There is, however, a special case in which this probability may be indirectly a reason for not resolving to do what would otherwise be best; i.e. where this resolution would only be right if followed by subsequent resolutions. The problem thus presented is considered later.