Question on optimization

[Sanju Chand]

Hello...Is it normal that the number of trades placed by the optimization engine is far less that the number of trades placed by the tester using the optimal parameters returned during the optimizaton process..if so, what is the sanctity of the optimization engine...

[Brian Dee - Random Trader]

Arooba

IMHO

Its not unusual for there to be large differences.

Some EA's dont back test at all, some do but with optimistic results - sometimes very optimistic.

Forward testing (on single broker data) will frequently give significantly lower return than back-testing on MT data.

So, when possible, back-test on single-broker history data.

Be advised, there are some EAs that perform better on live data feeds than they do on any historical data.
These tend to be the statistically based EA's rather than indicator-based.

To sum up, yes, there are limitations in the optimization engine but it is still a very very useful tool
and in time you will learn the 'conversion of expectation' ratio for a particular EA (or family of EAs) that you create

Sounds tough?

Yes it is, its a steep learning curve followed by a vertical one :)

Worth it...? I think so!

Good luck

-BB-