Dr Faustus's Account Talk

[DrFaustus]

Just to clarify ... when I say " 1 hit," that means that pattern has occured exactly once since Jan 16, 2007. Statistics presented are compounded for all hits on that particular pattern ... they are NOT an average rate of return, nor are they an expected return for that day.

Hope that helps clear things up.

Dan

 

[BeaverState]

I'm sticking with the I-fund on those days. Whichever fund has done the best over the past 12 months usually does the best or near the best over the next 12 months.

 

[qibovin]

Just to clarify ... when I say " 1 hit," that means that pattern has occured exactly once since Jan 16, 2007. Statistics presented are compounded for all hits on that particular pattern ... they are NOT an average rate of return, nor are they an expected return for that day.

Hope that helps clear things up.

Dan

That's what I figured. 1 data point--hard to draw any conclusions with any accuracy, but worth noting in relation to other data. (and after this week we'll have another data point!) Thanks.

 

[DrFaustus]

Well shoot. It would have been nice if the S-fund had continued climbing into positive territory yesterday but at least I can console myself by the fact that it did better than the I-fund.

For Wednesday, we have double pattern 2.4. This pattern has occured 1 time since Jan 16, 07 for a compounded total of +0.00, -0.09, -0.36, -0.44, +0.25 for G,F,C,S and I funds respectively. Now this isn't a lot to hang my hat on - but what the heck ..

I'm gonna be 100% in the I-fund for Wednesday.

 

[DrFaustus]

Man, am I sucking pond water! Maybe I should ditch this two day pattern thing ... though I think it might have some benefit once we get more data points. Any thoughts?

 

[FogSailing]

Not every day is going to be spot on...IMO, it's an idea that looks to have some legs..I hope you keep it going a while longer...gather more data. If it pans out, it's always nice to have another weapon in the TSP arsenal..

Thanks for your efforts..

FS

 

[BeaverState]

Man, am I sucking pond water! Maybe I should ditch this two day pattern thing ... though I think it might have some benefit once we get more data points. Any thoughts?

Aren't you in effect simply making more patterns to track? Perhaps the new patterns aren't as predictive as the original eight. I don't know why Ebbs patterns work so well but I think it may have to do with greed and fear, and how humans react to it. That would explain why the patterns work no matter what the current news is.

 

[qibovin]

Aren't you in effect simply making more patterns to track? Perhaps the new patterns aren't as predictive as the original eight. I don't know why Ebbs patterns work so well but I think it may have to do with greed and fear, and how humans react to it. That would explain why the patterns work no matter what the current news is.

Or yesterday's pattern. But I agree that more data points might start to demonstrate any patterns (if they exist.) However, Ebb's current tracker is just getting to the point of being statistically significant enough to make accurate predictions, and he's still tweaking it with the contrarian ideas and using the S fund. Since he has 8 patterns and you are looking at 8*8=64 patterns, you may need to collect about 8-10 times more data before you get the discrimination for which you're searching. That'll put us somewhere in 2011 at which point you might be able to better display the new Ebbtracker^2 in holographic 3-D.:nuts:

 

[DrFaustus]

... But I agree that more data points might start to demonstrate any patterns (if they exist.) However, Ebb's current tracker is just getting to the point of being statistically significant enough to make accurate predictions, and he's still tweaking it with the contrarian ideas and using the S fund.

This has been my point exactly ... I have spotted some two-day patterns which are indicating results contrary to what the EbbChart would indicate, particularly on those patterns which are kind of wishy-washy one way or the other.

... Since he has 8 patterns and you are looking at 8*8=64 patterns, you may need to collect about 8-10 times more data before you get the discrimination for which you're searching. That'll put us somewhere in 2011 ...

Unless I can access to Ebb's database for patterns prior to Jan 16, 2007.

I'll think I'll go back to what I was doing before ... and will continue to update the two-day pattern database until I get some meaningful stats. As I see something which causes me to go in a contrary mode to what the Ebbchart is predicting, I'll be sure to post it here. Trying to guess where the market is going with these two-day patterns that have just one or two hits is insane.

So I'll be making an IFT to the F-fund for tomorrow (Thurs) and we'll see where we go from there.

 

[DrFaustus]

Aren't you in effect simply making more patterns to track? Perhaps the new patterns aren't as predictive as the original eight. I don't know why Ebbs patterns work so well ...

Yeah, I now have 64 patterns to track instead of 8. Its do-able though, and is not too labor intensive.

Ebb's patterns work because they are based upon the EbbTracker, which has (I gather) gathered a lot of historical data from various sources and then he performed some kind of pattern analysis on that.

Since Ebb is planning to take this to a premium service (I think that is still what he is intending to do), its doubful we'll get more insight into how the various patterns are generated. Sure would be interesting to review though!

 

[fabijo]

This new tracker for the ebbtracker will need a name. Is it the ebbtracker tracker? What happens when we start looking at two day patterns on the ebbtracker tracker? This reminds me of calculus class, where we just keep going from first derivative to second derivative to third derivative...

 

[ssafed]

........or Dr. Seuss' Bee Watcher Watcher!

 

[fabijo]

Fox in Socks

 

[DrFaustus]

Well if it bears following, it could be called EbbChart^2

 

[DrFaustus]

From the EbbChart, pattern 4 is on tap for Friday. We're coming from pattern 6, making this combo pattern 6.4

There have been 23 hits on pattern 4 since inception of the EbbChart and, as you can see, it has an excellent overall rate of return: +0.17, -0.44 +1.67, -0.38, +7.48 for the G,F,C,S and I funds respectively.

Combo pattern 6.4 only has 1 hit thus far, with the following rate of return: +0.00, -0.35, +1.45, +1.89, +1.84 for the G,F,C,S and I funds respectively.

This doesn't give me any reason to go contrary to Ebb, so I will be 100% in the I fund for Friday.

 

[qibovin]

Do you have yet or will you be posting any stats on combo pattern 4.5 for Monday?

 

[crhuss]

I'm good for Friday in the I, but ebb's tread talks about bleeding out on Monday and should we be in the G/F fund on Monday. What does your two day pattern suggest?

 

[qibovin]

I'm good for Friday in the I, but ebb's tread talks about bleeding out on Monday and should we be in the G/F fund on Monday. What does your two day pattern suggest?

O contrere.
Reread the thread and the system comments. The "bleeding out" refers solely to the three reds, but the statistics suggest otherwise, so as a result, Ebb has been playing this contrarian and will actually be going to the S fund which statistically has the highest percentage positive and highest total gain with pattern 5.
That said, that doesn't mean there won't be bloodshed on Monday. Still, in the bigger picture, IF there is (and it's a big IF), "'tis only a flesh wound" in the big scheme of things.
I do agree, though, that any light that the good doctor's two-day combo pattern analysis might shed on this controversial pattern would be welcomed.

 

[DrFaustus]

Well, we've had two hits on pattern 4.5 since Jan 17, 07. The stats for that are: +0.08, +0.18, -0.91, -1.18, -0.71 for the G,F,C,S and I respectively.

We've had 16 hits on pattern 5 since Jan 16, 07 and the stats for that pattern are +0.25, -0.45, +0.81, +1.55, +0.20 for the G,F,C,S and I respectively.

So I'm not sure where I stand for Monday. Pattern 5 is not a strong pattern for S ( only +1.55% over 16 hits ), whereas 4.5 shows a possibility for significant downside. At this point, I am going to wait until Friday to see what other news is in the wind before making a call.

 

[qibovin]

Well, we've had two hits on pattern 4.5 since Jan 17, 07. The stats for that are: +0.08, +0.18, -0.91, -1.18, -0.71 for the G,F,C,S and I respectively.

We've had 16 hits on pattern 5 since Jan 16, 07 and the stats for that pattern are +0.25, -0.45, +0.81, +1.55, +0.20 for the G,F,C,S and I respectively.

So I'm not sure where I stand for Monday. Pattern 5 is not a strong pattern for S ( only +1.55% over 16 hits ), whereas 4.5 shows a possibility for significant downside. At this point, I am going to wait until Friday to see what other news is in the wind before making a call.

More Ebbstrapolation:

There are 4 instances since 16 JAN 07 that pattern 5 turned up on Monday with the following average results (% gains summed and divided by 4):
-0.09% C, -0.16% S, +0.17% I
Here, the I funds stands out as being the only equity with a positive average, and it was positive 3/4 instances. This is consistent with the overall observation that the I fund tends to do well on Monday. Incidently, this average gain for the I fund is almost exactly the same as the overall pattern 5 average gain for the I fund (since 16 JAN 07) of +0.18%. (C & S cannot say the same.)
Before you get too excited, however, there was one instance in which combo pattern 4.5 occured on a Monday, and it went horribly bad, actually contributing primarily to the overall bad showing of the 4.5 combo pattern noted by Dr. Faustus above.
So where does that leave us for tomorrow's move for Monday? Well, if you take a straight up arithmatic mean of the per-day gains and percent of positive occurances for 1) All occurances of pattern 5, 2) All pattern 5's on Monday, and 3) All combo patterns 4.5 (effectively weighting this average toward that one really ugly day (What else was going on April 30th?) here's what you get:
C fund positive <25% of the time with average gain of -0.11%
S fund positive ~50% of the time with average gain of -0.16%
I fund positive ~64% of the time with an average gain that is basically flat
Given this, I'm not sure that Ebb's move to the S fund for Monday is the wisest. I may put a little in the I hoping for the Monday effect. Bottom line is, just as this 3 red pattern was designed to avoid dips, this pattern has seen some big dips occassionally, but on average shows gains. Ask yourself,
"Are you feeling lucky"