Thank you again for your detailed reply. I have come to similar...

Thank you again for your detailed reply. I have come to similar observations but was not able to quantify them in terms of probabilities. I have found that when quantifying certain well established reversal or continuation patterns by the means of backtesting in particular candle stick patterns they seem to be 50 / 50 at best. My programming skills are not yet up to the task of backtesting divergences but was therefore wondering if you were able to quantify your observations.