Most games fail because most game companies are stupid.

Most movies make a profit, and if they don't, they even get a 2nd chance later with DVD sales or a TV run.

You are mixing everything up. 443 is the total amount of games when you consider that every version of the same game is a different game (like Halo 3 NTSC/US, Halo 3 PAL and Halo 3 NTSC/JP). Those million sellers are a much higher number up to 126 (not every game has a JP yet, like GTA 4). Your revenue argument is really weak. You are assuming every game sold at 60 bucks. Some games are way cheaper and several blockbuster get pricecuts along the way.

I think you've just hit on something, probably by accident, RubangB. Games primarily have only one channel of profit, their initial release onto systems. It wasn't necessarily that way all the time; a lot of classics from the 1980s and 1990s had arcade releases first, then came to home consoles. And of course, the eventual re-release of games as ports to handhelds also ensures another wave of profits for certain developers.

But with the arcade dying, and handhelds remaining as they ever have been, new games are facing more obstacles to profit than they did two decades ago. Many of them only ever do get that first release, and nothing else. No prospects for future sales for at least a decade, when the most popular portable tech catches up to what is at the moment the current tech. And that's assuming they even do get the port treatment.

The video game industry really needs another release method. Maybe something simple, accessible, perhaps something you could use from an existing console. A release medium where developers can take risks with a new idea and see if it'll sell. Oh wait, we have that, don't we? XBLA, PSN, and especially WiiWare. Now if only developers would home in on that...

As I realized that I could not count profit I switched to revenue which is freely available information. 80% of games make 20% of the revenue. Or 20% of games make 80% of the revenue.

As development costs are fixed once a game is produced essentially, if you get 30 million in revenue from a game which cost 20 million your return on investment is 50%. If that same game nets you 60 million in revenue, your ROI is 200%. Double the revenue, quadrupel the profits.

Furthermore I showed that at retail the top 10% of games made 120 million in revenue or likely 60 million in the hand for publishers. If the average cost of marketing and development is 30 million then the average ROI for the top 10% of games is 100%. Furthermore I counted the sales/revenue from just one console, I excluded the PS3 from this and if it was included it would just further strengthen my point.

I make 600 widgets, and it cost me $950 to make them. If 550 widgets are stolen but I can sell 50 widgets for $20 each, I make $1000 so im still profitable even after I lose money on 550 of them. 

I was proving the rule that 20% of games make 80% of revenue, I could spend 30 hours working on it, but I chose not to. 443 games is just a number on metacritic. Bah.

I will try to provide better numbers. The following are part of 360 million sellers without a Japan Release:

Marvel UA, GTA 4, Guitar Hero 2, Mass Effect, Rock Band. So we got 120 games, not 42 (42 X 3 - 6). I may be wrong if the number you got from Metacritic included every downloadable game. I only took a quick look at gamefaqs and saw that the 400+ number was wrong.

Instead of making ridiculously complicated numeric claims, you could just admit that you mis-worded it and should have said that most companies are not AS profitable as the top-selling companies. To which I say, of course not. Profit, however small, is still profit. Just because a handful of businesses reap immense profits does not mean that the businesses on the lower end of the spectrum do not reap any. It is irrelevant, actually, that those big players do make so much; the fact that profit remains to be made is why there are competitors that make up the other 20% of the market. If it weren't at least marginally profitable, they'd get out of the business or be forced out.

I see what you're getting at, but there are some major and unwarranted assumptions you're making here. First, you're assuming that the publisher will get every dollar that the customer pays at retail: that is demonstrably untrue.

http://www.forbes.com/2006/12/19/ps3-xbox360-costs-tech-cx_rr_game06_1219expensivegames.html

"For the companies that do put next-generation titles out early, making a profit is tough. Namco Bandai president Takeo Takasu said his company needs to sell at least 500,000 copies of each PlayStation 3 game it creates to make a profit. Analysts predict that some other publishers will need to clear 1 million units to get in the black--and start making about $1 per game sold."

If a publisher is really only making a profit of one dollar after they've broken even, your return on investment figures go straight out the window (Note: I'm uncertain just how long that holds for. You're widget example is a perfect example of the flaw in your thinking: you're not going to sell widget at $20 a pop (competition being what it is) but even if you magically do, retailers and shipping are going to be taking large bites out of your numbers.

No, if we're to continue this discussion, we'll be needing profitability numbers, rather than trying to exptrapolate them from revenue. That, obviously, is going to take a lot of work (slogging through financial records).

You are making correct assumptions on wrong data. That 400 + games is either with every downloadable game included or different version of the same game counting as different games (PAL, NTSC, special editions). He said he took it from Metacritic. The revenue numbers is also based on a ridiculous assumption (all games sell at full price, there isn`t any bundles).

Truly, noname has a valid point there. And he has highlighted a key weakness in analyzing the cinematic-style games market: it's in decline, meaning you don't get an accurate picture of game costs as they are in a growing or stable market. In a declining market, of course expenses outdo revenue; that is one of the most obvious signs of market decline. It's an easy mistake to make, though, given how economics literature is phrased. Really, I wish they would not use the phrase "ceteris paribus" (all things remaining the same) in economics, as nothing ever remains constant in a market. It gives students the wrong idea about how things work...