvious price level, then I believe the market will quickly discount the opportunity. Not because it's oh so "efficient", but because participants will become more active there. So it makes sense that illogical systems may work better. However, it may be the market that is illogical, and not your trading systems. The market indirectly infecting your adapting systems with its inherent biases. It's really not very logical that everything should have a set price tag at any point in time, or that that price tag is objectively correct at all times. To believe so is very illogical and based on circular logic. We know price is strongly ruled by supply and demand, which by themselves are not perfect at all times. We know most knowledge 20, 50, 100 and 1000 years before were deeply flawed in many ways, and should have no qualms to question current dogmas. If we are truly evolving beings, our current state is never optimal, no matter how convenient and comforting it may feel at the moment to live accordingly.
I still don't understand backtest, when we so freely can forward test. Possibly no patience??? To lose money!!!
Isnt backtesting part of forward testing??? Slight optimization of your backtest sample is then used in a different sample AKA forward testing? Forward testing without backtesting would be highly inefficient. Also, when you decide that you have done enough forward testing and look BACK at your results, isnt it now considered backtesting? Mind. Blown. And regardless, my original question still applies to forward testing...quit trying to start a pissing match and derailing my thread. You can freely start a thread of which Im sure there are 1000 that says "Is backtesting worthless?"
I had to think about that second paragraph, Im not sure if the set price point is valid (Saying this for purely discussion purposes, not a right or wrong thing) Systems or at least mine are more of a if X happens then do Y. I dont think we are trying to find a correct price but more make rules for patterns....maybe...Im still chewing on this in my head.