Quote Originally Posted by Phaedrus View Post
Yes, it is a bit unsettling that some "weak" signals appear to be (and in fact are) better than some "strong" signals. Nevertheless, over the long term all doubt as to their relative merit is removed. Backtesting the Ichimoku system on the SP500 over 10 years or so tells us that following ALL signals would lead to a 25% loss, but by acting on Strong signals ONLY, a 20% gain results.
Phaedrus,it seems then that you now have the answer to your original question:
"I have seen it stated that "Eliminating the Chikou Span and ignoring weak crossovers enhances the traditional Ichimoku method’s accuracy." This would mean that signals such as that at (2) would be ignored - a pity, yes? " In other words, "No!" That is, we can tolerate the seemingly lost opportunity of an occasional gain from a "weak" signal in favour of a more likely, but smaller, long term gain from using only the "strong" ones.

Then we might wonder about the algorithm behind the original chart post which is effectively classifying the SignalStrength into seven categories, because now you have proven that only the "strongest" category is useful.
In the long term. (the komu descends again ..)

"Ph: ... with a lot of bath water thrown out, I guess it is inevitable that a few babies go with it! ..."

LOL. i'm now just waiting for a chance to throw this into a conversation somewhere! Cheers