I have a few tales I could tell about a certain, recently bought out games studio, who's staff were until recently building games using Maya on 2ghz P4s with 512mb RAM, Geforce MX440s, backing everything up to their own *personal* hard drives [the network drives were less reliable, and slower, than a USB hard disk]. Meanwhile, rather than invest £20,000 on better equipment for the artists machines, the MD chopped in his six month old Merc CLS for an F430.
I have to stop now before I start thinking about the other tales I have been told.
Suffice to say I'll never get involved in the games industry - from all accounts it's vicious, vicious work, but the payoff is that it really is only the best talents who make the good games - my mate referenced above is really chuffed when the games he and his guys and girls slog their guts over day and night get good reviews, which, more often than not, they do.
Just never kid yourself that the only way to making good games is to spoil your employees - Rare and MSGS seem to be the exception to the rule in that regard.
As I understand it, the only way to 'break in' to the games industry is to have a team of fucking good people who can build fucking good games, and be prepared to work your fingers to the bone until someone sees the potential and picks you up or you get a good license - star wars, indiana jones, bond, etc.make a cracking game with a famous name attached, and suddenly you're all over the press and get attention from the big boys [See: RareWare, with both Nintendo and MS, Travellers Tales and EA are another] or you get something huge on your hands [like WoW, for example] that then starts looking after you in return...
You can make crap, with a famous license [FIFA is a *perfect* example - it went through about three iterations/sequels back in the mid 90s with almost no changes whatsoever, each version being as crap as the other...] and survive, but if you just make crap, you're dead in the water - especially with development costs these days. You can't just get someone with a pallette to animate some sprites, change the colours and details here and there, drop it onto a dev cart/disc and send it off to be published - not when people are expecting photorealistic lighting and millions of on-screen polygons these days - the money involved, and the risks, must surely be astonishingly scary by now?