The touched lines and the signal arrows stay with the chart now. Lines are missing from the first half of the day, as the code came in around that time.

The one major thing missing from this journal is an analysis of the day, with what I can learn from the trades taken and from the ones not taken. I did not have time to study too much what went well or wrong. Journaling takes time from coding and this must be the one thing I don't like about doing it. Also, I was expecting more interesting discussions with hardcore traders.

For yesterday, with what I can see on the chart and not having all the alerts here, it would be something like this:
-Good entries: long at the bottom of 4 and short at the top of 5. They worked.
-I had three counter trend entries on bar 19 and 20, which were losses at the full extent of the stop order. I should probably always look at the TICK in the last few bars and, if there was no sign the bears can take the TICK below -300 or so, then no entry counter trend.
-anually, I would have entered at 1351.75 and my strategy generated a signal there for a different reason, but was overridden by something within the context, probably the target based on the gap not being reached yet. It was a pretty small bounce, but there would have been no drawdown, which is how I like to trade.
-Some good signals on bar 43, bottom of 51, 52, 67, all rejected for reasons I don’t have now, since I cleaned up the database. I will make sure to keep the detailed history for this kind of journal entry, so I can learn more from it.

ToDo:
-Create a new entry type based on the calculation that gave me the 1351.75 short.
-Code to look back at the last few TICK bars and new exception type.
-Qualify the news that keep me out of trades, even when the type should not impact the index. Not sure where from yet. For now, only look at the USD based news.
-Work on the trading range code.
-Now that the code to cache certain passive chart objects, optimize the code to save only the ones I need exactly when they get generated instead of regularly looping through the chart objects and filter by attributes.
-Active lines, sensitive to touch, get generated automatically on the 5min chart. Also, I can send active lines from the 15min and 60min charts to the 5min chart. I need code that automatically generates good lines on 15min and 60min, so I don’t have to send them myself.

The day started well yesterday, too. Even with the manual long entry at 1357.25 missed by one tick, I am still up. The strategy generated another good long entry at 1357.5, the low of today, but rejected again because of the natural gas news coming up. I really need the code to look at the possible impact of the news and not keep me out for any unrelated news.

I guess they realized how natural gas is ubiquitous and cheap, because the price is at around 10 years low now. J

nakachalet 

...and/or for those stingy traders placing tight stops to demonstrate their cowardliness toward unplanned and unforeseen risks.

Not sure I get the idea about how tight stops demonstrate that or what is wrong about it. I thought even your trading plan also relies on getting out of trades as fast as you know the trade is no good. I can see a good number of reasons where tight stops make sense. I have many trades where a tight stop is required because of the reason of the entry. I also have code that looks at the PnL and gets me out if over a certain value and when the conditions of the entry don't look good.

nakachalet 

i'll try to go back sometimes over the weekend to my archive to locate those pix with their captions and repost just the captions for those pix, K? so those pix would make more sense, my apology <wish that whoever writes the program codes would in the near future include captions that come with each corresponding pix, pls. thx in advance, coders.>

Thank you very much! Looking forward to studying your captions.

It is funny you mention how programmers should work on captions. I thought about the same listening to Al Brooks' summary at the end of the day and, because my indicator collects a lot of alerts and events, I now have my own web page that shows me the "story" of the day, bar by bar. It's a lot of work and I don't think it is worth doing it only for the sake of making others understand how your chart went, but a nice thing to have, for sure.

nakachalet 

pajot, you are in luck, just have had a loser on 6e a couple of minutes ago and it was on video too. could have limit loss to just a 3 or 4 tics but then the label would coverup other significant info, so stop was placed at 7 tics; sold 2 6e at 1.3137, and subsequent stop was placed at 1.3244. that short video is attached. you know, traders who eat, sleep and play with their chosen indic; could pretty much foretell by looking at their indicators acting in concert, whether or not, if they should bail out before their stops would get hit. it is like foreseeing what is going to happen in the next few seconds before the event actually takes place. enjoy the short video at my expense.... lol

Thank you for the video! What do you capture videos with, please? I also like traders who post the mp4 file, since I can see it on my iPad.

Regarding the "foreseeing", my own theory says that whatever our eyes and minds can notice by looking at a chart with price action and indicators, an algorithm can do it better if written properly and based on a correct description of the "feelings" traders get when looking at a chart. I know our brain’s activity is hard to mimic outside our heads, but the major reason for why I think a computer can do this better is that the domain is very limited. When doing technical analysis and price action, there are not many things on a chart, and it is mostly about math. Even with angles, velocity, patterns, if we can take the time and understand why we get a certain feeling and under what exact circumstances, it can all be programmed.

nakachalet 

pajot, you are a blessed programer who can write your own codes but numerous others like myself have no idea whatever about programming. wish i could be as talented as you are, so i can work on my own codes too; but i just have no talent in programming--got to repeat a class in FORTRAN (?) in '65.... never could forget that.... LOL

Glad you can tell. I wasn’t sure it was coming across. LOL. Thank you!

nakachalet 

pajot, i can categorically assure you that those practically 'holy grail' indicators are practically all here buried chest deep in indicators section. and it only costs 50 usd to avail oneself to everything, practically....

If you don’t want to post them in the open, PM me directly, please. J

nakachalet 

know what, pajot? there is nothing personal in what i am going to say, ok? but it is an idiotic nakachalet axiom that.... what a trader wants to do or wants to learn on his/her personal level, is really entirely irrelevant to trading successes....

How would I possibly get offended by what you said!? I completely agree with you here. It is my path to do it in code and I know I have taken the hard way. I wanted to, because I love programming and solving hard problems. It is an “excuse” to add more chart time to my education without losing money. It also helps me understand how the market works and the code is a way to persist that understanding outside of my head.

nakachalet 

if a trader wants to become more successful, more profitable, more consistent; the shortest avenue to that utopia is NOT what that trader wants to do or wants to learn.... but to do what other successful, profitable and consistent traders practise, day in and day out.... until they become habitual in one's own trading....

I guess my style is different here. I love to solve problems without being too interested in how other people do it. I could learn and I am willing to save myself some time, but it is hard to discern between people who BS as a rule and have no idea what they are doing and people with substance. The effort of trying to mimic others, not knowing what works and not, is too much for me. I have tried and wasted my time.

I am very aware of the fact that money can be made in a lot of simpler ways and I wouldn’t spend hundreds of hours on thousands of lines of code if I had it. Trading is a way to achieve my goals and believe me when I say that I would rather be on a boat in the middle of the ocean or circumnavigating with my daughter than spending my days and nights programming. I just don’t have a good system that works yet. Again, if you are willing to share, PM me! J

nakachalet 

pajot, you will be surprised that there are many greater, much greater traders on board here, each specializes in his/her own instruments and markets. and to become proficient at even only 50% of what they are doing.... would take two or three years of on line live screen time. many traders whom i admire immensely also have already practically spelled out everything concerning what a successful trader must develop personally.... spend some extra time on what they counsel, if you have time, K?

I am approaching three years now, with most of the trading days covered for the full extent of the RTH and some overnighters.

nakachalet 

pajot dear, i am really not that bold in telling you or others what to do. by no means, pajot, believe me. nevertheless, just humbly wish that you would not try your very best to REINVENT the wheel or whatever indic are already available on board.

PM!

nakachalet 

happy and profitable trading everyone. apologized for my babbling. now, enjoy my losing 6e trade.

No babbling here, but honest and open opinions. I appreciate you taking the time to come back after those out of the blue posts. Thank you, again!

Right now, my strategy allows both short and long entries, because I collect stats for each entry type. Of course, there are other restrictions that may cause skewed stats, like the minimum allowed distance for doubling down and the fact that each signal type has an importance assigned by me, which makes certain signals have priority above others.

I was in a long ES trade today, with 1.75 profit at 1365.5, with a target of 2 points, so 1 tick away. A very visible measured move target from a lot earlier was sitting at exactly 1365.5, but my strategy did not decide to take the 1.75 points profit because of that, so the price bounced and I got stopped at two ticks profit.

This example and many other situations give me a permanent ToDo for the future of my strategy: implement tweaks for current trades and adjust decisions for new ones based on upcoming predicted action. I put something together with that in mind, a while back, but I need to actually process the info in there. Right now, I only display it for my eyes. A simplified version looks like this and it gets updated with each new price level:



, so I have some levels in my code and I need to use them more. I don’t check this list too much now, for my personal benefit, because all the levels are visible on the chart anyway.

Another slow day in ES and I can see how people have enogh action to work with, but I definitely enjoy faster days, longer swings, more trending. I think it has to do with my style. I like to be more like a sniper, in and out quickly. That's another reason I am working on my code: my strategy (I hope) will never get bored.

I could see how I had nice profits on a couple of trades and my strategy kept me in and even allowed one to turn into a (small) loser. I will adjust the ATM for that particular entry type.



Studying the stats NT helps collecting is a must, as they will tell things over long term that you may otherwise miss. The way I do my trades is to have a dedicated ATM for each entry type and I love how NT let's you see which ATM was used for each trade, so you can draw conclusions.

Here are the individual trades. I could see trades unfolding and would have moved things around, mostly take some profit on that large losing trade, but I won't interfere with my ATM that way, to keep things honest.



Regarding the chart, I love some of the entries and, most of all, the call on the manually drawn channel, which bounced on bars 62 and 69, at the exact tick.

Also, my calculation was calling for a 1357.25 bounce, but ES bounced at 1357.5 and that became the low of the day. Pretty close. I like the automatic entry with trend on bar 67, a good sign that I am not only looking for reversals and counter trend entries.

Just checked why the entry on the channel line on bar 62 was rejected (the gray arrow up). The reason was the "strong" bar right before, against the order. That is not a strong bear bar, so I have to check my code for how I define a strong bar there. Damn!

Here is my first stab at sharing the chart on futures.io (formerly BMT) and, to be honest, ever. I just ran a replay of today, at 50x speed, at high speed. I usually clean up my database before I run replays, so the couple of trendlines from yesterday were not there. The channel I drew today is not there either. It's just a clean run with all default settings. The speed also alters the results a bit, because of the lag in reacting to the price. I will try to post a replay every day, but I will probably miss the start of some in my tick recordings. Enjoy!

March 8, 2012 ES replay - YouTube

March 5th, 2012 replay - a bad day - YouTube
Sucky day, a lot of tweaking work. Still, an interesting day for what can be learned from it.

Another less than great day, at least for me and my indicator: -$169.66. On a positive note, I have figured out why the 15min and 60min EMAs do not plot properly on the 5min chart and will hopefully fix it soon.

Here are the trades so far:



and the chart:



I am wondering what traders do in days when the ATR stays low. Are you adjusting your ATM depending on the size of the ATR? Thanks!

I guess the only positive thing after days like this is that the strategy is not screwing up big time, so I don't end up bankrupt. Other than that, days like this trick my indicator because I don't handle the tight trading range properly. I used to have an exception that prevents entries if the ATR is below 1.5 - 2 points, but I turned it off after days with barely any trades.