The week starts out 'interesting'. Gold is only a few cents from the 'magical' $1600 level. I think the only magical level is $10,000, and the musings on when that gets taken out. If it does, and the moves were proportional, that prices gas at the pump around $30/gal and crude around $400. Imagine food prices then. The Swissie is exploding, and the EUR/CHF just set another new record low, as parity is now the talk. And this has our futures down a whopping 1.75 points. Wow, pretty scary, the ES down 1.75 points.

"So it is written, so shall it be..."

Originally posted by jimkane

Now here's where it triggers nicely, gives the look like it may go down, then snap traps all the suckers and kills them. Watch and see...

Well, I'm not sure I have stated my premise. I was more stating I must have a premise. A premise for me sounds something like 'The market is going up here. I can see that because the price action is going from the bottom left to top right of the chart. I think it has some room to run because no major lines or other things confluence for awhile. I am looking for a predefined setup to get on board the trend. I found one here, an ABCD at a .382/.447 and key lines. I look at the confluences overhead if it were to take off, and I can get 3.0-3.5 reward/risk, so I'll take the trade.' To me, this sounds remarkably similar to roofer's premise. I need trend to push my setups, roofer's push comes from the characteristics of the predefined setups and how the algos operate.

The idea/point is that if you are trend following, you have to have some clue as to the trend. If you use predefined extreme to extreme setups, you likely could care less about trend. How can one trade the trend if you don't know what the trend is, or have no bias? I posit that you can't. I was trying to get people to see that if I can show, in advance, the direction the market is pushing, using only the tools, then perhaps that may be a learned skill, one that trend traders would want to learn. So, I post things that should help one to do chartwork and maybe be able to learn more about trend.

If you don't like trend, here's my suggestion: don't be a trend trader. Follow and use techniques more like roofer's, and forget mine. It's an easy choice. You like trend, develop something more aligned with my work. You don't like trend, develop and implement something like roofer's. Easy-peazy.

Originally posted by roofer

jim has stated his premise.... here's mine....the computers run the market.everything i know they know. i can not beat them at their game.. all i can do is hitch a ride by hitting predefined trade setups in the context of the geometry given by medianline sets.each trade has a target given by lines. if i can not see minimum 4/1 rr i pass on trade.

Here's a question to roofer, but I'd welcome comments from anyone. I've been pondering this for a long time, with no resolution. Any input that would help me sort this out, I'd welcome. I'm thinking about the .886 gap on 5 technique. Now, I can see how the algos can easily be programmed for gaps, .886's, no gap closure on the 5-min, and so on. Any high school programmer could do it in his/her sleep. But the geometry? How do they program for that?

Yes, like the Autoforks software, they could program for every last possible set off extremes, and if any hit, it's a go, but that's not what is happening in reality. The geometry that roofer so frequently shows, it's not always easy to find. Sometimes it is from a gap set, sometime a more obscure set, and so on. And notice that the examples roofer shows, they all have the really sweet look to them, something we can see by eye.

I can find many examples where the geometry is there (in a way an algo would find acceptable by a set of rules that would also produce the examples roofer shows), but our eyes would say immediately that the geometry isn't good. And these setups, by and large, do not get 'algoed'. How can that be? I simply can't believe all these algos are able to produce the same geometry as roofer, and be able to discern 'good' geometry from 'bad'. I just can't buy that.

So, there must be some other way they are doing it, some other explanation. They must have some other parameter(s) that will almost always only happen when we have our geometry. I'd like to understand this and figure it out. This way, I can use an algo and not have to draw any lines. If they can do it, I can do it. So, anyway, any explanation for how they can see the same geometry and be doing it another way, I'd love to hear some theories. It is going to take a lot to convince me it is being done with sets like we do, in a pre-programmed way, and it can filter itself this good. I just don't think they are that good. And if they are, I'm literally terrified...

jim- i agree that they are using something else... but i don't know what it is... but i think that what is alot of work to me... is nothing to the computers.. so i think they are ''that good''...there are so many good traders using different techniques to come to the same trades. for example, i have recently incorporated in my thinking the work of the best elliot analyst i have ever seen.my work gives me an edge over his followers who do not use their own work. at every important pivot in his work the last 4 months, there has been one of my predefined trade signals pointing the way.when his work has alternative counts, frequent in elliot, one pointing up, one down, some of his followers complain and whine when they get trapped on the wrong side... each time this has happened,,, there has been a stack of buy or sell signals pointing the way unknown to those who do not do their own work.

Jim, what do you mean when you say "our eyes would say the geometry is not good." How can geometry be bad?

I'm not convinced the algo's are using the same median lines as roofer all the time. On the QQQ's, I drew 2 other sets before I saw roofer's set and even after I drew another set. So all together there were 4 sets with lines going into his areas. The algos could easily have programmed in 2 of them without using the more minor swing low that roofer used.

How heavily do you think the algos rely on measured moves? It seems like I see a lot of median line sets that capture a measured move. I wonder if that is what the algos key on when it is looking like they key on a median line set?

I've been playing around with waves and finding abcds in the waves.
This is just practice work and could be totally wrong. I sure hope Roofer eventually shares what he is learning from the elliot wave trader he mentioned.

Here are some charts.

TNA 60 minute with abcd's in what may have been W2 and W4.



TNA 60 minute with waves colored pink and green. After looking at Spy and QQQ I am not as convinced that TNA is doing a wave B as I commented on the chart.



QQQ 30 minute waves with possible abcds. The gray median line set is a Schiff with anchor A point the low on 6/17.



I looked at Spy 30 and 60 minute but it didn't seem to offer anything additional so I didn't draw it up.

ches if you email me privately i will give you info.

I guess what I am saying is that not every time price hits a line, or lines confluence and price hits them, it's an area that produces a potential trading edge. With experience one can make an attempt at filtering out these areas that just don't have the right look to them. I'm not really talking about subtle, almost invisible variations here, I'm talking about 'obviously has potential', and 'obviously looks totally wrong'. This is frequently something that can be seen by eye real easily, but is very, very hard to program into a computer. I wish I had a chart example off hand for you, but I don't.

Originally posted by tx_ches

Jim, what do you mean when you say "our eyes would say the geometry is not good." How can geometry be bad?

I'm not convinced the algo's are using the same median lines as roofer all the time. On the QQQ's, I drew 2 other sets before I saw roofer's set and even after I drew another set. So all together there were 4 sets with lines going into his areas. The algos could easily have programmed in 2 of them without using the more minor swing low that roofer used.

How heavily do you think the algos rely on measured moves? It seems like I see a lot of median line sets that capture a measured move. I wonder if that is what the algos key on when it is looking like they key on a median line set?