You can get a bit of this by consciously separating solutions from fully understanding the problem. Once you've started developing solutions, you become attached to them, you start reasoning using those solutions and their properties and tradeoffs as anchors, and it's harder to abandon them.<p><a href="http://lesswrong.com/lw/ka/hold_off_on_proposing_solutions/" rel="nofollow">http://lesswrong.com/lw/ka/hold_off_on_proposing_solutions/</a>