Thanks, yes it’s a real eye opener and I now have a headache!
Managed to tweak the EA so over 4 years of testing it lost 63 pence!
So no good really but it does show just by tweaking things you can get a 5% improvement, useful to know if I can get one that actually makes a worthwhile profit in the first place.
I’ve got load of ideas, probably none original, but let’s see what can be done. AI is great but also can be a real pain as well, like having a 2 year old!
Well I am hoping this is a bullish pennant forming right now on XAUUSD, fingers crossed I am in the trade.
Eek! More than 15p per annum?! Last of the great financial disasters. 
Yes; a more serious point altogether and essential to know. Especially if you can do that 2-3 times, you can move many things from overall loss into overall profit.
You probably don’t score anything extra for “originality”. Maybe nothing’s really completely “original” anyway? I wouldn’t worry too much about that.
Happy with this, for once I actually did a trade using some analysis and didn’t just jump in feet first!
tradingview. dot com /x/SxSBIGsh/
(can’t post links still!)
Now just over half way to my first phase of the challenge with Maven.
I hear that! 
So you just have to repeat once whatever you’ve successfully done once, and then you’re funded? Sounds good.
No it’s only the first phase of three, but it’s still a nice milestone to be at!
I chose small phases of just 3% as it seems easier overall.
Sorry, misunderstood. But wishing you very good luck for the rest of this and two more little phases - take care! 
I’ve got a XAUUSD scalping EA that is profitable on Monday and Fridays now
That’s backtesting over 18 months.
Need to see if I can tweak it, it also looks to be profitable Mon-Fri for around 11am to 3pm, weird.
Yes, that would certainly concern me as “potential over fitting.” I wouldn’t be too confident about that one, moving forward.
Now got one with 80% profit over a year!
It’s rather unorthodox though, need to refine it.
What’s generally regarded as the most “stable” currency pair?
EUR/USD is generally regarded that way. AUD/USD is also relatively stable, and USD/CAD.
That’s “overall,” anyway - their relative stabilities do change, from time to time.
Well gold is certainly volatile right now.
Well, I have a profitable EA now, but how much profit do you need to make it worthwhile? Mine is only around 20% a year, you can get funds that perform better and are lower risk.
For me (for most people, I think?) it’s not about the profit %. It’s about how much bigger the profit % is than the drawdown %. In other words (and these would be the words of an investor or a risk manager or whoever), it’s about the drawdown %.
You can always adjust the “stakes”, if the drawdowns aren’t too dangerous.
A system that makes 20% per annum with a PTT drawdown of 4% is no better or worse than a system that makes 10% per annum with a PTT drawdown of 2%.
If you show any “profitable” system to any investor or risk manager or whomever, they’ll instinctively scale the PTT drawdown figure up or down to 4% (because that’s an industry-wide “maximum acceptable risk figure”) and then adjust the profit per annum accordingly, depending on how much they had to change the drawdown to make it 4%.
Of course, I’m not suggesting that your acceptable maximum drawdown figure has to be 4%. We all have our own degree of risk aversion, depending on all sorts of factors. (Mine, I admit, isn’t far off 4%!).
So: what was the biggest drawdown over the period for which the profit was 20% per annum?
Let me do more testing and I will let you know, I still think I can refine this and make it better.
Good luck. Make sure it is “testing” and not “backfitting,” though: it’s very easy to confuse the two and come up with something that was perfect historically but loses money later. Another reason why forward-testing and out-of-sample testing are absolutely essential. 
Right I’ve had to step away from building a robot as my head is bashed. Will possible come back to that once I have a easy to define strategy that works in real life!
But it’s been a very useful exercise for a number of reasons, back to trading now.
Are we talking literally or metaphorically, here? 
You’re not alone. It turns out there’s been some kind of software glitch, over this. You might want to add a quick post to the thread below, just to alert the authorities?
Is XAUUSD hitting 3k today?