I want to talk about piracy again, but I'm going to remove morality from the discussion because it just brings the language Nazis out of the woodwork and I think this discussion can actually work without morality being brought in for this viewpoint. The reason I want to bring this up now is because recent information has led to the possibility of an equilibrium being found. For the sake of simplicity, copy protection is not included as a factor for this simple model.
An evolutionary stable strategy (ESS) is defined as a strategy which, if most members of a population adopt it, cannot be bettered by an alternative strategy. (Dawkins, 1976) It is not necessary for all of a population to adopt a strategy for an ESS to emerge. In fact, it is possible to find a blended ESS which depends on percentages of the population adopting complementary or opposite behaviors.
In a software ecosystem, we have three behavior patterns: producer, buyer and liar. Producers create the software, buyers purchase the software, and liars pirate the software.
For producers, there is a cost associated with the creation of the software and a cost associated with handling support after the fact. In order to maintain an ESS, a producer has to have its initial release bring in enough resources to cover the cost of the initial release, the cost of supporting the release and the cost of the next release at a minimum. Any less and it risks dying out.
For buyers, there is a cost associated with the acquisition of the software and a gradually smaller support cost per buyer for the producers. (The reason that the cost is gradually smaller is that as patches and support documentation grow, the cost to the producer to support the release shrinks.)
For liars, there is next to no cost associated with the acquisition of the software and a much more stable support cost per liar for the producers. (The reason that the cost is much more stable is that liars cannot get the patches.)
As long as a producer has enough buyers to support his release and the support costs for the liars and the buyers, a producer can continue to produce. Experience and evidence released by other companies show that on average, 80% of technical support contacts are from people who have purchased the software, 15% are from people who may purchase the software, and 5% are from liars (people who have not purchased the software). While we cannot quantify the number of people who have not purchased the software, we can quantify their expense on support.
Regional experiments with game pricing has shown that for most games, the initial cost does not affect the number purchased, just the curve at which they are purchased. In other words, if a game will sell 300,000 units, it will sell them quickly at first with rapid drop down to a low level at $50 or at a slow steady pace at $20, but it will still only sell 300,000 units. So cost for games is not really one of the factors that can easily be adjusted.
The two factors that a producer can adjust are quality and quantity on a sliding scale based on the number of buyers in the market. A very short, high-quality game and a lengthy bug-laden game can cost about the same to make. A producer is trying to strike the balance between the duration of the game and its quality. In practice, the "money spot" is less than one support contact per thirty copies sold.
Now as the number of liars in a marketplace increase, the amount of resources that a producer can devote to the production of their next release decreases due to the increased support cost of the liar's market share.
At least at the moment, we have achieved a bit of equilibrium in the PC market. Five years ago, that equilibrium was at about $2 million to produce a product with about 350,000 buyers. Nowadays, the balance point has shifted to $750,000 to $1 million with 250,000 to 300,000 buyers. Fixed costs per unit have risen even though the prices that we sell games at has not. Pirated goods are more available on the PC due to the proliferation of broadband and the number of liars willing to pirate has increased in part due to the extremism of the anti-piracy lobbies of the RIAA and MPAA. Attempts to continually monetize games (MMO's, ad revenue) have had mixed results with most MMO's not being profitable and ad revenue averaging under $2 per unit sold.
Essentially, our equilibrium is smaller products like PopCap and Stardock titles. "Galactic Civilizations II" cost $1 million and sold approximately 300,000 units according to a Stardock spokesman. While there will be some big budget bets, they will be fewer and further between.
Even digital distribution isn't a godsend because rather than the cost of the unit and the retailer margin, we have the chargeback margin and the ongoing cost for bandwidth. Digital distribution may grow your marketplace, but people are still (rightfully) skittish about using their credit cards online, European debit cards won't always work on American networks, and to be fair, it further reduces the opportunity cost necessary for one to become a liar. In short, digital distribution is a wash.
The current generation of consoles is still trying to hit an equilibrium. The biggest advantage that consoles have is that there is a higher barrier of entry for a liar to take from a producer on a console (the cost of modifying the console) and there is a risk associated with modding because the act of modding could result in damaging or destroying the console. As a result, liars don't really become a problem for consoles until a soft mod is released.
So to sum up, the marketplace is evolving to handle the larger number of pirates by reducing the amount spent per title. If the number of liars continue to increase while the number of buyers continues its slow decline, it is very likely that the PC market will return to the days of hobbyist developers being in charge (in which cases systems like Microsoft's XNA initiative will seem more prophetic than anything), but unless we can come to some middle ground between the anti-piracy and copyright-abolition agendas, we will continue to see fewer and fewer high-profile bets on the PC.