How do you know when you are onto a winning system?

I think that everyone is looking for the holy grail (well except for those that believe it don't exist, and they are probably the proiftable traders!). But how do you know when you have hit upon a system that actually works?

I think you can draw a comparison with Elliot waves, regardless of what happens, you can zoom out on a chart and see that it is following that particular pattern. If I can explain:

You could have a system that is a total fluke, and for 10 trades it has 10 wins. It profits (for extremism cases) £10 million. However, subsequently you make 100 trades with it, and it loses £100k each trade - leaving you break even at the end.

Likewise you could have a sysemt that has only one win and a 1000 loses, and yet you end up being +1 pip on the system.

Zoom in on these, and the 1st example would be the better system, zoom and out and the 2nd would be considered better (assuming that you stayed with the systems - after all this is all hypothetical)

This can be replicated for any system, over any amount of timeframe over any amount of trades. You may think that you have a winner, but if you could see in the future, you may see it losing millions...

Is a system only as good as it's last trade, last 10, last 100 etc?

What testing would prove to you that it is good ong term/short term? What is more important?

I think that the people who have not stumbled across something that works for them, expect to find diamonds, probably do find them, but toss them aside because they have a bit of dirt on them... they are not shiny like we expect them to be.

A bit of a rambling post, but I am actually interested to hear some answers or thougts on the original question!

I'd consider a system profitable if it runs for a long enough time with a stable profit chart curve. I'd make sure it has gone through the bad phases, where all possible bad scenarios for it were played out. For the short-term trading system, several months is usually enough. For the long-term (daily, weekly charts), it would probably take several years to confirm the system profitability, but the problem is that the global market conditions could change at that moment making the strategy useless .

There are a lot of excellent trading systems, which you could find here at FF. Thats's not the question! The main issue could be based to the MM-Concept which you have to use in conjunction with the trading system and your risk-tolerance. That's all!

I wouldn't say "a lot of excellent"...there are a number of profitable ones with low hit rate/profit/higher risk, but the ones with high hit rate/profit/low risk, I think can be counted with one hand.

Back to the question, effectiveness for a full year should be enough to determine a successful system. I'm talking day trading by the way, I believe swing trading would be a different matter.

check your bank account

when you do not have to rely on the opinions of others or yourself. The system itself is profitable without outside interpretation.

When you test to see if a strategy works, you'll have a database of data. This'll tell you if it's running normally and what to expect in terms of averages.

Win ratios, drawdown, average win/loss in pips, average gain per day/week, consecutive losses, ect.

Whole point of spreadsheeting your trades is to track performance and compare it to past data.

Here is the thing about testing. A strategy will only work according to the instructions of the strategy. So when testing, initially you are testing to ensure that the entries/exits are as expected. Then after you verify correct functioning, you then check for profitability. Most people do it the other way around, either because they have blackbox system and have no idea how the components work, or they are expecting some magical gain from throwing a bunch of parts/systems together and hoping it will just 'work'

You stop searching for another system.
You like trading that way
Your account balance reinforces the same.