The Reversal System - Engineering Background
....
So when the market is trending, it's fine. But it IS when the market TURNS, that traders make a killing of their money, but you will NOT get the correct timing from Armstrong/Socrates.
....
That sums it up nicely. So what is the reason for this from the technical / engineering perspective?
What constitutes the Reversal System? Note: The following does NOT cover the fraudulent hindsight-based
Revision Signals which are mislabeled Reversals, and as such they are not Reversals in the sense of the following.
1) Extreme detection is applied to find the time and value of the previous high / low in a time series.
2) To the result of the extreme detection, a simple segmentation algorithm is applied to split the time between the extremes into segments. At the borders of these segments, the reversal values are looked up.
That's it. It nicely even explains how double and triple reversals are generated.
Let's go into this with some more detail:
Imagine the reversals are generated by slicing the time going backwards from the generating extreme to the previous opposite extreme into time slots. This is the segmentation algorithm. There might be some kind of projection involved which is needed for reversal values that the market has not yet visited, but nothing fancy.
The reversal values are somewhere near where the price line crosses the four vertical boundary lines of these time slots. In some cases, depending on the shape of the price curve going backwards to the previous extreme, there are multiple reversals with same or near same values, which are then the double, triple reversals.
The reversals are near previous / projected resistance and support values naturally because that is where the price movement is flatter, and where the vertical time lines cross the price line more easily / more often. It does make sense. Doesn't it?
These reversals work, if the trend is clear, price wise going back into the direction of the previous extreme or some kind of projection. They are elected one after the other, just like following your nose.
If however, the market does NOT move in this simple fashion, but the market is in some kind of consolidation period, then the reversals are elected in a chaotic fashion.
Guess what: You buy the high and sell the low!
That is because the reversals are near resistance and support, and you are getting whip-lashed in the trend channel between them with no end in sight.
It is this situation where we need the ambiguous human interpretation of the forecast arrays to provide excuses for the failures. Here, recently MA produces user feedback about them trading AGAINST the failing reversals, praising them for doing so in his public blog. See
Trading Against the Reversals.
Also, in such movements, failing weekly reversals are effectively degraded to daily reversals or any time period by providing another excuse as described here:
Reversals & Timing. This really destroys any confidence in the reversals because it is a complete mess, just for the purpose of ambiguity and perfect deniability.
The Socrates system does not have any means to detect whether the market is in a mode where the reversal system works or not (trending or turning). Therefore, over time, using this system, a trading account will be depleted, basically because markets are moving sideways longer than they are declining / rising as nicely as the reversal system wants them to. Or, when the market is turning and there are no signals indicating the turn.
Now let's combine this knowledge with actual observations and an experimental implementation.
I am re-publishing here the work of
decipher11.
Since MA is trolling this blog to try to bury any information demonstrating that Socrates is a fraud and the product of his megalomania !
This is how reversals are created - far from quantum physic as stated by Martin Armstrong - since just human intelligence can generate them quite easily
Every time the market exceeds the previous tick's high or low, a set of reversals is generated. If that new high or low is not exceeded for the next tick, that reversal stays.
There are 4 prices in a set of reversals.
These are determined from the timings 3, 5, 13, 26. (for stocks he seems to use 4, 8, 16, 23, however for the Dow he seems to be using 3, 5, 13, 26).
So if a new high is made, the low from the previous 3rd tick is a bearish reversal, the low from the previous 5th tick is another bearish reversal, the low from the previous 13th tick is another bearish reversal, and the low from the previous 26th tick is another bearish reversal.
The same is true if a new low is made, in this case the previous nth high is used.
If there is a holiday and the markets were closed, that still counts as a day.
You can generate reversals yourself at all levels of time.
------------------------------------------------------------------------------------------
MA constantly says that the plebe (useful idiot) doesn't connect the dot as his system does, meaning no market move alone, everything is correlated.
Which is a cleaver observation.
BUT using Socrates you will never find any correlation, all markets are "analyzed" separately from each other .... not dot connection, no correlation
Socrates is everything except Artificial Intelligence !
For those interested a pinescript to use in tradingview calculating reversals multi-complex system
written in 10 min, can be improved with elected reversal not being displayed anymore
//@version=2
//decipher11
study("[RS]Reversal", overlay=true)
Mod = input(true, title="box checked : 3-5-13-23 Reversal (Commodities and Dow) - unchecked 4,8,16,26 Reversals (Ind Stocks)")
r1 = Mod - 3 : 4
r2 = Mod - 5 : 8
r3 = Mod - 13 : 16
r4 = Mod - 23 : 26
topl1 = valuewhen ( high >= highest(high,13), low[r1],0)
toph1 = valuewhen ( low <= lowest(low,13), high[r1], 0)
topl12 = valuewhen (high >=highest(high,13), low[r2], 0)
toph12 = valuewhen (low <= lowest(low,13), high[r2], 0)
topl13 = valuewhen (high >=highest(high,13), low[r3], 0)
toph13 = valuewhen (low <= lowest(low,13), high[r3], 0)
topl14 = valuewhen (high >=highest(high,13), low[r4], 0)
toph14 = valuewhen (low <= lowest(low,13), high[r4], 0)
tl1 = plot(topl1, color =topl1 != topl1[100] - na : red, linewidth=2, style=linebr, offset=0)
th1 = plot(toph1, color =toph1 != toph1[100] - na : green, linewidth=2, style=line, offset=0)
tl12 = plot(topl12, color =topl12 != topl12[1] - na : red, linewidth=2, style=linebr, offset=0)
th12 = plot(toph12, color =toph12 != toph12[1] - na : green, linewidth=2, style=line, offset=0)
tl13 = plot(topl13, color =topl13 != topl13[1] - na : red, linewidth=2, style=linebr, offset=0)
th13 = plot(toph13, color =toph13 != toph13[1] - na : green, linewidth=2, style=line, offset=0)
tl14 = plot(topl14, color =topl14 != topl14[1] - na : red, linewidth=2, style=linebr, offset=0)
th14 = plot(toph14, color =toph14 != toph14[1] - na : green, linewidth=2, style=line, offset=0)
//
hey Marty, can you imagine, if I had an hour to lose I would have been able to rewrite SOCRAPS !
is a charlatan, and he spent 11 years in jail for that reason but he has not changed.
Read this blog
starting here to find out more about computerized fraud.
See
armstrongecmscam.blogspot.com for a more compact view of major findings posted in this blog.
