How many sales are lost to piracy?
Most people discussing this have to realize that not every download is a lost sale. A few pirates will be people who buy the game and then also pirate it. Some are people who are too broke to afford the game. Some are people who don't really want the game, they're just obsessive hunter-gatherers caught in an environment immune to scarcity.
So how many sales are lost, then? What if it turned out that I was totally wrong and that it is possible to enact perfect DRM on an open platform? How many of the 90% would buy if they simply couldn't pirate? This is very tough to measure. The Ricochet story I linked above talked about the publisher adding some DRM and finding that they garnered an additional sale for every 1,000 pirates they stopped. I can't distill their numbers without spending a thousand words on warnings and qualifiers, but in the end we have no way of knowing how well these results would translate to other titles.
If the 1 in 1,000 number was at all indicative of the rest of the industry, then it would be almost worthless to fight pirates at all. "Perfect DRM" would only result in an additional 0.9% increase in sales. Less than a single percent.
What we would need to really judge this figure would be some data from the big-name publishers.
How much does DRM help stop piracy?
Well, "none" if you're talking about widespread torrent-based internet piracy, anyway. We don't even need to cite a number for that one. All we need is to think about it. The process of pirating Galactic Civilizations II (DRM free) and Spore (extra-zesty DRM) is exactly the same: You find it and download it. How can the DRM impact the rate of piracy when the DRM doesn't exist for the pirates? (Unless we're talking about DRM adding to piracy by causing people to pirate out of protest, which is another whole can of worms.)
The only possible impact DRM can have on net piracy is to (maybe?) slow down the cracker and delay the game from appearing on the torrents. But that number seems to have more to do with how famous the title is. Spore was available to pirates before release day, so it's pretty hard to imagine the DRM ever did anyone any good. If the game was released DRM free and appeared on the torrents (say) a day earlier, would that have changed the sales numbers?
But the kind of piracy it can stop is casual piracy. Simple friend-to-friend sharing of discs and installs can be thwarted by a simple disc check. Most gamers don't have the skills to edit an executable and bypass even something as simple as that, much less tangle with SecuROM.
But it's important to note that the 90% figures we see above come from both DRM and DRM-free games.
How sick is everyone of this entire piracy argument merry-go-round?
How many pirates are jerks?