However, the outcome of research and development activities is not
completely haphazard. Arrow (1962) argues that the initial a priori probability
distribution of the true “state of nature” (e.g. the firm’s understanding of the
technological principles or the feasibility of combinations of capabilities) is
relatively flat. With each successive research step, the a posteriori distribution
becomes more defined. So, researchers make the best possible choice, given the
capabilities and action options at their disposal, and given the information about
the future and the consequences of possible actions