We present a simple model of spatial evolution that avoids several problems that arise with more complex networks of players. We consider a world where pairs of players are matched forever. These players learn from the whole population but they are more likely to learn to strategies used by...