The method I developed is based on the analysis of pure price, instead of information distorted by technical indicators. In fact, if the basic postulate of the technical analysis on repeatability of history is true, there will be no necessity to study behavior of indicators to answer the question about the further movement of the price. It will be enough to find such dynamics of the prices in the past, which would be similar to recent price dynamics investigated, for implement the best ever technical analysis. I use algorithms, based on fuzzy logic and statistical methods together with artificial neural networks to search for price movements similar to present one. Then I divide the area of future possible price movement into regions and calculate the quantity of similar price movements that corresponds to that region. More of past movements pass to that regions, more the probability will be that the future price will reach that region.