Hi Nicholas,
Answer to question 1) The EA’s are the same. They are coded differently as MT4 and MT5 have different languages. So yes you can use MT5 for testing and then MT4 for live trading. However, be aware of one thing: If MT5 is on a different broker or different prop firm than MT4 then there could be some difference as each broker or prop firm will have somewhat different histortical data.
Answer to question 2) One week or two weeks is not sufficient backtest to see how the EA will perform over the long term as during such short periods of time the market could be trending without any significant retrace or drawdown. During longer periods of time such as one month or longer then the market is more likely to have some small retraces and even a drawdown. If you are going to run your EA for a month or longer than you want to know how it will perform during retraces or drawdown. So it is recommended if you are going to run an EA for a month or longer that you at least backtest it for one month.
Answer to question 3) There is no right or wrong answer to this question. I have used the best 1 to 3 EAs and have found that the 10 EAs they were selected from actually performed better than the few I selected. Then again I have had the opposite occur. The one nice thing about running all the EA’s is that they will compensate for each other. What I mean by that is when one or two EAs are going through a drawdown the others most likely could be profiting, this helps to minimize drawdowns. So, since it is difficult to decide to run 1 to 3 EAs or the whole lot, what I do when I want to have more than a 1% account risk I will then run all the EA’s to minimize drawdown. If I just run 1 to 3 EAs then I keep my account risk to 1% because the drawdowns will most likely be greater. So, running the whole lot with a 2% risk also helps to increase profit. So running the whole lot helps to keep drawdown to a minimum while at the same time increases the chance of greater profit.
Hope this helps!
Alan,