1) Demo account (simulator software)...use only the one at your broker. 2) First design a trading plan with trade strategy "before" using a demo account. Thus, your trade strategy should have already been backtested prior to demo trading. Too many traders make the mistake of using a demo account while they are still trying to design a trade method...resulting in them being too long on the demo and that in itself will create problems when they move from demo trading to live real money trading. You'll hear about it...traders that trade well in demo but lose money when they switch to live real money trading. 3) Primary purpose of a demo account is to familiarize yourself with the trade execution software prior to live trading with real money on the same trade execution software. Simply, if you have a trading plan with trade method and its been backtested...you're now ready to demo trade to familiarize yourself with the trade execution software prior to live real money trading.
The biggest reason I went into quant-hft-algo land is because I hate the sweaty palms and blood pressure of manual trading. Software makes trades with ruthless cold blooded precision that I never could myself. It has no hope, no fear, no greed. Just math.
I do agree, all my failures in the beginning was not having a well tested Trading Plan, I was just too lazy wanting to learn how to program. I have had fears my entire life, LOL and it is funny as it shows my weaknesses in life, but you get to a certain age where it gets easier and having more of a "I don't give a crap" attitude, perhaps I can help other traders see my weaknesses so they don't feel alone. In the beginning I had fears of losing money, then once I got to consistently making money, I had fears it was dumb luck I was doing well and wondered where the end was coming. There at some point like a decade into it, my fears were the more I made, my thoughts I had less. Now after so much time has past, my fears are more of I don't remember where I put my glasses or I just had my car keys, LOL. I don't have fears any more about trading, it could be the 100k trades I have taken, I know I have worked through every kind of problem at least three times in my life, nothing surprises me as it is in my Trading Plan which is always ongoing. As far a Demo account, I think it is needed, I still use it today on new systems I have made, I make a new system almost each month, some people good a crossword puzzles, my hobby is trading systems. Once I have demo'ed it, it gets real time traded for six months then coded for automation. Most systems I build will make profits based on good money management rules, but often system doesn't have low enough losing percentages as I like averaging down.
I too am continually futzing with it. Even though markets never change fundamentally (i.e., imbalances between demand and supply), they do vary, e.g., the winter and summer quarters. If one isn't making the necessary adjustments, he'll either be taking a lot of unnecessary and useless trades or he'll be missing a lot of trades that could have rung the bell. Not that this takes very long. But it does have to be done.
I was an options mm. Toward the end, I made markets electronically and automated. With that type of trading, you can't make money without risk. You can't get rid of all your Greeks at eod without a cost.
A good system is "autoadaptive". It should first see what kind of market you are trading and adapt itself to take optimal profits. So no matter what market you are in you should always make profit on a short term basis (day or week, not trade by trade). There is no universal system to trade all markets at all times in all kind of trends (or no trends). So a system should exist of a number of subsystems. It's all about number crunching. All markets have prices, these prices are the basis to work on. Whether you trade futures, forex or stocks should not make any difference, because a computer calculates always the same way with numbers. 1+1=2 no matter if it is a quote from futures, forex or stocks. It will always stay 2.
When I say there is no universal system that can do this, I mean a system without subsystems. A system that has several subsystems can do the job. A systems should select a subsystem to do that specific trade. That's the only job the systems should do: monitor which subsystem should do the trade at that moment. So you need a lot of different trading systems, each working under specific conditions. But trades can switch from one subsystem to another before the trade is closed, because conditions can change. They can even changes several times within 1 trade.
Actually, I don't need any of that and never have. But that doesn't mean that there aren't people who do. But getting back to the OP, this may or may not be about whether or not a demo account is a waste of time if one is trading computerized systems. I have no idea. I have no interest in computerized systems. Perhaps the OP does. Perhaps she couldn't care less. But that leaves the question of why not? What exactly is the downside of testing one's system before putting down real money? If there isn't one, at least a rational one, why does this subject keep coming up year after year after year?
The million dollar question is how your system can sense the market, namely is it trending, sideway, level of volatility and etc. It is easy to say this in hindsight (as most ET here did), not easy to predict it in real time. Hint: The usual BS education material such as 50MA cross 100MA and etc will not work, at least not have any statistical significant data to proof it works