Time hides Nothing
- Joined Oct 2005 | Status: Pip Slappa Extrordinaire | 1,012 Posts
You are in a maze of twisty little passages, all alike.
Old Dog's Taming of the Beast? 34 replies
Old Dog Fib Trading - EA Development 18 replies
Old Dog Fib Trading - EURUSD 399 replies
5 Tricks of Trend Trading, by Rob Booker 25 replies
DislikedI think the old refers to the basic two strategies: breakout and countertrend. The new trick is that the ea decides between the two based on back testing. Or I could be totally wrongIgnored
DislikedSo which file should we be using? Steve's updated one or Ron's updated one? Or either b/c they both do the same?Ignored
DislikedRonald, after reading your posts on this thread I realized that your settings must be much more tight or with more restrictions than the default settings and this is because you have said that your EA have taken only 8 or 9 trades a week, well I m getting hundreds of trades per week.Ignored
DislikedA 50 winrate is a little low. You might as well flip a coin.
It is impressive that the bot is still profitable though.
I wonder if you could make it be profitable with a 25 winrateIgnored
DislikedA 50 winrate is a little low. You might as well flip a coin.
It is impressive that the bot is still profitable though.Ignored
DislikedFor everyone who wanted to join, the webinar is now starting, ask away.
Room Name:
ronaldraygun
Organizer:
ronaldraygun
Start Time:
August 5, 2010 8:51:45 PM EDT
Dial-in :
702-473-3463
Passcode :
766253
Meeting URL:
https://my.dimdim.com/ronaldraygun/Ignored
DislikedBy limiting the trades to a higher than 50% winrate we're just introducing an additional filter which would make the trades more likely to be successful, however we could just limit trades to a higher success rate with similar effect, could we not?Ignored
DislikedI notice there was discussion earlier about weighting more recent results more heavily. Perhaps one way of doing that is to apply a linear gradient, so that the most recent are weighted X times as much as the oldest results.Ignored
DislikedThis backtest system could also be utilised for a number of other trading strategies. It would be interesting to see it applied to some other easy-to-code currently zero-edge approaches such as MA crosses, rsi trades and so on, to see if an edge can be found. More complex approaches might prove more fruitful, however.Ignored
DislikedI haven't had a chance to look through this code properly yet, but I'm wondering how difficult it would be to plug in different logic for the backtest system. Any tips on this, Ronald? Or is that not within the scope of what you intended to happen in this thread?Ignored
In terms of using the structure of this EA to run the backtests for other systems, that's mighty difficult. Every time I want to write an EA that optimizes itself, I have to write from scratch.
DislikedAs I explained on the webinar, this method may be tricky due to how the optimizer works.Ignored