Alan Wake 2 (for example) did not spend a decade in development, but somehow blow “pretty much 100% of the budget” after launch.
You can maybe salvage that sentence fragment by insisting we’re talking about multiplayer-only “live service” crap that goes on for years and years after launch… but the topic you named is distribution. The marginal cost of software is essentially zero. Supporting customer N+1 is a rounding error. Valve basically has a monopoly on PC game distribution and only employs a couple hundred people. Do those salaries cost money? No shit. But relative to, conservatively, half the money spent on PC games? Fraction of a percent.
“Keeping people employed” takes a lot of money because making a game takes a lot of people a long time. Shipping is the cheap part. Has been since CD-ROMs. In many infamous cases, people were not kept employed once their game shipped, because all those people were not necessary to make all of the money off of the game.
Most games don’t get years of full-staff support. They simply do not need the army of artists who made the whole the thing from scratch, when all that’s happening post-launch are support jobs. Some games are lucky enough to get significantly more content - paid or otherwise. But they’re the exceptions.
At some point, development ends. Nobody’s paying the full team that made Sekiro to keep making Sekiro. Mostly they moved onto Elden Ring, and at this point, surely they’ve moved on from that.
That’s what’s “keeping people employed.” They made one game. It launched. It got finished-ish, generally not long after. And then their managers put them on another game.
These games cost so much money because those higher-ups will put a thousand people on one project for years. The factors pushing budgets and scope up up up are mostly competition and marketing. Having a billion dollars to throw at a project leaves you in rarified company - which plays nicely with how spending half that money spamming one game’s ads is more cost-effective than dividing it up between ten games.
Big games also enjoy a better “long tail.” FromSoft can keep selling Sekiro, forever, for an upkeep cost of approximately fuck-all. No huge staff is dedicated to those files on a server. The cost of keeping them available isn’t literally nothing, but it is practically nothing.
And no, the cost of those things is not “marginal” or “a fraction of a percent”.
Hey, good thing that’s not what I used those words for! Marginal cost is a specific economic concept. It’s the cost of selling 10,000 units versus 9,999. For games? That’s essentially nothing. Which is how Steam handles a supermajority of sales and distribution with a tiny little company.
You’ve mentioned it a couple of times now, and it`s very important to remember that a) Valve is an extreme outlier. They’ve spent the past twenty years crowdsourcing, distributing or automating as much work as humanly possible, including ways to get the user to pay for the bandwidth. The reason that number of employees was news is that it’s absurdly small. And I still bet their server costs are astronomical.
And, b) that Valve isn’t making the games. That’s one of the bits they’ve externalized. They made games when they needed to seed their platform. Now they don’t have to, so they don’t. They do charge a big chunk of money for the privilege of selling your game (as well as to cover those huge server costs), but the developers/publishers are the ones paying for the development team’s cost, which continues to tick up whether you’re making the game or selling the game, and regardless of how many copies you’re selling.
As for the actual point. It’s not nearly as easy as “the in-house guys do the next game”. Development effort isn’t fully flexible. Your modellers or UI designers don’t work on the same parts of the process as your systems engineers or, obviously, your community managers. Prototyping doesn’t engage the same people as the production stage. Resourcing is an ongoing struggle that makes up the full time job of a ton of people.
Not that it matters that much to the point at hand, because the point at hand is that your operating costs remain flat whether or not you have everybody workingefficiently. Yes, your project managers are constantly working to optimize that output, but that ticker is always going. For a digitally distributed game it’s not (as) unit-based, sure, but it’s time-based, but this entire conversation hinges in that you are misrepresenting how unit-based it was with physical distribution. From the publisher’s perspective the only difference is how much revenue you get per unit sold. Manufacturing and distribution are just another percentage of that sale you don’t keep, like taxes or the platform/retailer cut.
The math is pretty ruthless: there is a cost per active user, which is the corresponding fraction of the cost of making the game (including shipping and manufacturing, if physical), the cost in marketing per onboarded user, the corresponding fraction of the cost of running the studio (again that ticking clock) and the cost of supporting them, among other bits and pieces. There is also an average revenue per user. The second number needs to stay higher than the first. This matters more to live service games because they get to tally those numbers constantly, but it is not different for one-off games, it’s just that the cost and revenue get counted all at once.
Alright, I’m incredibly deep in the weeds, but I’ll add one final note: the argument that scope is being pushed up also lacks nuance. Scope is more flexible now than it has ever been. There are major hits out there made primarily by a single developer. Small indies, mid-sized indies, double A releases, mobile games… You’re not wrong that AAA majors use their size to avoid mid-scope competition by chasing massive evergreen moneymakers, but that doesn’t meant that’s all (or even the majority) of the industry.
Yes, neatly demonstrating how this was supposed to be a conversation about distribution.
Development effort isn’t fully flexible. Your modellers or UI designers don’t work on the same parts of the process as your systems engineers or, obviously, your community managers.
… yes, and some of those parts are support issues, while others are not. Coders stick around for patches. Modellers, not so much. Next game.
Aaand you’ve deleted all your comments in this chain the very minute I opened this one. Fuck me, I guess.
deleted by creator
Alan Wake 2 (for example) did not spend a decade in development, but somehow blow “pretty much 100% of the budget” after launch.
You can maybe salvage that sentence fragment by insisting we’re talking about multiplayer-only “live service” crap that goes on for years and years after launch… but the topic you named is distribution. The marginal cost of software is essentially zero. Supporting customer N+1 is a rounding error. Valve basically has a monopoly on PC game distribution and only employs a couple hundred people. Do those salaries cost money? No shit. But relative to, conservatively, half the money spent on PC games? Fraction of a percent.
“Keeping people employed” takes a lot of money because making a game takes a lot of people a long time. Shipping is the cheap part. Has been since CD-ROMs. In many infamous cases, people were not kept employed once their game shipped, because all those people were not necessary to make all of the money off of the game.
deleted by creator
Your in-house guys do the next game.
Most games don’t get years of full-staff support. They simply do not need the army of artists who made the whole the thing from scratch, when all that’s happening post-launch are support jobs. Some games are lucky enough to get significantly more content - paid or otherwise. But they’re the exceptions.
At some point, development ends. Nobody’s paying the full team that made Sekiro to keep making Sekiro. Mostly they moved onto Elden Ring, and at this point, surely they’ve moved on from that.
That’s what’s “keeping people employed.” They made one game. It launched. It got finished-ish, generally not long after. And then their managers put them on another game.
These games cost so much money because those higher-ups will put a thousand people on one project for years. The factors pushing budgets and scope up up up are mostly competition and marketing. Having a billion dollars to throw at a project leaves you in rarified company - which plays nicely with how spending half that money spamming one game’s ads is more cost-effective than dividing it up between ten games.
Big games also enjoy a better “long tail.” FromSoft can keep selling Sekiro, forever, for an upkeep cost of approximately fuck-all. No huge staff is dedicated to those files on a server. The cost of keeping them available isn’t literally nothing, but it is practically nothing.
Hey, good thing that’s not what I used those words for! Marginal cost is a specific economic concept. It’s the cost of selling 10,000 units versus 9,999. For games? That’s essentially nothing. Which is how Steam handles a supermajority of sales and distribution with a tiny little company.
You’ve mentioned it a couple of times now, and it`s very important to remember that a) Valve is an extreme outlier. They’ve spent the past twenty years crowdsourcing, distributing or automating as much work as humanly possible, including ways to get the user to pay for the bandwidth. The reason that number of employees was news is that it’s absurdly small. And I still bet their server costs are astronomical.
And, b) that Valve isn’t making the games. That’s one of the bits they’ve externalized. They made games when they needed to seed their platform. Now they don’t have to, so they don’t. They do charge a big chunk of money for the privilege of selling your game (as well as to cover those huge server costs), but the developers/publishers are the ones paying for the development team’s cost, which continues to tick up whether you’re making the game or selling the game, and regardless of how many copies you’re selling.
As for the actual point. It’s not nearly as easy as “the in-house guys do the next game”. Development effort isn’t fully flexible. Your modellers or UI designers don’t work on the same parts of the process as your systems engineers or, obviously, your community managers. Prototyping doesn’t engage the same people as the production stage. Resourcing is an ongoing struggle that makes up the full time job of a ton of people.
Not that it matters that much to the point at hand, because the point at hand is that your operating costs remain flat whether or not you have everybody workingefficiently. Yes, your project managers are constantly working to optimize that output, but that ticker is always going. For a digitally distributed game it’s not (as) unit-based, sure, but it’s time-based, but this entire conversation hinges in that you are misrepresenting how unit-based it was with physical distribution. From the publisher’s perspective the only difference is how much revenue you get per unit sold. Manufacturing and distribution are just another percentage of that sale you don’t keep, like taxes or the platform/retailer cut.
The math is pretty ruthless: there is a cost per active user, which is the corresponding fraction of the cost of making the game (including shipping and manufacturing, if physical), the cost in marketing per onboarded user, the corresponding fraction of the cost of running the studio (again that ticking clock) and the cost of supporting them, among other bits and pieces. There is also an average revenue per user. The second number needs to stay higher than the first. This matters more to live service games because they get to tally those numbers constantly, but it is not different for one-off games, it’s just that the cost and revenue get counted all at once.
Alright, I’m incredibly deep in the weeds, but I’ll add one final note: the argument that scope is being pushed up also lacks nuance. Scope is more flexible now than it has ever been. There are major hits out there made primarily by a single developer. Small indies, mid-sized indies, double A releases, mobile games… You’re not wrong that AAA majors use their size to avoid mid-scope competition by chasing massive evergreen moneymakers, but that doesn’t meant that’s all (or even the majority) of the industry.
Yes, neatly demonstrating how this was supposed to be a conversation about distribution.
… yes, and some of those parts are support issues, while others are not. Coders stick around for patches. Modellers, not so much. Next game.
Aaand you’ve deleted all your comments in this chain the very minute I opened this one. Fuck me, I guess.