[lshubmcw] Chess endgames via MapReduce

Chess endgames could be straightforwardly solved with iterations of MapReduce, by the traditional algorithm:  On each iteration, start with the positions whose value are known.  The Map step does a retrograde step generating all previous positions.  The "magic" happens in the Reduce step, where identical "previous" positions become collected together.  From the Reduce point of view, we analyze if there is enough information from successor position values to calculate a new known value.

Obviously, for large endgame tables, this will take a lot of computing power and disk storage, but those are easily available in commodity MapReduce services such as Hadoop.