With no abundance of platforms, be it mobile or console based, it seems video game discovery regardless of game genre is happening in a random fashion. What startups, existing services do you see helping create more signal from all the noise?

I think that right now there is an opportunity for someone like a Game Stop to own that space. I have not seen anything cohesive.

What are some of the features / qualities you are looking in that game discovery service?

So several personalized guide/search startups are reportedly moving in this general direction (Ness, Stamped, Matcha, etc.).

The challenge is that games present a unique challenge compared to other verticals (restaurants, moves & tv, etc.).

The first, and biggest, hurdle is your data backbone. How does your service know which games exist? There's MobyGames & Spong, I'm not aware of them having an API but maybe you can straight-up license their data. Beyond that, the next decent source seems to be the open games database which has fewer than 11,000 features.

Once you get that, you have to build your recommendation algorithm which can be based on genre, etc. similar to what people are doing for movies and restaurants, but games present a unique curveball here too as to which platform to recommend over another.

Not such a big deal in terms of XBox or PS3, but it makes a huge difference in tablet vs. computer. For example, I've been waiting to play Bastion and grabbed it when it came out for iPad. I can see where I would LOVE the game, but I can't stand the way it controls on the iPad.

Dakota there are definitely some challenges, but none that i think that can't be overcome. I've been surprise by the absence of a service that takes a Netflix taxonomie approach to solving the problem, that is treating platform and traditional genres at arms length and creating a new taxonomie by which to tag all games ie foreign, thrilling and suspensefull instead of chinese-born MMO.

Do you think the main problem is just coming up with a new taxonomie for game classification?

I believe the primary problem with game discovery today is that many publishers haven't really bothered to keep up with the developments in marketing strategy and digital distribution technology.

Indeed, their marketing eco-system hasn't changed since the 80s: rely heavily on gaming journalism (whether off-line or on-line).

I think a recent example of a beautifully executed promotional strategy was Modern Warfare 3: they understood target audience, modern marketing landscape - and facilitated discovery.

My take is that the real winners will be startups that offer high quality outsourced advertising, marketing and viral promotion to game publishers (and maybe, just maybe for those that dare, game developers directly).

Oh I don't think anything I listed should be thought of as insurmountable barriers, rather that's just the work that needs to be done to get your foot in the door. It is just that I suspect the data isn't as readily available as it is for movies, tv, restaurants, etc.

Once that work has been done is when the real challenges will unearth themselves.
- How good is your similarity data?
- What player preferences transcend traditional genre divisions?
- How do you figure out which version/platform of a game to recommend?
- How do you decrease friction between the player accepting a recommendation and actually being able to play the game?

I think the future of game discovery lies beyond genres. What Netflix has done with movies is pretty good but I do not think it is deep enough. As video games evolve as a medium, there will be a greater focus on the actual content of the game. By this I'm talking about things like the emotions/tone of the game or the overarching theme.

For example: Discovering games similar to say Gears of War is not just about finding a third person shooter with monsters. There is additional data points such as games that deal with the trials and tribulations of war or the loss of loved ones.

Brilliant insight. So true. Gaming can be directly tied into personal preferences, interests and passions. By using social profiles and personal interests gaming companies can not only hyper target recommendations but even make in game customizations that's are deeply personal.

I don't know how all of this ties together, but libraries could be a great place for game discovery. A lot of libraries now are lending out games (I started two collections here portlandlibrary.com and here cmclibrary.org) and many libraries also have a lot of programs around games (here's a national gaming day that the American Library Association runs: ngd.ala.org)

At my library, we play a lot of FIFA '12 for XBox 360. Many of the teens I serve don't have the money to afford these systems. We're their arcade.

@Dakota, some great insights indeed. Again a little surprised no one has made the connect your FB likes and we will tell you which game you might like.

@Ian, I hear you, and although i think the Netflix taxonomie can be improved upon, i think it would infinitely better than how games are currently classified ie MMORPG, RTS, FPS etc etc.

A better classification tied to a more personal tie whether to your interest or interest of your friends seems like a great roadmap to improve game discovery. Sounds like we just need to hit some VCs next :-)

@sean see my last comment

I think Dakotareese identifies a key hurdle: the data backbone. I also agree that it's not an insurmountable hurdle, but it's going to take a lot of work. Perhaps use thegamesdb.net as a starting point.

All the other suggestions can be catered for in a good taxonomy: it should track attributes not only at a genre level, but also at a personality level (as Ian suggested).

So the way I see it there are three steps to building this: firstly the taxonomy, secondly populating your data structures according to this taxonomy, and thirdly, building the service on top of your data structures. As a brilliant value add, also include promotional services to game publishers (what Oliver Woods said).