Disliked{quote} K's recurrence indi in the attachments. Also note the ~4X figure may also be seen in the stats as: Freq of possible problem bars / Freq of true problem barsIgnored
Having the following two probabilities:
- P(t_r): probability that a potential transient bar (left transient) will become recurrent within the next h bars.
- P(r_f): probability that a left recurrent bar will become fully recurrent within the next h bars.
The following seems to be one of the most important implicit assumptions:
P(t_r) > P(r_f)
This can be verified empirically by comparing the following values, which can be taken as approximations to both probabilities:
- percentage of potential transient bars becoming recurrent (already part of the indicator)
- percentage of left recurrent bars becoming fully recurrent (has to be added to the indicator)