> […] They also buy a lot of expensive games too, […]
That is the interesting point: gamers aren't divided between full pirates and full legit. Many people are both. In the case of the pirates who surround you, I notice that a significant part of their budget is devoted to the legal game market. So, if they suddenly couldn't pirate any more, would they buy more? To what extent?
Anyway, the fact that people aren't either full pirates or full legit can't hide the elephant out in the room: pirating behaviour is way more effective at downloading than legit behaviour is at buying. So 20% of piracy behaviour accounting to 80% pirated game doesn't sound odd: it just mean that when legit behaviour buys you one game, pirating behaviour downloads you 16.
Also, a gamer which is 80% legit and 20% pirate will download 4 games for each one he buys. And if you turn his pirating behaviour into legit behaviour, he won't give 400% more, but 25%. Significant, but not the catastrophe music majors would like us to believe.
Finally, 20% of pirating behaviour doesn't necessarily mean a 20% decrease in legit behaviour (and 20% revenue loss). It could just be a 25% increase of gaming behaviour (and 0% revenue loss). This is not all. Sometimes, games are pirated prior to purchase, to try them. Sometimes, players buy the game when they are finished playing them, because they liked it. And sometimes, not pirating the game would just mean not playing it at all.
I won't deny that piracy doesn't cause revenue loss. But we can't deny that the loss is much, much lower than the rate of pirated games. I bet that like in the music industry, the biggest costs are middle men. And the internet makes them less and less useful.