Valve is very famously terrible at actually shipping games. Mostly they’re good at taking 30% off of most of the PC gaming industry and occasionally hiring teams that made popular mods to polish the mods. https://www.pcgamer.com/valves-unusual-corporate-structure-c...

Is there any evidence that this approach outperforms? If you rest on a cash cow, basically any methodology that doesn't literally slaughter the cow is going to work out.

Isn't the fact that they're the market leader in what they do proof enough? Would they have created/found this cash cow if they were a traditional company?

People try to poke holes in their ways, yet they have made legendary games, amazing hardware and absolutely dominate the game distribution market.

Basically all of their competitors use the traditional management style. What other evidence would you expect?

A sample of one isn't enough, and in any event their success is not reproducible. They had a major title in Half Life 2, convincing people to install Steam, but that was their last major title and it is not clear that they used this process at the time.

Going forward: Someone else created Counter Strike, someone else created DoTA, someone else created Portal, someone else created Team Fortress, someone else create Left 4 Dead. Valve took these projects on and polished them, but they didn't really create something new until Artifact - which failed in the market.

What you also don't see is all the unannounced projects that went nowhere, because that's what you get when people just get to play around on their own, indefinitely.

> someone else created Portal,

A team created Narbacular Drop, and that same team then worked on Portal as part of Valve. It's not "someone else" by any stretch.

What? Aren't those games Valve orginals since the beggining?

In the sense that these games were conceived by teams working at Valve, using this management structure, the answer is: No, none of them are.

Most of them have their origin in community mods, some already having gained substantial traction on their own.

Sure, the original idea might've already been there, but they built the game all the same. Saying it "wasn't them" is misleading. They built it from scratch, with existing concepts from mods.

Maybe their ideas department isn't the greatest, but their execution on these games was very good in the long run. They popularized hats and make absurd amounts of money from it.

Let's not forget the Steam Deck. Early on still, but an excellent product in a whole different area.

That handbook is from 2012 so 12 years old now. Are there any updates?

I interviewed there, and took a tour of their office. From what everyone said and what I saw, I don't think the handbook is outdated.

Valve is also fairly bad at maintaining focus on smaller projects. High-profile green-field projects like Steam Deck are fine, but lower priority tasks, particularly maintenance work, suffer. (For example, most of Valve's first-party games - including major titles like Half-Life and Portal - have been unplayable on macOS for nearly five years because there's no 64-bit build.)

I’m on not saying you’re wrong, but old games don’t really count as evidence of lack of maintenance. Games generally don’t receive maintenance unless there’s an ongoing revenue model, which there isn’t for Half-Life and Portal.

Valve did a big marketing push last November for the 25th anniversary of Half-Life, which included a major update for the Windows version. Meanwhile, the Mac version won't run.

https://www.half-life.com/en/halflife25/

TF2 and hats?

Valve most likely doesn't use that handbook in full any longer. Lots of things changed.

For instance, they have a round table where everyone would judge everyone else. So if too many people don't like you, the axe is coming. Sounds miserable.

Don't they infamously have a shitload of fiefdoms and informal structure?