In working for both first-party (EA/Sony etc.) development studios and third-party (Kuju/Argonaut etc.) ones, I've noticed the following functions that publishe

In working for both first-party (EA/Sony etc.) development studios and third-party (Kuju/Argonaut etc.) ones, I've noticed the following functions that publishers fulfil. I've mainly worked on AAA titles, with a smattering of MMOs, so this doesn't necessarily apply to other segments - specifically, I'd expect indie games to work very differently. This is doubtless an incomplete list. 1. Money. I've put this first for obvious reasons, but as the article alludes to, the industry cycle itself is incredibly unsteady (I can look up the exact figures, but I recall that something like 50% of yearly sales of boxed games are done within one month around Christmas), alongside the studio financial model being obviously unsteady (work for 3-5 years to release a game that generates 90%+ of its lifetime revenue in 2 months). When you look at it like this, publishers look like the subservient ones: a studio that produces a flop often has another title in development with another publisher, and just misses out on royalties; a publisher that has a string of flops goes bankrupt. If this relationship didn't exist, AAA studios would be less likely to stick together (as the failure of even one game would disperse the team) - that means no Blizzard, no id, no Bethesda, no Rare, no Lionhead, no Bioware - essentially any studio who have ever pushed back a game, or released a market failure or a slow-burner would probably be out of business. 2. Platform-holder pressure. Sony, MSFT, Nintendo have (no so much now) incredibly strict rules about what games can do on their platforms, ranging from "no printf() calls" to "no GUI elements within 12 pixels of the edge of the screen" to "your game is not allowed to crash after being played for 48 hours straight". As a developer going straight to platform, you have zero leverage with the platform-owner (remember, they know that if your game flops you're gone as a company). Going through a publisher, well, far be it from me to imply that any underhand tactics, such as threatening to delay other titles so they don't coincide nicely with the platform-holder's plans, might occur. 3. Boring stuff. Games need hosting, translation, websites, promoting the game to retailers (which is a more adversarial process than you might expect), supply chain logistics, getting the game rated, getting manuals written, dealing with returns from physical shops, collector's editions, merchandise, blah blah blah...after a decade of work in the industry, I'm aware that my knowledge here barely scratches the surface of what's required to put out an AAA game once the code is signed off and the assets finished. This is doubtless the easiest function of publishers to replace, but no-one except publishers offers this service at the moment, and any studio who deliberately goes without this is shooting themselves in the head. 4. Quality. This doesn't always work well, but a publisher is often the first party to see the game outside the studio, so they have the best information to predict the state of the art when your game launches. Advice like "those shaders aren't good enough" is never easy to hear, but it's always better than releasing a game that is panned for dated graphics. 5. Media relations. This doesn't get talked about much...but publishers are the ones with all the power over both paper magazines (exclusives, pre-reviews etc.) and online games sites. In return for access to games before the release date, there's usually a subtle or not-so-subtle pressure to give the games a positive review. (This is why games that are given an average of 8/10 being regarded as a flop.) Media outlets that don't follow this protocol are simply frozen out of the pre-review process. I don't think this is a good thing for the industry, but in the same way as point 3 above, any developer who forgoes this process is putting themselves at a disadvantage relative to other development studios following a more traditional model.