41 minutes ago, Tom Sloper said:

Here's one I came up with a few years back. Okay, so 18 years ago. 

Awesome, 18 years is a few years long time for a kid2 seconds for an older person. Ignore the gray in my hair.

There's no one-size-fits-all, but it is something you can estimate: Look at similar projects and use them as a guide.

These days a AAA game costs on the order of $500M to create, with about $200M spent on development costs.  

Full-sized big games (that aren't quite AAA) cost on the order of $50M-$100M in development costs.

If you want to say "a game like ${A} or ${B} or ${C}" you can get closer, go do some research on those titles. If you can't find firm numbers, look at the size of development teams and multiply by salaries, look at published operating budgets for publicly traded companies, look at financial statements you may find online, look up estimates from aggregators like VGChartz. If all three are similar to what you want to make and they cost $75M, $45M, and $60M each, then you can use that for your own estimates.

Don't forget that a startup has additional costs since they don't have a large library of resources to rely on nor existing piles of hardware, nor a history of working together in productive ways. Each of those has more costs.

Game credits can be quite revealing.

3 hours ago, frob said:

There's no one-size-fits-all, but it is something you can estimate: Look at similar projects and use them as a guide.

These days a AAA game costs on the order of $500M to create, with about $200M spent on development costs.  

Full-sized big games (that aren't quite AAA) cost on the order of $50M-$100M in development costs.

If you want to say "a game like ${A} or ${B} or ${C}" you can get closer, go do some research on those titles. If you can't find firm numbers, look at the size of development teams and multiply by salaries, look at published operating budgets for publicly traded companies, look at financial statements you may find online, look up estimates from aggregators like VGChartz. If all three are similar to what you want to make and they cost $75M, $45M, and $60M each, then you can use that for your own estimates.

Don't forget that a startup has additional costs since they don't have a large library of resources to rely on nor existing piles of hardware, nor a history of working together in productive ways. Each of those has more costs.

Game credits can be quite revealing.

Thank you. 

Don't forget about editor software, content production pipelines. Even if you use an existing 'engine' which supports dragging 'standard' assets 'into the editor' somehow, for an 'AAA' game you'll find yourself in need of building very custom data pipelines. That involves some more engineers, tech artists, and a lot of time, plus a crystal ball so they design and implement it in such a way that ART rework will be minimised. You can't avoid custom technology for a big, original game.

Also don't forget about platforms. Even if your 3rd party 'engine' claims to support all the various consoles and whatnot, chances are you'll still need your own engineers doing a lot of platform-specific work and fixes.

I don't think it's even remotely possible to create what we all very vaguely understand as an AAA game with 10 people in full-time developer roles in 2 years from scratch. For a second game of the same genre, almost unchanged technology, and well established pipelines, it'd still say that 10 people and 2 years won't cut it, still an order of magnitude below reality.

9 hours ago, pcmaster said:

Don't forget about editor software, content production pipelines. Even if you use an existing 'engine' which supports dragging 'standard' assets 'into the editor' somehow, for an 'AAA' game you'll find yourself in need of building very custom data pipelines. That involves some more engineers, tech artists, and a lot of time, plus a crystal ball so they design and implement it in such a way that ART rework will be minimised. You can't avoid custom technology for a big, original game.

Also don't forget about platforms. Even if your 3rd party 'engine' claims to support all the various consoles and whatnot, chances are you'll still need your own engineers doing a lot of platform-specific work and fixes.

I don't think it's even remotely possible to create what we all very vaguely understand as an AAA game with 10 people in full-time developer roles in 2 years from scratch. For a second game of the same genre, almost unchanged technology, and well established pipelines, it'd still say that 10 people and 2 years won't cut it, still an order of magnitude below reality.

What if you are already using a pipeline for CGI assets in Houdini and those will be the main art assets with anything extra need being shot in the studio? If you already have the offices and motion capture and sound stages? 

10 hours ago, Hodgman said:

That's a triple-I team, not a triple-A team  

This is hilarious!! Triple me team, time to get the robots out.

Those off-the-shelf 'CGI assets' might not perform well enough and you might not be able to have as many of them (on screen) as you'd like for your specific game

You're still implying that everything's ready. It never is, things evolve and what worked 5 years ago isn't often enough today. Your team will have to constantly accommodate for that continuous change. That requires people and time. Even if you throw really a lot of money into it (buy a lot of 3rd party/outsourced solutions and assets), you might need a surprising number of your own people integrating it.

I'm not sure if I'm making much sense or just sound too pessimistic, I just see how much man-time things seem to take in a AAA studio...

The way I see it is that even if you have one development team preparing all the tech, pipelines and workflows and then you suddenly start your other team, the A(AA)-team which you're talking about, to make you a game in 2 years, you'll still have to have that first team upgrading the tech, pipelines and workflows in the meantime for the next game. And 10 people will suffice neither of the two teams and also they'll never work well in separation.