Dev kits and engines, while expensive, aren't too great of an expense. Making your own engine does cost quite a bit, but that cost becomes shared if you use the same one with other games.

The issue with creating your own engine aren't so much costs as increasing a lot the development time for the game so making it longer between the time you start development and the time you start getting back money for the sales of the game...

But in the majority of the games today the biggest cost is content creation.

The ever increasing amount of devs used to create ever fancier graphics. some of these games take hundreds of devs many many hours each so you can see that these costs would escalate quickly

The 1st major cost is salaries to employees, with the advancing hardware comes a higher paid developer to make content. If you have a team of 100 people all making 50K-120K a year, then you a monthly burn rate of up to 1 Million a month right there (don't forget about of those upper management types that earn the real the big money as well).

The 2nd major cost is your operating cost - the money it takes to keep the lights on, run the facility and operate the equipment. Building rent, energy costs, IT people... ect.

The 3rd major cost is the money you spend on software if you're not going to build the game engine from scratch. Unreal can cost 3 Million for one title if you want it to be royalty free from Epic.

Add all that together, and just over two years you got a game that is costing you 30 Million - and two years is almost nothing when you want to make a GOOD game, good games take time. 2 years of development can almost certainly equal a not-so-good game. Most AAA big name titles need 3-4 fours years, especially if you are creating original IP. Once set, pumping out a sequel on the same tech, same platform becomes an easy 2 year job with quality almost in the bag.

I didn't even mention marketing costs, or the cost of slowing down production to create demos and the like.

Thanks for that. Won't some money be saved because the early work on a new IP won't involve many people, just some artists working on ideas before the real numbers are needed for coding?

all this plus

food

spa visits

hair cuts

New hardware if something fails

overtime pay

and more

I don't work in the game industry but I do work for a software company. Nearly 80% of our cost is paying salary/benefits. the other 20% is everything else (rent, computers, severs, travel, ect.)

Nope.  Because a new IP requires almost a full size team working in longer pre-production times.  You can spend alot of time making something that isn't fun because it may take six months of dev time before you can honestly critique it.  You then either scrap it and start over, or spend more time evolving the current architecture into something that might be fun.  It's hard to make a judgement on something that is half done, creating original content can be a giant black hole of development.  This is why you see games like BioShock take up to 5 years.

Creating art by itself gets you nothing, creating code by itself gets you nothing - everything must be built in tandum, cross discipline - and the less people you have, the slower this process becomes.  More games than you could count have failed to make it to production because the ideas for the game, for one reason or another, fail to come together in a reasonable amount of time.  Had 2K not come in and fund the last part of BioShock (even though they pretty much raped Irrational Games in the process), BioShock would have never been made it to the shelf - even though it had been in development for years.

I'm just saying I know Uncharted went about a year in pre-development while the main team was working on Jak X, and that was a new IP that got done in 2 years for $20m