oh and forgot to say, of course a great submission as ever!!

here's what i think the correction should be (starting line 60):

    if (up[index] > up[index - 1])
    {
        HH[customindex] = up[index];
        LH[customindex] = double.NaN;
    }
    if (up[index] < up[index - 1])
    {
        LH[customindex] = up[index];
        HH[customindex] = double.NaN;
    }
    if (down[index] < down[index - 1])
    {
        LL[customindex] = down[index];
        HL[customindex] = double.NaN;
    }
    if (down[index] > down[index - 1])
    {
        HL[customindex] = down[index];
        LL[customindex] = double.NaN;
    }

Hi there.. there is still a small issue. the HL and LH labels are the wrong way round. i.e. the LH is indicating the HL data point and the HL is indicating the LH data point ;)

potentially the only drawback i can see is that you MAYBR used m1 bars as the backtest datasource. it doesn't correlate using IC Markets on raw data for the same period.. shame. but good effort.

should use:

Despite you pleas, i'm afraid i am going to say thank you- LOOOL…

as ever, a splendid example of good programming habits and a willingness to share.

top of the class!

not sure if i'd call ChatGPT a friend or not :D

but yeah, will be almost impossible for even a subset of those to align and fire in a trade. 

 @emaile2008 -  on looking at the source code, this is definitely due to all the activity taking place on the OnStart() event, rather than being OnBar(). You could refactor everything out to OnBar() and add further conditions so as to play when required.

hi @tambekema how do we test this one out?? do you have a time limited demo?

looking fwd to giving this a try if you do.

cheers

jim

hi @tambekema how do we test this one out?? do you have a time limited demo?

looking fwd to giving this a try if you do.

cheers

jim

+5 once again, top drawer submission and totally selfless. thanks once more for providing such high quality working examples.

excellent work - some would say it's a close copy of the plugin called Currency Strength Board

well done for thinking this one out and making it available as a bot, never mind a plugin!!

ok.. how?? :)

thanks marwan.. watched the video. the tick data source definitely needs to be tested with some of the coding approaches (not a criticism, just a long held observation) as we see 76% DD on both equity and balance. if you were able to achive the same equity curve on tick data as on m1, i think you'd have folk breaking your door down for this.

Looking fwd to future updates. thanks so much for your time and efforts on this.

i can become a bit of a bore on this one, BUT.. are you backtesting this using tick data, rather than m1 bars (i.e. as the backtesting data source).. i.e.

Would love to see the results using tick data if that's possible.. thanks.. appreciate your efforts.

look like impressive results. was this backtested against tick data ?? i.e.

Tick data is the ONLY reliable data source to use and results will almost certainly differ significantly if using m1 (or any other bars) as your source.

Would love to try a time limited demo of this and happy to share any results obtained.

thanks ..jim

I've commented before on previous algo shares of yours, but just wanted to say it again.. THANKS for taking the time and effort to share these very useful fully functioning examples, which you share without need for praise, nor payment. A real bonus to the community.

+5

nice ACADEMIC pursuit and nice coding style. It's great to see a solution that does what it says on the tin. however, I would urge folks to never use martingale in any shape or form. the statistical evidence just doesn't support its use. I would urge folks tho' to look at scaled staking based on using balance percentages and/or REDUCING exposure when/where appropriate.

this isn't a criticism on your code, which I've said is pretty neat -just the principle of martingale which i've been a vocal objector to for too many years.

yes, as jay mentions, no matter what timeframe you are targeting, you should use tick data as your historical data source as this represents with 99% accuracy, the data that is occurring in real-time. using m1 as the data source is next to useless and will almost certainly skew your results to produce a very favourable chart which can't be replicated in live. You should adjust this in the [Backtesting Settings] tab as such:

Rather than using:

Once you do this, your backtesting/optimisation should reflect the expected outcomes on LIVE.

Hope this helps.

hi khalid - thanks for your work here, sounds interesting. do you have a demo version that you can share for ctrader??

thanks

jim

thank you for the update :)

do you have a demo or any other info on backtesting etc?? would be interesting to see what things looked like if a bit more detail was exposed.

thanks

really excellent addition and can see many other uses for this (such as pulling json feeds from the TV page etc).

nicely implemented

i have to say, having looked at other algo generators, the code produced by your utility is much less esoteric and is akin to what we as developers would write, thus making it easy to take the basics produced by the tool and manually enhance s required.

Excellent work, looking fwd to seeing the entire set of indicators being included in the package, as well as scope for custom indicators being available.

whoever got you blocked is obviously a complete w*nker as everything you've given is given generously and for free. 

welcome back!!

Luke - good luck with this one. I do know from experience that the lack of a demo version, even with verified fxbook records/ctrader screenshots, is going to be a hard sell (especially on premium prices). 

Might be advisable to offer a time limited version that only lasts for 7 days after installation. That way folk will see similar backtest results and more importantly, live execution results for themselves. Thereafter, the incentive to purchase will be greatly enhanced.

Just my thoughts as i like the sound of the above but would be disinclined to take a punt for the quoted price.

All the best

jim

yup -definitely this appears to just be a default project with no logic in place… wrong bot posted maybe??

what would be life changing would be if it was actually ctrader based -lol. charlatan :)

this has got many similarities to one found here: https://ctrader.com/algos/cbots/show/4178

uncanny ;)

whao - what an amazing contribution. these may not be used for common everyday situations, but are definitely going to plug the gap on those cases where the standard averages just don't cut the mustard.. excellent submission!!

there are a few fair value gap examples on here (tho none that i have immediate links to). I'm sure you know how to identify those in code. from there, it's just a matter of using the ‘gap’ that been discovered to fill the rectangle area and doing some tick maths to determine the centre line (discount area).

I have created my own version of this that also incorporates the ldn/ny killzones as shaded areas. however, it's not a project that is easily shared as it uses a few 3rd party components also. However, what I am saying is that by breaking this problem down into bite sized chunks, you'll be able to create the indicator you're after. I'll see what bits of code i do have that would help with the FVG identification tho.. and drop here if it'seasily extracted from my codebase.

might be an incorrect url, but your live fxbook returns a 404 error. 

the devil is in the detail!!

quick question. are you testing this against tick or m1 bars?? this is an important factor with this type of bot as both will produce wildly differing results..

thanks

hello there.. looks interesting (if a bit pricey without a test version).

What I'm not able to see from the above is whether you are testing against 99% accurate tick, or if the default m1 bars are being used as a reference for the backtest. I know from my own testing that the results vary dramatically between tick data and m1 bars as source.

ping me privately if you prefer (jim(dot)tollan(at)me(dot)com) . 

cheers

jim

I've said it before on this work, but this has great potential. 

Quick question tho, I notice above that you mention using Psar/WWMA etc. However, on downloading this from ClickAlgo, I couldn't see any parameters for those. Is this due in a future release??

Anyway - loving the progression here, feel it could be quite special once fully formed and as i mentioned before, would love to see a cBot using a similar ML approach as that would be really cutting edge.

thanks again

sorry -just saw that you had responded. what i was meaning was this it would be quite nice to see a cBot that used the same machine learning capabilities/engine that you mention is core to this indicator. that would then allow a bot to place trades based on having assessed the market as being bullish or bearish. an additional layer of interest could be that the machine learning runs over multiple timeframes to offer confluence to support the trade entry and exit.

just a few random thoughts :)

you've certainly done your homework here. i like that you wrap up a number of strategy approaches inside a single cBot - that really allows for targeted backtesting.

will definitely be taking a good look over the coming week or so. thanks for your efforts…

for anyone else reading (re my response below this), i made a balanced comment about my suspicion on this submission: https://ctrader.com/algos/cbots/show/3955

… thereafter, there seemed to be a few comments (from breckon556) that targeted submissions that i'd commented on. the amateur sleuth in me deduced that there was a  casual connection. 

cynically, some might say that by bad mouthing submissions that i'd praised, my observations on said bot (https://ctrader.com/algos/cbots/show/3955)  would be rendered obsolete ;)

breckon556 - that's quite a bold accusation. i've trialled this guys works in the past and they are pretty decent value.  i doubt this is a get-rich-quick scheme. 

fwiw - the description and the images capture a lot of the base elements of ICT. i for one would probably take the chance on $29.

btw - just marvelling at the odds of you joining today and the only comments made are in threads that i've commented on. maybe i should do the lottery this week!!

what's more interesting is the number of folk leaving comments that literally only joined today and guess what, their first ever review all happens to be on this bot - no others out there.. lol

if i'm honest, my guess is that there's no AI (embedded in the bot) and it's just a simple fast/slow moving average bot. just a hunch!

an interesting addition would be to have the ldn and ny session killzones added as configurable ranges as these closely tie in with ICT trading timezones. likewise, adding session high/lows would help identify potential entry points.

just my random thoughts.

i have to add, as someone that “attempts” to trade using ICT concepts, this could be a very useful indicator, especially if it can be used inside a cBot to semi automate entry and exits!!

great work

would love to see a cbot using the same approaches that you use in this indicator, or conversely, an example that uses the indicator inside a cbot…

great stuff.

I have to say, your builder platform looks pretty promising. Would be interested to see the roadmap for releases as this could be a great tool for kicking off cBots which are later handcrafted in visual studio.

for anyone interested, here's the link to the builder (i googled AlgoBuilderX to see what it looked like)

https://algobuilderx.com/

fair enough - so why not do the reader a courtesy and put some text into the description to alert to that fact…

as i said, if we don't object to stuff like this, the crowds will just keep on walking by! I hope you understand my point.

argh… again, more clutter and non cTrader targetted rubbish…

moderators, please occasionally do a bit of erm, moderation. the uploads area is becoming a sh!tfest of rubbish due to a complete lack of oversight and supervised moderation.

vince -please check your uploads before committing them - this is a waste of time as there is no info, no upload, no clue as to what it does. if this demonstrates your level of attention to detail, i have some misgivings about what the PAID content quality must be like.

sorry if this sounds like an attack but if we don't self moderate rubbish, then we all suffer.

this is an interesting approach. i've been kinda sidetracked of late looking at ICT concepts and therefore have been ignoring traditional TA. However, I have always found VWAP to be quite attractive representation of price action and using it in this way may definitely allow one to SEE those narrow bands of indecision prior to price change.

will definitely look into this…

thanks  -simple is good!! ill take it for a spin

when i email you today at 16:30UTC, it will be the 4th email requesting further info on this bot.

the fact that you don't answer or acknowledge your emails says a lot about how the customer care side of things are likely to go here if anyone were to sign up for this bot.

shoddy….

what an excellent concept. this is something I do all the time in other walks of life (strategy builders/rules editors for business processes). Hadn't thought about it in terms of cTrader, but definitely makes good sense to have something that allows you to quickly pull up a quick foundation for a rule.

well done and excellent contribution.

I know you don't specifically mention in the preamble, but; are you using m1 bar data, or tick data for your back-testing. I've noted it numerous times that m1 bar data will give apparently stunning results only to be confounded when trading live. However, if using tick data, the results are in a very similar ballpark +/-5%.

Apologies if this sounds condescending, but was curious to know what data was used for the back-tests as this will vary wildly in reality if m1 bar data is used.

what's even funnier than the fact that the code isn't appropriate for cTrader is the fact that 39 folk have downloaded it -lol

nice profit profile. would be interested to know why you didn't profile the entire period on your backtest. you only went as far as 23 june -was there anything in the trading window post 23rd june that altered the profile??

nice starting point for comparing different candle patterns. one little nice change might be to define how many candles in a row, rather than a static 3. this code might work:

        [Parameter("Number in a row", DefaultValue = 3, MinValue = 1, MaxValue = 6, Group = "Standard")]
        public int ConsecutiveCount
        {
            get;
            set;
        }

        private bool IsConsecutiveBullishCandles(int index)
        {
            bool isConsecutive = true;
            
            for (int i = 1; i <= ConsecutiveCount; i++)
            {
                isConsecutive = isConsecutive && IsBullishCandle(index - i);
            }
            return isConsecutive;
        }

        private bool IsConsecutiveBearishCandles(int index)
        {
            bool isConsecutive = true;
            
            for (int i = 1; i <= ConsecutiveCount; i++)
            {
                isConsecutive = isConsecutive && IsBearishCandle(index - i);
            }
            return isConsecutive;
        }

just a thought.. cheers...

one small thing - I can't seem to find a reference anywhere to the Session class. Anyone -any thoughts??

private readonly List<Session> _sessionList = new();

??

ricky - this is EXACTLY what i'm going to do with this indicator. I found a neat strategy definition that will fit this perfectly..

https://www.youtube.com/watch?v=vfRjJheqgVw

hi logictrader. I feel you're going to have to produce more compelling details and also offer up a trial version to entice buyers. as it stands, it would be quite a push for anyone to invest $5k in a cBot without having run it on their own system and without having seen sight of the required parameters or any Trade Statistics.

just my humble 2c

Hi there - very good results when running similar backtests on ICM. However - everything seems to go off when we run from feb 2023 - jun 2023. did something occur in the market over these months that the bot is unable to deal with??

It's such a shame as I had very strong feelings about this one - maybe a future version will address whatever anomaly is occurring since Feb 2023.. anyway, thanks again for this bot, otherwise works fine on the historic data you modelled it on - really appreciate the effort!!

i'm just trying to think about adding either a Fractals or a Divergence indicator for confirmation/confluence. again - WIP!!

whao - this indicator literally wraps up a lot of the insights that i've been trying to *capture* regards supply and demand zones. or rather should i say, this targets those zones in a completely diffetent way without having to map out the zones. i think this has huge potential to provide high quality entry/exits based on your descriptions.

will let you know how it goes... thanks for sharing

is the ClickAlgo trial version available yet?? would like to give this a run thro!! 

yes, i made that mistake early on and found that switching to raw tick data required a pretty major revision on my optimisations and to an extent, my logic!! just be aware.