To be honest, it seems that neither model works. Within the last year we've seen Daybreak cancel EQNext, shutter landmark only months after its release, and proclaim H1Z1 their new flagship franchise. Funcom similarly has just life-supported all 3 of their remaining MMORPGS (Anarchy Online, Age of Conan and The Secret World), while launching similar limited-scope survival horror games like The Park, Conan:Exiles, and the incoming Secret World: Legends. Moreover, Turbine has dumped LoTRO and DDO onto Standing Stone while shuttering the Asheron's Call franchise, then turning their attention to the mobile market.
Time marches on, and it may just be that the MMORPG as we know it has run its course, being too costly no matter how it's done to sustain itself over the years as its fickle user base constantly shifts games. Even successful ventures with more limited massively multiplayer like Elder Scrolls Online and the Destiny franchise seemingly cringe away from the MMO moniker.
PS/ Downix replied at the same time as I did, and I hadn't seen his post while I was writing. I was replying to MGLZadok.
Consider the possibility that the problem inherent to all of these failures isn't the economic model, it is the giant company on top that demands extremely high returns on investment.
Ten years ago, people were talking about the demise not of triple-A MMOs, but triple-A *everything*. That all of video gaming was on life support, and mobile gaming was going to replace everything with its lighter financial burdens and lower break even points. Ha ha, instead mobile
became triple-A in all but the platform, with mobile games running budgets of tens of millions of dollars of development costs and extremely high production value thresholds.
It doesn't fundamentally cost a lot to make an MMO. It doesn't cost a lot to operate an MMO, at least not in relative terms. The vast majority of the money spent on game development, MMO or otherwise, is in content development and the customization of engines and tools to support that content development. To put it another way, the expense of building an MMO, as I see it, are the things you spend money on to differentiate your experience from everyone else's, to give enough people a reason to play it. Usually, that's done either with the content itself or the unique features that content delivers.
An interesting question to ponder is: when City of Heroes was running, what was its set of unique features, and how expensive were they to deliver. I'd argue that CoH's primary unique features were things intrinsic to its design and the core of its operational structure, and those things were not expensive to deliver on their own. Money is the brute force method of making a unique game. But we had a perfectly viable game platform that did not contain either a state of the art technology platform nor unique-featured content. It did contain two unique features you don't see in other MMO platforms: a high degree of customization freedom, and what I can only describe as supremely broken gameplay. They let us be overpowered behemoths and just made it work for the game. That's not going to attract everyone, but it would attract some people as a differentiator. And it wasn't expensive to keep those mechanics moving forward.
In my own line of work, there are lots of times I run into situations where people can literally make millions of dollars, but no one is doing it because most of the people in a position to try want to make
tens of millions or hundreds of millions of dollars and it just isn't worth their time to pursue something that will only make maybe a million dollars. If your threshold for success is a hundred million dollars, probably 99.9% of all game ideas are doomed failures. If your threshold for success is "get paid a decent amount of money and be able to pay your contributors as if they were professionals" most game ideas are probably nominally viable.