One of the most remarkable thing to me is Nash's proof of the equilibrium.

The whole thesis is about 27 pages and the proof itself is 7/8 pages (double spaced) and use basic linear algebra. Even though it extends a previous result from Von Neumann (where N=2) significantly.