I think the bias is in the way the selections are made.
If I remember correctly, 5 or 6 past players (I think John Eales is one) sit down and watch all of the games played that year and pick the top 3 players.
That means that games between Italy and Scotland are given the same importance as games between Australia and New Zealand.
And players in the top teams lose votes to their team mates, whereas good players in average teams stand out.
Its a flawed system, but when trying to make it quantative, its the best I can think of.