OnLive

I guess I am jumping on the bandwagon here, but for some reason today I could not stop thinking about how blindingly obvious it is that OnLive is perhaps the least-likely-to-succeed gaming technology idea I have ever seen.

The idea of how the subscription business model would have to work has occupied my brain all day. Like anyone who has ever played a video game over the Internet, I am highly skeptical of it as a technology. I see I don't need to talk about that because Eurogamer eviscerated it already. I am a little late to the party on questioning it's bandwidth assumptions, TheGameReviews also did that already (though I would have a lot more nuanced argument that I may make later).

What I question, is assuming they've magically programmed these two concerns away (and iuf they did, why on earth they're selling a video game equivalent to DiVx instead of making bajillions of dollars selling the technologies to, say Microsoft or Google is unfathomable), what must their pricing model be in order to profit, and is that level sustainable in the marketplace?

I thought about it, and came up with a simple-ish equation:

F>[(M*r/Θ)*C]

In this equation:
F= the monthly fees, aside from game purchases.
M = the cost of one hardware machine they need to run a game for one user
r = the ratio of machines to users required, expressed as a decimal.
Θ = the expected lifespan of one of the machines
C = colocation costs per Machine per month

Roughly, let's estimate the numbers as follows in what I assume to be fairly conservative numbers:
M = $1000
r = 0.25 (meaning one machine for every four paying subscribers)
Θ = 18 months
C = $22 (or the equivalent of $700 for a 16-slot rack running 32 CPUs at a fairly inexpensive co-lo facility with enough bandwidth to serve)
That leads us to a monthly subscription cost of at-least $36 to break even on operations. It doesn't service debt, it doesn't pay for marketing, doesn't pay back development costs or the investors much less pay management and overhead.

If I go crazy low I could change the variables to $800, 0.1, 24 and $18, that still puts you at $22/month break-even on operations and assumes they'll have exactly the right facility/hardware spending in place to service their customers.

My high end guess would be $2000, 0.4, 15 and $30, leaving a break even price of $83/month.

Now, maybe their service, despite the aforementioned technological problems, is pretty darned good. What I say, is a new XBox 360 Pro and 2 years of Live! service costs $400, or $16/month, there's no way anyone will pay for OnLive if the cost is much more than that, I don't see how it's possible, I think it's more likely the cost will be at least double that.