A Grid Computing Economy 3 – Online Gaming

Part 1 described the mechanics of an economy based on the unused processing power of desktop and laptop computers on the internet. Part 2 gave a detailed example of a digital media store like iTunes that could use this spare power to offer free music while still compensating the artists fairly. In this part, we'll do a comparison of some popular free Massively-Multiplayer Online (MMO) games between the revenue they earn and how and the revenue they could earn through the spare computer power of their users.

A lot of assumptions are going to be made for these comparisons but they will be marked as such. If any of these assumptions seem unfair or inaccurate, feel free to leave a comment and I'll come up with a better estimate. The most central assumption is that a desktop computer has 10 cents per hour of spare processing power when it is in normal use. This is based on the one dollar per computer per hour that Sun Microsystems charges for their grid .

Consider two games that offer some sort of permanent basic play for free, Maple Story and Runescape. These games offer free basic play and generate their revenue by offering premium service. Maple Story (MS) sells in-game items for aesthetic and practical purposes, and Runescape (Ru) unlocks large portions of the game for a monthly fee.

Let's assume that 60% of the players of any given MMO are casual. They play for 20 hours per month (5/week). In both MS and Ru, these casual players are almost a total loss. Only about 1% (presumptively) of these casual players ever buy premium items or membership. If premium play costs $10/month then each casual player is worth 10 cents per month in gross revenue.

Compare this to a hypothetical grid-based game. If a computer being used typically generates 10 cents per hour, let's estimate that a computer running an online game has fewer cycles to spare than a typical computer; 5 cents per hour instead of 10. Those casual players generate 5 cents per hour of play, and at 20 hours per month that means a casual player is worth a dollar per month. This is roughly ten times what the games traditionally manage.

Another 30% of the player base is supposedly moderate players. These players are in the game about 30 hours per month (7-8/wk). Because moderate players reach higher levels than casual players, they are more likely to have a need for premium services and items. Assume that 10% of moderate players buy premium items or memberships at $10/month. This means that a moderate player is worth a dollar per month to a traditional 'free-to-play' MMO. However, by 5 cents/hour model, a moderate grid game player generates $1.50 in that same month. That value didn't come from the player's pocketbook, it just came from her computer.

The last 10% represents the hardcore players that clock 60 hours per month or 15 per week playing a given MMO. These players are very likely at the top rungs of their games and so will definitely be using premium services. Each hardcore player is therefore worth $10 per month to a traditional game, but only $3 per month to a grid game. So for Maplestory or Runescape it would be better to have a hardcore player on the premium system instead of the grid system. However, for a hardcore player of three-years, those unused processor cycles represent $360 in savings.

Also, taking the weighted average of all the group revenues, both the grid and premium systems generate about $1.30 per month per player. This means that each system is equally fit for supporting the development and service of the game. However, because a game on the grid economy costs the player nothing for the full game, a game with a grid system should draw more players and therefore more gross revenue than a game that charges or advertises in order to allow full play.

Therefore, for both the players and providers of free online MMO games, generating revenue from utility computing is a better business model because it creates value from what is otherwise waste.