<![CDATA[Forex Software — Strategy ambiguous bars?]]> https://forexsb.com/forum/topic/6270/strategy-ambiguous-bars/ Fri, 19 Aug 2016 12:33:28 +0000 PunBB <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38255/#p38255 Thank you for detailed answer! Now things a lot clearer smile

]]> Fri, 19 Aug 2016 12:33:28 +0000 https://forexsb.com/forum/post/38255/#p38255 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38251/#p38251 "Enter Once" has no trading logic. It is responsible to set a max count of trades per period (ex. bar, as in this case).
FSB can open at maximum two positions per bar - one long and one short. When you use Enter Once, the program opens only the first signal.

..

The problem with the backtest of the above strategy comes from the particular combination of Price Move entry, the SL, and the TP. This causes the "Pessimistic" algorithm to take wrong decision for the entry point and the protection. The algorithm announces an ambiguous bar in such case. 

When the bar starts, the program sees that there are two entry points with corresponding SL and TP for each of them. These are 6 potential trading points. The FSB Pro algorithm tries to compose a "pessimistic" route between the points without looking in the future. This works in 99.9% of the cases, but some times it doesn't. Fortunately, FSB Pro comes with a Comparator and a Monte Carlo tools and reveals such problems easily.

If you think, such complex algorithms are drawback, imagine how MT tester works. It simply draws and follows a jagged zigzag line between the OHLC points and calls that "Tick mode". No ambiguous bars - no problems for the traders.

]]> Fri, 19 Aug 2016 09:15:08 +0000 https://forexsb.com/forum/post/38251/#p38251 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38250/#p38250 HI,
I think the problem is with "enter once", it should have bug somewhere. Sometime ago I played with it and found similar unrealistic results, however not reported it here and disregarded this indicator. However not so sure if the problem is with it, but it my guess smile

]]> Fri, 19 Aug 2016 08:31:01 +0000 https://forexsb.com/forum/post/38250/#p38250 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38239/#p38239 Thanks for the quick reply!  Knew there was something wrong but had never seen the ambiguous bars impact a strategy so dramatically before.

]]> Thu, 18 Aug 2016 21:19:51 +0000 https://forexsb.com/forum/post/38239/#p38239 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38238/#p38238 The problem is in the ambiguous bars. Try the Comparator.

http://s4.postimg.org/tkbyu660p/fsbvseas.jpg

]]> Thu, 18 Aug 2016 21:07:53 +0000 https://forexsb.com/forum/post/38238/#p38238 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38237/#p38237 It is clear that the backtest result is highly overestimated. I'll try to find what is the problem.

http://s4.postimg.org/g4oy4pxix/fsbvseas.jpg

]]> Thu, 18 Aug 2016 21:05:45 +0000 https://forexsb.com/forum/post/38237/#p38237 <![CDATA[Re: Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38236/#p38236 And the mql..

]]> Thu, 18 Aug 2016 21:00:59 +0000 https://forexsb.com/forum/post/38236/#p38236 <![CDATA[Strategy ambiguous bars?]]> https://forexsb.com/forum/post/38235/#p38235 In fsb, the attached strategy looks awesome (both against my broker's data and to a lesser extent the fsb demo data.)  It did great in multi pair testing, monte carlo etc.

https://gyazo.com/ddd52251e404f3484f31dfc0613d45c0

When I tossed on demo, it's performance wasn't nearly as pretty.  In the strategy tester in MT4 it dug a hole to China.  So... is the issue that it's got a bunch of ambiguous bars or is there something else I should be considering?

]]> Thu, 18 Aug 2016 20:59:37 +0000 https://forexsb.com/forum/post/38235/#p38235