Here's a diagram that helps me to see Intra Bar Gaussians. This is a simplified view. Actual volume might be different. The idea is to be alert whenever the close is hesitant.
I have been restricting myself to studying only ES p&v and am posting this question because I think it's relavent to this month's laterals discussion. If not, please feel free to ignore. I have read and reread this parapraph, and continue to be stumped by what it's saying. If the blue arrows are wrong, I'm having a hard time seeing how the decreasing black arrow can be right. If one splits the bar labeled "this bar", then what you end up with is a red bar lower than the prior red bar (first sign of change on a bar to bar basis), then 3 increasing black bars, ending in a bar that is higher than the gaussian peak ( the peak at the "this bar", which is actually lower because it was split). What I see overall (starting on the increasing red just prior to the "this bar") is increasing red to increasing black, with a little B2B embedded in there. Furthermore, the comment in the above quote about the last red bar being a flaw is even more baffling, as the point of the quoted paragraph seems to be that no change has yet taken place. But, I thought that flaws only occurred in the dominant direction. So if that last red bar is indeed a flaw, then by definition, change has already occurred. In summary, it seems like the quoted paragraph is arguing that change has not taken place, but the best I can read the gaussian is that change has taken place. And, I am left confused by the suggestion that the last red bar is a flaw when the quote suggests that no change has occurred.
it argues that " how easy it is to find yourself several levels down into resolution levels when your intent was to remain at the Forest Level."
I think you have to make a step back from the chart, and "see", and draw what it makes sense. On such a long, high volume bar 1-2 ticks can't make a huge difference, isn't it so? You have to get the correct feeling of what's happening, to extract the information from noise (there'll always be noise). <img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=1615952>
CNMS2, this is precisely my point. I created a 12 minute video which outlines how I'm currently looking at these situations and a sentence by sentence review of the quote from Spyder. A number of you have requested I do a volume/gaussina video, but currently I have too many unanswered questions to do justice to it. I hope that this little question video along with a few more can clear up these holes in my knowledge and then I will do the video. http://webpages.charter.net/coolmoss77/question.html
My point is that all these rules have to be applied in their intended spirit. They create a framework. You have to extract from the price action the meaningful information, by filtering out the noise. And there's noise on all axis: amplitude, time, phase, etc.. When you monitor on the 5 minute OHLC chart, but trade on a slower fractal (level 1, 2, ... forest), your brain has to smooth / filter out the less meaningful info. When you monitor the same 5 minute OHLC chart and trade at tape level, you have to be aware of what's happening intrabar if you want to make the most timely decisions. The compromise is what Jack called the Tampa tape: to adjust your monitoring fractal in function of the current market pace so that your traverses are tapes. As the market pace changes, your fractal will change. You can finesse your entry on a faster fractal than the one you trade, but then you step back (up) to the proper fractal for the current market pace, where you see the Tampa tape. These kind of adjustments make almost impossible to automate efficiently the whole method, imho.
I think that you have to define "late" for your trading fractal, not for your monitoring fractal. <img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=1616089>
Unusually high for a retrace, the red volume announces a change of trend on a higher fractal (dailies). <img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=1616100>