Twoblink and "The Secret Sauce"

TwoBlink,

I think your approach as an explorer of both the mathematics and the idea of mechanical and innovative trading is right on. It is the ability to move beyond where we are now and stretch our selves that truly brings about change and evolves us and our a technology... Keep on keeping on..

Sorry for my post here and I will be happy to remove it if your ask me to...

Thanks scott

Thanks for the kind words Scott.. I just want to share my ridiculous line of thinking, in hopes it triggers either PRODUCTIVE dialog or helps someone else think in a new direction.

So, back to the show..

As far as lagging indicators, I just had a new way of thinking.

Like everybody else, I think:

http://www.trending123.com/patterns/97e4f4f0.jpg

If I had gotten in 3 bars earlier.. Look at all that wasted pips!!

But, now, I'm kind of thinking something differently. Are all pips the same?? What if those 3 bars were just PADDING, just the rice on the sushi, and I'm waiting on the fish part?

http://www.fotosearch.com/comp/CRT/C...15414-49DG.jpg


Visually, I see 3 bars pointing up, before I'm in, because of the lag of the indicator, but what if I see that as the plate? I'm seeing the Sushi. I'm seeing that I missed the rice and the wasabi..

But what if I looked at it more like this:

http://img78.photobucket.com/albums/...on-sashimi.jpg

and those 3 bars weren't "lag" but merely the plate. Then I'd be eating "just the fish".

What if I look at the trend like a fish.. I don't eat the tail, and I don't eat the head, just the body of the fish.. If I look it like that metaphorically, then entering late seems ok to me.. because you confirmed an entry. Exiting late is ok too, you confirmed an exit...

you could make a better entry by drawing a straight 3 point line but thats not my point.

the point is, whatever your probablilities and indicators tell you, it is most important what you do when they are wrong. but OK, that can be delayed till you find your sauce.

the pips in the start and end of a move can be insignificant if you manage to capture most of the entire move.

keep posting, it is a very interesting read and mind refreshing aswell.

Certainly a MA is one type of slope indicator. Linear regression is another and not based on MA at all. There are many others in mathematics.

Scott

What a dis-respectful and rude post. Leave the man alone to his thinking...

Manni,

2 options:

1) Stop reading my journal
2) Until you are Euler or Newton or Leibniz, try shutting up about things you don't know..

Or Option #3) Do both 1 and 2. Until I see an indicator or a math theorem named after you, I'll assume you don't know any math..

And you wonder why people don't want to share any information with you..


Anyway.. back to the topic at hand...

Attached is the problem at hand that I'm trying to solve.. or I'm wondering if there's a need to solve it at all..

So now we get into the "prediction" vs "interpreting" problem. This one is going to drive me crazy for a while.. However, I've tweaked things to the point, where I think VISUALLY trading is quite possible now; the trend seems obvious and strong, and the trendbreak seems obvious and strong, as do the Support and Resistance channels... I just get ticked off when I see my broker hunting for stops..

I am however, exploring the use of tunnels again, and avoiding them. Also, slope based entires seem to work really well. Wait, I take that back, slope based confirmations seem to work well.

But if again, I'm believing the military, then the filtering should occur FIRST. Don't use bad data.. How do I filter first??

OK, so the question is, before they had computers to calculate Bollinger Bands and the like,.. how did they do it? How did the oldschool people do it?

I love movies, so one of the quotes from the movie "The Italian Job"

"...your dad did it all by touch.."
"Well, technology guarantees speed and accuracy"
"There's no guarantees in this business"

I hope I didn't misquote.. I'm pretty sure this is 100% accurate..

Anyway, started thinking, so at the end of the movie, Charlize Theron, ended up having to do it by touch anyways..

So I'm wondering if I should get rid of all the indicators, and just stare at the screen... do it old school...

While the pursuit of a strategy that you can use to be consistenly profitable is a worthy undertaking, one must keep in mind the information garnered from the likes of posters here or even from some books in print may or may not be true.

The information of Gann and his legacy as proposed by authors such as Williams and Elder are based on the comments of of Gann's estranged son who later recanted his statement.

Gann's trading ability were legendary. Yes he continued to study the markets and tried many new and different things, but to conclude he was not a good trader is very short sighted. If you do some research you will find statements to support this claim of his trading ability.

As to what he left and did not leave as an estate, it stands to reason, if he was as intelligent as I think he was, no one will really know the exact amount. But from business associates and contacts that I have made through the years, that 50 million most like to throw around is substantially lower than what may be the actual.

So I'm looking at tunnels and responses right now..

I'm almost close to arguing that the market is a conditioned response.. I ring the bell, the people short the Yen

So if I apply Pavlov's theory to the forex market, that means all I need to do is to FIND THE BELL..

have you ever experimented with neural networks or even markov chains?

the forex market is rapidly changing and the trading system needs to adapt (hopefuly with no or little lag).

Yes, CRM114 is probably the most accurate Markovian discriminator, and like other spam engines, it's adaptive..

I think it's like those 3D pictures you see at the mall, I can almost make out a shade of the image itself, but not quite.. I'm waiting for the 3D pic to pop out and hit me in the face..

That said, based on what I've done today, I'm working with a system, that in its current form, is kicking OCC's butt, and that's a startling surprise to me..

Every system I've come up with, has improved upon the previous one in accuracy, and dropped about 50% of it's weight in the process..

"Agile and Responsive", "Simple and Effective". So far, I'm meeting these.

Tomorrow, I will be slapping some intelligent trailing stop to the system, and seeing how it looks..

Vegas talked about models, I'm talking about probability and conditioned responses..

If I'm CSI and I need some doggie drool, I'll just ring a bell. So if I'm short, and I need to have the price tank, I just have to find the right bell, and everybody shorts.. But I'm probably not the bell ringer, I'm just got to know how to market will react as part of its preconditioning..

Hi TwoBlink,

This is your Journal, which is different from a normal discussion thread. Journal owners have the option to ask others not to post to their Journal. Some people are incapable of complying with such a simple request. Therefore, you just tell me which posts to delete and they are GONE!

Tell me via PM and I'll take care of you. I want to see more of your journal without having to wade through the negativitiy.

Thanks TwoBlink

Dialist,

Thanks. I pretty much had not planned for anybody to reply to my journal (afterall, that's what a journal is) and also planned that if someone did, it would be intelligable questions..

I "hope" all the BS is over, and I can just go on with writing my journal. But it might come down to requesting that ALL other entries which are not mine, be deleted.

Thanks for the offer!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

So, got in a good 5 hours of sleep, and ready for more action!!

I feel like the victim of bate and switch.. What the pros are doing, and the what the pros are selling, aren't the same thing.

I mean, they aren't "lying" to you, but I think the emphasis they put is in the wrong place. Or.. perhaps, it's in the right place, but we aren't listening.


It seems there are 3 parts:

1) The Price
2) The Time
3) The Money Management

Everybody's focused on price, and now, I'm not so convinced that it is where the sauce lies..

So I am going to, for the moment, give up "price" as an avenue of where the sauce is located...

So we shall begin to explore the other two.

Time seems like a fairly strange creature; it's constant, but it's preception keeps changing. The "magic" might lie in which timeframe you are trading off of..
"DayTrading" is a subjective word, 5M charts are daytrading, as are 4H charts..

Time seems a bit difficult to wrap my hands and mind around, so I will leave it for the time being...

Now Money Management, this seems like it could be where the sauce lies. Mathematically, even if I only had 1% chance, with enough money, no limits, and I martingale it, sooner or later, I'd be rich...

So today will be deemed MM day for me.. I will explore MM...

I haven't discovered the secret sauce yet, but I at least now know the ingredients that go into making the secret sauce..

Made a lot of progress today..

Dug up an old article of an interview with Ed Seykota.. He said there is no Holy Grail, but when you read between the lines, he tells the secret!

I'm excited! I'm jumping up and down! I'm dancing like the Roxbury guys!!

"What is love.. baby don't hurt me, don't hurt me, no more!!"

Hi Twoblink,
A note to say your post is fantastic .
Please continue your post.

So I'm having what I call the "Ed Leedskalnin" Dilemma right now..

So how do you tell the Prince from the Pauper when they look alike? How do you know if you've found the secret sauce? This hadn't crossed my mind.. until now. This is a bit of a problem.. I feel like "How do I know this is my child?" before DNA testing was available..

So let's start with the assumption that I know the secret sauce. How could I confirm or deny? Or do I have to bust a Ed Leedskalnin and build it, before it's verified proof?

It seems if I know the sauce, then I just use the sauce, and I'd be rich. But if it was the margarine and not butter, I might not know until I've put it into the mashed potatoes and then, I find out, and now I'm out of potatoes..

So do I just assume that what I currently have, is the secret sauce and just start implementing it?

So the problem I'm currently having is with some of these calculations. They are theoretical in nature, and so I don't know how to work with them..

Grr.. I think this means "If and What-if" trials..

So if x happens, then I do y.. like a paralleled SIMD machine..

But that means a lot of "case specific" rulesets.. Which I don't like..

hi TwoBlink,great job with u'r exploration..

So when I was in junior high, I was suppose to take woodshop as a class (you know, make useless things and kill trees) but the class was full.. and as I tranferred into that school, I had last priority.. so I ended taking... Home Economics..

Let me just say, people who laugh at home ec, just are dumb. You GOTTA eat, and I learned more in that class, than all of the rest of junior high COMBINED.

I un beknownst to me, the girl who later became my first girlfriend 5 years later, was in the same group as me in home ec.

So we had the same ingredients, same teacher, same oven.. my cookies tasted great, her's tasted like crap. It is probably God's way of telling me something; but no gf I've had, was that great of a cook. Now that I'm married, my wife can't cook to save her life. Good thing I took home ec.

So what does home ec have to do with trading?

To the casual observer, my first gf and I, did the exact same thing. How is it possible for two people to do the exact same thing, and get such different results?

So, either Sir Issac Newton was incorrect, or we didn't do the same thing. Guess what? We were doing the same thing. This is why I always think, you can't really make money with someone else's system. You have to tweak it. You have to understand it, and I mean innately understand it. Otherwise, you will think you are doing the exact same thing, and in reality you aren't.

So back to the cookies.. I followed my first gf from taking out and measuring the flour, mixing the dough etc.. not a single mistake. But her cookies just tasted like crap. So.. We started "regression testing". and we tried it forwards and backwards. She's prepare the flour, and I'd mix. (which is kind of stupid, since we used an electric mixer). and she'd scoop the cookie dough out, and place it on the pan. Perfect cookies. So we switched it around, I prepared the flour, and she mixed, and I scooped out the dough.. again, perfect cookies. We broke down all the steps it takes to make cookies, and we did every combination possible..

and the results? If I had ANY involvement in ANY of the steps, the cookies turned out great. The ONLY time they didn't, was when she did 100% of it. To this day, it's true with the cookie for me and my wife..

So.. What makes my cookies so good and her's so horrible? I HAVE NO IDEA. Today, 15 years later, I still don't know.. But what I do know is, she doesn't have an innate understanding of cookies. My wife doesn't understand food either. I'm surrounded by women who don't understand food..

There has to be little subtlties that aren't noticed that some people do that other's done.

But it's really difficult to find that little difference. BUT.. here's the good news.. I caught it once.. (When Mars, Venus, and Neptune were aligned..)

I noted the vegetables my wife cooked tasted salty.. or rather, they tasted bland but was sitting in very salty salt water, a feat I am unable to reproduce.. This however, I knew was because of when she added the salt.. Her timing was off just a little bit in the "when to put seasoning". So, through close scrutiny, you can find the difference. The frustration of course is, there __IS__ a difference.

Reading Ed Seykota's article, and I quote:


So to the untrained eye, Ed, my hero just said that my quest was futile.

But... OBVIOUSLY there __IS__ a secret, otherwise, I'd be making 280,000% on my money too!! I'm just happy when I'm not -28%!! So the secret is hiding in plain view. Ed knows it, but he just doesn't know he knows it, or he knows it, but just doesn't think it is much of a secret.

So, back to baking. I have another friend, her cakes were more like pancakes. She loved to cook and bake, and every year for her birthday, she'd ask me for a Cheesecake for her birthday. It takes me about 11 hours to make a good cheesecake, but I do it for her every year. One day, I was over at her house, as she was prepared to bake. She shoved the cake in the oven, and turned it on..

HELLO!!

I pulled the cake out, and I told her, she's gotta preheat the oven. She said she never knew that.. While we were waiting for the oven to preheat, I took a spoon, tapped the side of the cake pan a few times.. tap tap tap, rotate 5 degrees, tap tap tap, rotate 5 degrees.

So, 10 minutes later.. the oven's nice and warm.. and so we throw the cake in there. I stick a small pan of water under it, so it won't get too dry. I also took a look, and her oven was a convection oven, and so I just the time on the recipe down by about 20%.

So... PERFECT CAKE..

She was so excited.. so giddy..

her cake never rose because she used a cold oven, which is a no no. Her cakes were usually too dry, and so the pan of water helped. Tapping makes the bubbles go out, and so her cake didn't have strange holes... and the circulating air just cooked much faster.

Now nobody told me to do all this, this wasn't on the recipe, but it was just something I knew how to do to compensate for situations that I've met before.. (Experience vs. Education) I guess.. now pretty much any mother who's baked can tell you most of this.. but they don't... The PROBLEM you see, is people "in the know" seem to assume that you know it too.. it's the assumption that is the killer.. So she was so excited about her cake, and how I had done some "slight of hand" to make it all better. ??? Preheating the oven is as old as time.. adding a pan of water to the oven, or tapping to get bubbles out.. all that has been done before...

And then I discovered something.. You pretty much CAN'T FIND A BOOK THAT TALKED ABOUT DOING ALL 3 AT THE SAME TIME. The problem is that the information is not gathered together in one coherent place.

The other problem is that you have no idea how to integrate all these ideas, nor how they all influence the final product, nor how they INFLUENCE EACH OTHER.

So, back to trading again.. I was Made in Taiwan : But grew up in LA. I'm accustomed to pretty much food around the world. Having lived now in Taiwan for almost 4 years, I'm no stranger to Taiwanese cooking. When you understand the people, the culture, you understand the food. Wait, errr.. when you understand the food, then you understand the people and the culture.. something like that. Anyway..

So my white friends come and visit me, and they are just at their wit's end, trying to decide what to eat. They basically think everything here is disgusting... and so they end up eating McDonalds..

This morning though, it struck me.. What if I'm just like the white guy when it comes to forex?? What if, I'm living amongst a bunch of other white guys, and we all grew up on fried chicken and pizza, and so we have no idea who "real" chinese food should taste like? What if we have been AVOIDING the good stuff, because we are too stupid to know it's the good stuff? Caviar? My wife thinks it's just disgusting salty little seeds that smell and taste fishy.

OH MY GOSH, WHAT IF NOT ONLY DO I NOT KNOW WHAT A GOOD TRADE / TRADING SYSTEM LOOKS LIKE, BUT WHEN I SEE ONE, I AVOID IT LIKE MY WIFE AVOIDS CAVIAR???

Houston, we have a problem!

So now I gotta start digging and understand "The Anatomy of a Good Trade", then I can graduate to "The Anatomy of a Good Trading Platform", then I can graduate to "The Anatomy of a Viable Trading Model", then I can graduate into "The Automation of a Trading Model into a Trading System".

If I just said, "A good trade is any trade that makes you money."...

Wait, that's a problem though, because what if I risked 100 pips to make 3 pips? Yes, I made money, but that ratio is ugly like Monica Lewinsky...

So we shall start with the disection of:

The Anatomy of a Good Trade

1) Makes Money
2) Locks in Profits
3) Cuts my Losses
4) Lets my profits run indefinitely
5) Has some sort of emergancy brakes
6) Should know the probability before the trade
7) Should know the risk/reward ratio before hand
8) Should know what the entry signal(s) look like
9) Should know what the exit signal(s) look like
10) Should have NO SURPRISES
11) If I get stopped out, should be able to calculate recoup rate in number of trades
12) Should have risk to reward ratio > 1:4

Unanswered stuff..
a) Is there such a thing as "optimized" timeframe..
b) 4 shifting variables:

1) Bet/Lot size
2) Stop loss size
3) When to take profit
4) Do we get in, during that bar, or at the close of the bar / open of the next bar.

My OCC really showed its weakness when livetrading vs backtesting because of when you enter. OCC was based on beautiful entries and great money management scheme. But if the beautiful entry is taken out, it is not just a lack luster trading system, at least on the 4H. If I had to trade next bar with OCC on the 4H, that's way too long. When you have 4 variables, how do you optimize them since they effect each other so greatly, and thus your profit? Run it through a computer simulator?

I pretty much an still wondering how the people of the old did it before computers... But according to Ed's article, he spent half a year programming this stuff... So I guess they did it the hardway.

We can pretty much calculate ATR.. We can take the "Day" as the fundamental unit of Forex. We calculate the ATR for it, and then calculate the average movement per timeframe, and this should yield us, the density of price movement per bar..

That's a great idea..! Gonna do that later tonight!! I'm going to plot price movement density to timeframe. Shove the answers into excel, and take a look at the graphs. This will yield the "optimized" timeframe. Or does it? It will yield the densist timeframe, which does me good how?!? Great question, get it done first, and see if anything shows up..

http://daytrading.about.com/cs/educa...l/aa092799.htm

Excellent!!

http://www.investopedia.com/articles.../04/091504.asp

Excellent as well! That about.com page really has me thinking..

Multiple variables are difficult to adjust for.. So I have to nail down the personality first.. Then I can roll out percentages of winners vs win/loss ratio vs. bet size. But that "might" end up determining my time period for me..

I'm almost convinced that timeframe is 1H ~ 4H. Hmm.. time to write some simulations...

That was fast.. about 30 seconds after I finished the last posting to the journal, I started writing down a few things that were very interesting to me..

And immediately, am faced with a problem that is fundamental and I don't know if I can overcome it..

So I called it the "Cow lick vs Me Love You Long Time" problem.

When you have shorter MA's, you react faster, but you get quite a few "cow licks", ones bar might be spikey and it ruins everything else..

BUT.. if you move the MA to longer time periods, "Me Love you Long Time", the curves are smooth, as there is less weight given to the cow licks, BUT the reaction time is like that of the kids on the short bus..

So false signals vs delayed reactions..

I know I know!! Time to start reading about DSP signal smoothing!!

There is this machine that TV companies buy..

The First generation machines: What it does is, it checks out the amount of "white" on the screen. If the white is over 80% or something like that, then they snipe 1/10th of a second from the frame.. So blown frames get stolen some time.

The Second Generation Machines: Not only do they steal time from the white canvas screen shots, they now do interpolated comparisons, to see how much "difference" from the first screen to the next. If it's less than 1%, then they snip a 1/10th of a second off as well..

I kept asking, Does this actually do anything???

My buddy in Burbank gave me the math of it..

"See, we can steal about 1.9 seconds per 1/30 hour show.."

Yipee, wow, big deal.. but wait, it __IS__ a big deal..

so that comes out to about 3.8 seconds per hour.
3.8* 24 = 91.2 seconds.

Each commercial averages about 30 seconds. That means this machine, will steal enough time for THREE commericals a day for them.. Pays for itself within a few months!!

So I have decided, darn it!! I'm going to do the same!!

So...

I took another look at OCC.. but in a different light.. and.. it's lookin' pretty good!! With the newer modifications, it's doing quite well!

What did I do?? Phase shift.. rather timeshift.

If I know how much "lag" I'm getting, I just move it back that much, and tada!! No more lag.. well, actually, that means my MA's are correctly super imposed on their bars..

this seems to be a great fortune teller.. When I use it with something more accurate...

So right now, I have OCC based on Hull's Moving Average, instead of EMA, and I have it -5 bars. EXCELLENT.

The reason is the slope change is very easy to see, and prepares you for the market moving against you..

But you can't cheat father time.. So now my history is accurate, I'm missing 5 bars from my indicator...

So now I have accurate history, but not accurate present information.. This is bad..