Hi Sharks Like you, I don’t use any indicators for discretionary...

Hi Sharks
Like you, I don’t use any indicators for discretionary trades (algo's are a different story).
I do apply heuristics & draw in key levels (Support & Resistance).
I'm also a statistical trader and measure trade probability using in and out of sample data+actual trade results.
I only enter trades presenting a probability above X (80%) based on stat models (in Reality 76% WL & around .4RR).
Predicting price (Turning points and targets) I'm fairly accurate. If I throw time into the mix the accuracy stats blow out.
I measure price action in waves & my accuracy measure is based on hitting a price target within the current wave.
Looking at my charts above in this thread, the slope of a typical wave is reasonably consistant/predictable so with this I see limited value in attempting to predicting time.
Volume spread analysis (VSA) also suggests that price+time prediction is dependant on the strength of the players on either side of the bid. I haven't found a way to predict the near future presence of bull/bear power and based on this have concluded attempting such prediction is futile (It may be that my quant is lacking).

I don't understand your comment "Time-wise all I do when I validating my setups is generate the stats for each of a selction of timeframes following entry, i.e. 1 bar, 2 bars, 3 bars, etc, and the stats end up telling me the optimal timeframe for each setup."
Are you saying you enter a trade then determine the optimal TF to manage the trade?