Cursor Chart Instant Repositioning feature request.

Questions about MultiCharts and user contributed studies.
bowlesj3
Posts: 2180
Joined: 21 Jul 2007
Has thanked: 227 times
Been thanked: 429 times

Cursor Chart Instant Repositioning feature request.

Postby bowlesj3 » 09 Jul 2009

This idea is directly related to the “replay cursor instant start” idea I submitted on this thread. http://forum.tssupport.com/viewtopic.ph ... ght=replay

Here is how the feature would be used. The trader positions the cursor on a selected bar (in my case a bar on the “10-second-bar-chart”) and then they press the back quote key (or any user selected key which is free for use preferably on the left side of the keyboard for a right handed person) which causes the chart to jump to hide all bars after the bar where the cursor was located when you press this key (it is important to note that only the one chart jumps – not all charts in sync - however it might be handy to have this as an option somehow). I would be using it to speed up a manual replay technique I have developed just last night which I figure is going to improve my trading a lot. If any short term discretionary traders are interested I can post it (it is not applicable to longer term trading). It could be a bit long.

Thanks,
John.

User avatar
arnie
Posts: 1594
Joined: 11 Feb 2009
Location: Portugal
Has thanked: 481 times
Been thanked: 514 times

Postby arnie » 09 Jul 2009

Interesting idea John.

I'd like to see that replay in action.

Regards,
Fernando

bowlesj3
Posts: 2180
Joined: 21 Jul 2007
Has thanked: 227 times
Been thanked: 429 times

Postby bowlesj3 » 09 Jul 2009

Hi Fernando,

Long one.

Why I figure there are more discretionary traders.
I have no way of knowing for sure but I suspect replay is going to be used mostly by the short-term discretionary traders (not so much by long term). We have discussed our systems privately somewhat, as you know. As you know I don’t attempt to fully automate my system even though I have done a lot of programming and am very comfortable with it. The trade setup is way too complex and the triggering is boarder line too complex (I don’t think I could ever trust it fully). Seriously I figure it could take 100,000 lines of code to fully automate my whole system and as great as I think MC is I don’t think my machine or MC could handle that (I might be dead by the time I am done and missed all those manual profits I could have made to boot). So I restrict MC to a workload reduction tool and use it as much as I can. I normally try to enter on bounce trades (sudden price direction changes) and stalls before this (sometimes after and sometimes I use the stall to decide between 2 or 3 walls that are close together). I find most canned indicators and all indicators that have any averaging of any form at all are way too late for me so that is why I do this. Although most traders on this forum may be (trying at least) for full automation I suspect that there are actually more discretionary traders who trade by price action or technical methods. To find them you may need to look outside the forum. If I find it too complex others will too. I often wonder if TSS has surveyed this externally and internally to the forum. Anyway here is the manual method I am now using for replay. The nightly practice lines up perfectly with what I am doing during the day but is much faster.

How one might use “Cursor Instant Chart Positioning”
You have to practice your technique to be able to trade a system like this. I now practice the completed trading day every single night without fail (kind of like Tiger Woods practices his game). I don’t practice the setup portion of the trade but rather the entry point technique. It is done without a live feed. I now do it by first positioning the chart to a point a few small waves or tiny waves before the actual best entry point for the trade. I see the small waves on the 10-second bar chart. There are about 20 to 30 of these trades every day so that is a lot of chart positioning and that is why the above idea would be a useful speedup for me. Next I use the left hand to press the right arrow to move the chart bar-by-bar and I press the pause key when I want to execute (pause causes my database program to force TWS to send out the order before auto-fire occurs). My eyes do a 3 cycle rotational sequence when each new bar appears (3 bars and the sequence starts again – during live trading this would be 10 seconds for each task but during practice I do it a lot faster). So the manual control of speed using the right arrow allows me to choose the speed of the moment depending on how sleepy I am after a hard trading day :-) Actually it is true (not a joke). Part of the process involves my systematically looking for a stall near the wall bounce that I have programmed MC and my database program to trigger. On the 3 way cycle of 3 bars I rotate my attention to these tasks (check for forward stall, check for backward stall, syc myself up). Yes, I still have a reflex hesitation habit but it is getting better and this is the main reason I need to practice training my body to just react when I see the stall trigger (don’t think – just do it – most of the thinking was done during the setup phase using a very short checklist to stay consistent). Hesitation is kind of like an eggshell. It is very valuable when you are learning to protect your account but eventually it becomes a negative force and you have to break through it before you can fly properly so to speak. Old habits die-hard and it is really hard to break this habit when you are also trying to read the market price action for price stalls. That is why I actually hit the button during practice (it is a physical thing). By the way, during the live trading day the database program and a text reader announces which of the three items I should focus on (check for forward stall, check for backward stall, syc myself up). All three activities are critical for me so I have to have the programming ensure I cover them off systematically (don’t totally neglect one over the other). After having said all this I am flexible with this rotation both during the trading day and during the replay at night meaning it s not always one item per bar. I am starting to think that 10 seconds for all stall checks and 30 seconds for syc up is about right. They are now parameters in my database program. Regardless of these details or improvements, the main idea is to use the replay to get in the habit of ensuring that stall checking and button pressing become a fast efficient reflex habit.

I have read that some traders can’t break this hesitation and they try to fully automate to work around it. I would rather just attack the problem directly using replay practice. If it hits the bounce before the stall the syc up is to make sure I don’t turn it off. I practice this as well at night but I have it programmed to turn it off if the approach exceeds a tolerance I have. At times I delay it and enter on a subsequent stall (often it is a break and maybe a surge - an expansion ending to the parent wave). This is another reason I don’t fully automate. There is no way to know this is going to happen until it speeds up excessively out of context with prior like waves.

By the way, I figure you could do the approach I do during the live trading day with MC alone by using the playsound feature in MC to play a wave file that suggests what to focus on as the trade approaches. To tell it to start the 3-way focus as trigger approaches maybe you could place an arrow out front of price to signal you want the announcement every 10 seconds. Nightly practice would be the same as I described earlier.

I tried it live today and it is a lot better. Much more flexible than what I was doing before. It will take a few weeks to see the full benefit I figure.

A real replay would make it seem more real. The ability to run the studies during a real replay may be handy. For it to match what I am doing with the right arrow step action it would need a very fast speed up and slow down feature. I think both features are very useful.


John.
Last edited by bowlesj3 on 13 Jul 2009, edited 1 time in total.

bowlesj3
Posts: 2180
Joined: 21 Jul 2007
Has thanked: 227 times
Been thanked: 429 times

Postby bowlesj3 » 11 Jul 2009

Here is a very interesting idea to consider for your replay TSS. Have a feature that can process a bars worth of data to run the studies if the trader presses a button (that way the studies could run and the trader could control the replay in a stepping action rather than have it run continuous). This in addition to the ability to instantly change the speed by pressing 0 to 9 while it is running in a continuous mode. Make it so that pressing this step action button instantly switches it to step action and pressing one of the 0 to 9 buttons changes it to continous replay at the desired speed. All this on top of the ability for the trader to position the cursor and press a button to have it instantly restart from a new location (staying in the same mode it was in before - step or continous). The chart that is highlighted would control the step action bar and the instant restart with the cursor too.

Something to consider. I certainly would use all these features.

User avatar
arnie
Posts: 1594
Joined: 11 Feb 2009
Location: Portugal
Has thanked: 481 times
Been thanked: 514 times

Postby arnie » 11 Jul 2009

Hi John.

Very good stuff what you've written.

Basically, you'd use the "replay" feature to automate yourself instead of "automate" your MC written study.

Playing over and over the trading day and studying all setups generated for that same trading day would create the phsycological path necessary to trade as if a complete automated trading system was being applied.

We could select a specific bar, or a specific minute to start that replay. We also could, when selecting which bar ou minute to start the replay, chose the speed of that same replay as you said. For example, 0 to 9 could be, 0 for continuous mode and 1 til 9 indicating between 1 and 9 seconds for each bar.

It's natural that full automation is the goal since this would leave you with a lot less stress moments, but it's also natural that not all trading systems are apt to be fully or partially automated. The trader must have the clarity to acknowledge where he is and where his trading system is so he can act and not wasting his time trying to accomplish the impossible.

Regards,
Fernando

bowlesj3
Posts: 2180
Joined: 21 Jul 2007
Has thanked: 227 times
Been thanked: 429 times

Postby bowlesj3 » 11 Jul 2009

Hi Fernando,

I am not sure if I would use the term "Automate Myself". LOL. The setup portion of the trade (which for me is analyzing the implied direction of at least one level of wave above the one I am trading) is not practiced on replay. Also the finding of bounce points is not automated (I actually did but I dumped it since I got it right every time and my program screwed up now and then because new situations would keep poping up). The problem is that price bars come in so many combos and waves in so many shapes and sizes with sub waves, (not to mention that there is no way to know what shape is coming next that has never actually existed before) that you can't really trust your code to get it right every time. Gee sometimes I look at them and it takes a while to figure the how to actually mark the tops and bottoms and in the end I create a few variations to come up with an assessment of where the market is pointing (how can I write code for something like that - after 25 years of programming I know as well as any programmer that computer code is for speeding up something that happens exactly the same every time - waves do not do this, they vary greatly mostly because of the bounces that distort them and especially when these bounces layer and come into conflict). When they shrink and conflict into a channel or triangle it gets harder. You can't practice finding those in a systematic way. I have a procedure for contraction and conflict but I just remember it. Another thing, some people might try and sell you on their artificial intelligence program that could do all this. However why would anyone want to pay a lot of money for artificial intelligence when real intelligence is free and works a lot better?

It is the triggering that I practice because it happens fairly fast. It is a systematic way of analyzing the two levels of waves below the wave I am trading to determine if it is going to stop short of the bounce (or exceed it). If done well I should be able to get an entry point from 1 unit to 3 units off the top or bottom of the wave I am trading. Typically bounce entries are within 2 units of the top or bottom if you select the correct bounce. I use to use MacD at the very beginning. If you wait for the cross over you loose way too much profit (all averaging systems are the same this way). Bounce conflicts will whip saw any averaging type of system. My system can handle all this conflict as fast as I can see it and normally I see it approaching before it actually conflicts (assuming I don't make a mistake). Human error is the trade off. That is why I am thinking of team trading.


For replay, 0 to 9 would be some sort of speed adjustment. Probably from 0 normal speed to 9 which is gradually faster. This idea is directly from the method the Micro Focus cobol debugger uses. Maybe the back quote key could be step mode it TSS decided to put that feature in.

John.

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 15 Jul 2009

Thanks for the suggestions. We will keep them in mind.
We are working on replay feature and I think it will not disappoint you.
It will be available in MultiCharts 6.

bowlesj3
Posts: 2180
Joined: 21 Jul 2007
Has thanked: 227 times
Been thanked: 429 times

Postby bowlesj3 » 15 Jul 2009

Thanks Andrew.

I am curious. How will the data come in. Will a trader be able to pull up their work space off-line as they do now and start the replay without having to do any copy of data out to a special symbol (or separate data source as a copy). If they can go directly from normal off-line viewing to replay that would be very convenient. Even if it is necessary to press a simple copy button on the menu to prep the replay that would not be too bad (if that same copy of the data could be used for both replay and regular off-line viewing of the whole trading day). Either way I think there is a need to be able to quickly flip between replay anywhere on the chart and regular viewing with all the features we have now.

The other thing that occured to me is the step idea. If the step was put in, the speeds would still be in effect while the trader was stepping. In other words if they did a step on the 1 minute bars the current speed setting would effect how fast the data ran to get to the next 1 minute bar before it stopped.

I can't speak for other traders here but if the replay can actually run the studies, I could exactly practice the trading day running my database as I do during the day. As an example I have several ways I trigger including 3 ways I detect forward stall. Currently MC feeds info over to the database program and I get notified of these via popup menus with verbal notification and I have to decide if the final trigger should be set because there is no way a program can have the full overview I have of the market. If I decide to arm one of these it is armed to trigger on the next bar to bar stall (meaning I am not really just suppose to press a button to fire out the order immediately all the time - in some situations yes). So these every fast decisions could be practiced for the completed trading days over many weeks until they become auto pilot reactions rather than fairly slow decisions I have to think out.

Thanks again,
John.


Return to “MultiCharts”