1 Topic by hannahis (edited by hannahis 2015-10-06 18:20:22)

  • Registered: 2010-11-18
  • Posts: 1,179

Topic: Remove "Use Previous Bar Value" Option

Hi Popov,

I know you mentioned that to remove "use previous bar value" is attempting a "financial" suicide.

However, for 1min time chart use, the difference won't be too great an impact right to be considered as a suicide attempt right? 

By removing the "use previous bar value" isn't I maximizing the use of tick data? 

A tick chart will also allow you to “see” more trade information and work particularly well with cycle analysis.  If you’re waiting for the close of a bar to enter a trade, say a breakout, a tick chart will get you in earlier.

By restricting the FSB to calculate strategies and enter into position only after the "use previous bar value" isn't that defeat the very purpose and importance of using tick data instead of min data or worse, 4 hourly data, i.e, waiting for the 4 hour bar to close before executing a trade?

I would prefer instead of "forcing" FSB users to "use previous bar value", please consider keeping it as an option.  Furthermore I don't agree with the concept behind the use of this function.  Because we compromise trade accuracy for "statistical" stability.  It defy real trading practice, especially for any strategies that use beyond 30min time FSB time chart (unless you are a position trader who trade long term, tick data may not seem so important to such traders.  But for scalper trader, time is of the essence and every tick counts).

Maybe that's explain why sometimes FSB profit/statistical results differ so much from real/demo trading results because of this function (that's just my guess for the discrepancy.  Otherwise, what other probably reasons could there be?  And if there are so many other probably reasons then where is the yard stick whereby we can use to measure or develop our strategies if all these statistics are so arbitrary). 

If I can't develop EA based on the FSB statistical results because there are so many factors that could explain why such discrepancy between FSB results and demo/live trading, then what's the strength of FSB back testing for?  What's the point of using tick data when "use previous bar value" interfere with the calculation or execution of an EA.

As for me, I've been using FSB to input my trading theory and I try to tweak my EA to get the best results from FSB but sometimes, I just go ahead and use my EA even if FSB results aren't good because I trust my trading theory more than I trust FSB results.  My sympathies go to those who are trying to use FSB to figure what works and what doesn't.  FSB spectacular results but EA crash the demo account when put to test.  And if we can't rely on FSB results then it's such a daunting task and discouraging often times.  Now and then, we will have new users crying out loud "what's going on?, What's wrong?  I believe you are well aware of such discrepancies.  Most likely also figuring out a solution.

There must be somewhere that aren't right in the FSB calculation.  I really can't figure out what it is but to say the least, maybe "use previous bar value" may be the 1st area to look into.  It may be something else...whatever it is, let's work on reducing the discrepancies so that we can truly rely on FSB results.

Lastly, the onus is on the traders to exercise caution when testing out their EA on a real account.  If a trader is not wise, he is not wise...It doesn't matter whether they "use previous bar value" or not to commit a financial suicide.

For the sake of experimental growth, let FSB users have the choice to "use previous bar value" or not.  Moreover, FSB is a platform for all to do testing and breaking paradigms in our trading perspectives.  No harm having more options to venture into.  I may be wrong about "use previous bar value" as the factor for the discrepancy but still I would like to have it as an option rather than being "forced" to use it.

Despite all these...FSB is still far superior than any products out there.  And without FSB, my trading theory would be just a fraction of my imagination...but with FSB, it's a reality.  And I'm grateful to you for all your efforts and impressed by your genius, expertise and professional dedication.  Don't get me wrong, I'm just trying to figure out how to give my feedback in ways to make FSB even better.  And I never regret a single bit of buying the life time membership.  I'm a long time supporter of FSB and will always be smile

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Remove "Use Previous Bar Value" Option

For the sake of experimental growth, let FSB users have the choice to "use previous bar value" or not.

Long ago the "Use Previous Bar value" option was controlled by the users. Unfortunately, it was not found productive due to the lack of understanding of the problem. We fixed the problem by adding an intelligent system for controlling that option. It is so important that if a custom indicator fails  to comply with it, the backtest results become absolutely overestimated.   

I made an effort to demonstrate it now. I removed this feature from the Momentum indicator.

Here is the Momentum with removed Use Previous Bar Value:
http://s28.postimg.org/mikmd5gp5/screenshot_1223.jpg


Here is the backtest with the default Momentum options. No optimizations, no tweaks... just a simple Momentum with removed "Use previous bar value".

http://s28.postimg.org/eeci8eua1/screenshot_1224.jpg

The result will be the same if you set that option wrongly for the most of the indicators.

I don't think you want it.

You can find more info on that topic here: Use Previous Bar Value

We made all the best to make FSB as reliable as possible. We'll  fix any issue that deteriorate the safe trading immediately.

Trade Safe!

Popov's Website

3 Reply by hannahis (edited by hannahis 2015-10-06 19:41:39)

  • Registered: 2010-11-18
  • Posts: 1,179

Re: Remove "Use Previous Bar Value" Option

Thanks for your info.

To remove the "use previous bar value" do I have to do it one by one for each indicator or is there a programme code where I can disable it for all the indicators?

The topic for the "use previous bar value" is not really updated because the old FSB allows changes to be made for "use previous bar value" but the current FSB under the strategy tab only has the "recalculate strategies" option.  Is there a button to disable this function for all indicators?

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Remove "Use Previous Bar Value" Option

If we make an option for switching off the automatic "Use previous bar value" and if the user do not use it correctly, FSB will produce wrong results. Currently there is no way to make FSB showing wrong backatest. This is an ultimate traders protection and the main goal for the project.

If such option exists and a user uses it properly, it will work in the exactly same way as the automatic function. So making it is pointless and dangerous. There is no way to make it better than it works now. Even the fact you asking for it shows that you are not familiar with the process. I suggest you to use your time and energy to understand the FSB model instead of contradicting it. We are absolutely sure FSB works at the safest for the trades way we know.


..
Wow. This is my post number 4000 !!

Popov's Website

  • Pit
  • Licensed Member
  • Offline
  • Registered: 2015-01-11
  • Posts: 137

Re: Remove "Use Previous Bar Value" Option

Hi Popov

Lol This is my post number 4000 !!

Very good working.

Lg Pit

6 Reply by hannahis (edited by hannahis 2015-10-08 08:01:47)

  • Registered: 2010-11-18
  • Posts: 1,179

Re: Remove "Use Previous Bar Value" Option

Popov wrote:

Currently there is no way to make FSB showing wrong backatest. This is an ultimate traders protection and the main goal for the project.

Thanks Popov,

We are assuming that FSB back test currently is "accurate", i.e. what you see is what you get.  But there is a discrepancy in FSB back test results and the EA demo results.  We get fantastic FSB results but EA wipe out demo account.  How to address this discrepancy, if we assume the FSB back test is solid and stable.  What other explanation?  I'm just trying to figure that out and if we can pinpoint where the problem and address it...wow isn't wonderful when we have total confidence, when the day come when what we see (FSB statistical results) is as close to what we can get?  I'm a dreamer...and I like to break boundaries...to stretch and problem solve and yes, dreamers can be a pain in the neck for realistic people at often times smile

  • Registered: 2010-11-18
  • Posts: 1,179

Re: Remove "Use Previous Bar Value" Option

Popov wrote:

Wow. This is my post number 4000 !!

Wow, that's why I admire your professionalism and dedication in your attempt to ask our enquiries
so promptly and patiently.

Are you going to give away 1 Life Time Membership to whoever is your 1 millionth post? Ha, Ha, Ha

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Remove "Use Previous Bar Value" Option

We get fantastic FSB results but EA wipe out demo account.

The main reason for that is over-optimization. If I have "fantastic FSB results" and the strategy has more than 2 indicators, I sense a problem.
Please don't mix up wrong backest and over-optimization. These are two different animals. For good or bad, FSB is the best in both. The program shows the results but the trader is the man in charge of making the decision if the strategy has a chance on the market or not.

Popov's Website