A colleague and fellow Dallas IGDA board member, Myke Sanders, recently forwarded me a data analysis he did on the habits of players who purchase gold from gold farmers. It's a really fascinating bit of data, one that may give developers some revised perspective on how to combat the gold farming issue while continuing to avoid micro-transactions, implementing real money auctions, and/or monetizing the currency. While this isn't an overly legal issue, I still wanted to take an opportunity to comment on the findings. The full piece is available here (PDF).
Myke notes two specific spikes on each graph: There were spikes at single and highly repetitive (12) transactions within the 90 day window of the study. There was also a large spike of people making high value transactions. I would be interested to see a correlated data set of these two, but based on Myke's analysis, there were basically two groups of people: those buying a huge amount of gold once in 90 days, and those who bought almost weekly but for a much smaller sum. He reasons that the first group might be making purchases for a single, say, epic mount or piece of equipment while others might be buying to supply a weekly raid. He goes on to theorize that a loan system might satisfy the former while non-transferable buffs that are a better alternative to disposable equipment might satisfy the latter.
In general, I would tend to agree. It's almost surprising there hasn't been a 'Bank of WarCraft' to date, though I would have to say that it would need to be operated by Blizzard rather than by players (i.e. virtual loan sharks). Of course, Blizzard may need to employ an economist to at least design the system such that we don't end up with an financial meltdown in Azeroth. It would be very easy to over saturate the market with a virtual currency, which would lead to inflation and thereby throw off the balance of the game between the bank and random drops. It would also be interesting if you could actually earn interest on gold deposits. Another alternative might be a virtual credit market, but that gets even more complex. I guess the real questions is whether players want their virtual world to mimic the real world more or not.
The solution for the second group tends to be a little more complex, as it would take likely a pretty good revision to the overall mechanics of the game to get the kinds of buffs being proposed, or at least a major change to the game as it's currently available. Granted, this is not impossible, but I imagine the sudden change in strategy might not go over well with many players (a la the New Game Enhancements in Star Wars Galaxies) and may be better suited as an approach for developers of future games.
As a side note, it will be interesting to see if developers take advantage of the wide latitude granted by the Glider decision in order to go after gold farmers in a similar manner. The recent DMCA applications won't apply, but it is a similar inducement to break the EULA/TOS of most games.
I certainly hope that this will be the first of many studies into actual buying habits of those who utilize the services of gold farmers and even virtual commodity salespeople. While there has been plenty of anecdotal evidence over the years, only actual data like this will provide enough information to actually combat the problem, if your game is designed to avoid such outside monetary influences, by changing certain game design elements to be more in line with the way people want to play. Similarly, if you're looking to start a game that encourages monetization of one sort or another, this is likely a good source of information for your design decisions as well.