MapReduce is one of the most important algorithm types in big data. Perhaps its biggest draw is the fact that it can be massively distributed across commodity hardware to crunch massive amounts of data. This puts super computing in the hands of those without specialized parallel hardware. In a very basic example for a Hadoop class I did, I implemented a basic MapReduce simulator in Mathematica in order to illustrate how MapReduce works. I also got some extra credit for it, as well.