The statistical relevance model of scientific explanation was proposed by Wesley Salmon in 1971 as an interesting alternative to already existed models introduced by Hempel and supported by many other philosophers of science. The most important difference between the nomological models and statistical relevance model is that the latter tries not to use the very dubious term of “law of nature”. The first part of the paper consists of the overview of the Salmon’s model and of the main arguments which were raised by various authors against it. In the main part of the text all of those arguments which were meant to undermine the model are presented on an example taken from the economic practice. It is very popular among the economists and especially among valuation experts the so called “statistical analysis of the market”. The main objective of the analysis is to discover all of the factors which influence the market value of the particular product, in other words to explain the market value of the product. The example was taken from the social science (economics) for purpose as one of the thesis in the paper is that, the SR model can work quite well in physics or chemistry, but it is dubious whether we can really deploy it in sciences which try to describe and explain the various phenomena of human activity and behavior. The final conclusions are: The practical deployment of the model in social sciences are problematic, as it is too idealistic and therefore it doesn’t work properly. Against its initial presumption the model doesn’t avoid the problem of laws of nature. Although the law of nature is not a required element of the explanans, it comes back at the stage of proposing the initial candidates for the relevant variables. The hypothesis on, which variables can be and which cannot be relevant to the explained phenomenon are constructed mostly according to the intuitively understood causal relationship founded on laws of nature. The important postulate of homogenous partition is in practice unachievable what causes that the explanation is bound with the enormous risk of a mistake. The risk is quantifiable and can be estimated, but the estimation is depended upon experience and intuition of a researcher.