10 ms·

The Cabal: Valve’s Design Process for Creating Half-Life (1999)

188 points by thisismyswamp 6y ago

throwaway3699 6y ago
I think the Cabal concept is especially interesting in the wider context of Valve's "self organising" culture. It doesn't seem that this structure was designed, but was emergent.
- PicassoCTs 6y ago
  This structure is definitely not emergent, if left alone, human groups form little fiefdoms around local "lords", the hierarchy and bureaucracy of a feudal system emerges. To have working anarcho-syndicalism is a constant fight against this mechanism. Who decides how much anyone gets paid, and how many virtual crystal gold smileys someone gets for doing what he currently does? How many golden-smileys do you have to have on your shoulder to influence that decision?
  - marcinzm 6y ago
    I remember reading that Valve's internal structure nowadays is lords of fiefdoms even if they communicate externally that it's not. The only difference from other companies is that there's no org chart saying who the lords are, you have to discover that yourself.
walrus01 6y ago
In the context of Cabal and video games, I was expecting something about Destiny 1/2
Netcob 6y ago
That's a refreshingly good example of "design by ...cabal..." actually working out pretty well. I think the key element is how the members actually had to implement their designs. If you just come up with a lazy compromise, it'll start haunting you immediately, and not just after the worker bees did their thing. I think the "cabal" responsible for USB 3.1.1.1.gen2-420gig-whatever should have done the same, with each meeting ending in every member having to explain to their grandma why she can't connect her monitor to her laptop even though the cable fits perfectly, and that it's somehow _both_ the wrong cable and the wrong laptop port.
- danbolt 6y ago
  I've noticed that happening a lot in my time in the games industry as well. The person making top-down design role isn't exposed to the friction that their decisions are making down the hierarchy. I think the worst part is that they're not even necessarily as invested in the product as someone else, so they're kind of structurally going along with the inefficient situation they're in.
wgjordan 6y ago
(1999)
Blikkentrekker 6y ago
> If they walk into a room with other characters, those characters should acknowledge them by at least looking at them, if not calling out their name. If only Gordon returned this courtesy. Silent protagonists are such an awful thing.
- nitrogen 6y ago
  You are Gordon. Did you acknowledge the NPCs by looking at them?
  - Blikkentrekker 6y ago
    I realize why they do it, but it can never work because the n.p.c.'s do not react to me in return. I can talk to them, say things to them, but they will never respond to anything I say. Thus, the end result is that Gordon Freeman is a man whom many speak to, yet who never speaks back, and no one considers this to odd to say the least. — the conversations thus appear ridiculous. It worked better with Chell, because others do not that Chell does not speak back, and Chell actually canonically refuses to speak out of stubbornness, so it's amusing to see Wheatley being obviously uncomfortable from the fact that Chell does not speak back and is simply blabbering to fill the awkward silence.
    - partlysean 6y ago
      NPCs in Half-Life 2 mentioned a number of times that Gordon doesn’t say much and is a “man of few words.” I’m sure this is a nod to his silence in the first game as much as it’s a storytelling device. I’ve always liked that his character never said anything—feels like Gordon can be anyone. Same goes for the lack of cutscenes that take control of the player’s camera. You always feel like you’re looking from your own POV vs. being pulled out of the immersion.
      - Blikkentrekker 6y ago
        > NPCs in Half-Life 2 mentioned a number of times that Gordon doesn’t say much and is a “man of few words.” I’m sure this is a nod to his silence in the first game as much as it’s a storytelling device. That is an in-joke at best and does very little to rectify that n.p.c.'s have entire “conversations” with him and at no point ask as to why he's remaining quiet throughout it. — the conversations are simply completely implausible. > I’ve always liked that his character never said anything—feels like Gordon can be anyone. Same goes for the lack of cutscenes that take control of the player’s camera. You always feel like you’re looking from your own POV vs. being pulled out of the immersion. Perhaps if one go about his daily life talking to walls and being completely disinterested in what a man might have to reflect back. Whether you reply back to them or not, it always leads to an unnatural conversation where one party is talking to a wall that doesn't respond.
jimbob45 6y ago
“ Include an expert from every functional area (programming, art, and so on). Arguing over an issue that no one at the meeting actually understands is a sure way to waste everyone’s time.” This always sounds good in theory but ends up with unnecessary individuals falling asleep during meetings. I prefer either a staggered approach to meetings where the likely unnecessary people are only invited to the end of the meeting OR are simply “on-hand” for the meeting in case they need to be pulled in. I don’t care how critical the input is, if you pull someone into a three hour meeting and they only contribute five sentences, they’re not going to be happy with you.
- kaba0 6y ago
  It depends on the program domain, but I think it must have been quite fun designing a game as opposed to the client still doesn’t know what it wants and this API should instead do this.
- danielscrubs 6y ago
  They where the ones that where going to be responsible for it so how can they be unnecessary? He even said they started to say no more and more as the deadline started to loom over them.
- Jare 6y ago
  In most significant decisions in gamedev you want art, design and engineering (at least - QA, audio, animation and a number of others as well) to be involved. Some people from one discipline may be versed enough in another to cover the role, but the role must still be present, or there will be pain and work thrown away later on. (incidentally, the ability to understand an issue from these multiple but interrelated angles is what makes experienced people so valuable in gamedev, and their constant loss to other less crazy industries so painful)
ArtWomb 6y ago
Cabal. Design Bible. All grounded in the Single Source of Truth philosophy. With feedback from 200+ playtest sessions directly incorporated into the stream. And the result is the cohesive narrative vision in Half-Life. Contrast this with "What Went Wrong with Cyberpunk 2077" exegesis from Bloomberg this morning and I'm still not sure anyone can point to a single reason beyond "we were too ambitious". I wish this breakdown were more technical in nature. Like a good GamaSutra postmortem. I suspect tooling plays an enormous role. UE4 Blueprints are phenomenal for visual development. But does it scale to teams of hundreds of designers? Inside Cyberpunk 2077's Disastrous Rollout https://www.bloomberg.com/news/articles/2021-01-16/cyberpunk-2077-what-caused-the-video-game-s-disastrous-rollout https://www.bloomberg.com/news/articles/2021-01-16/cyberpunk...
- ffhhj 6y ago
  It's interesting to me how the physics bugs such as characters and object getting stuck inside walls, flying cars, failed inverse kinematics and so on have been the usual bugs for so many years now, but there isn't yet a methodology to prevent gamedevs from falling in those old known pitholes. These are very difficult to detect by QA and the way to fix them is tweaking some parameters and geometry here and there. There should be a way to supervise the physics engine, find whether future calculations will cause an invalid state on some object and choose a transformation that won't produce such conflict.
  - vvanders 6y ago
    As with everything in gamedev it comes down to one thing: performance. Anything that can be faked, approximated our outright hacked in the name of perf or memory will be. The canonical example for me(which was ironically is in Half-Life) is that the AI for the grunts is surprisingly simple. If you throw a grenade they were scripted to shout "grenade" and navigate to the nearest nav node. Tons of people thought there was really complex AI behind that behavior but it was just a couple of really well telegraphed "if()" checks.
    - indigochill 6y ago
jgilias 6y ago
> If the players are in the mood for more action, all they need to do is move forward and within a few seconds something will happen. This! I remember being scared to move around when playing Half-Life the first time. Flickering lights still make an instant association with crowbars to me. Great to read a bit more in detail how they managed to make something so great!
jokethrowaway 6y ago
The Cabal sounds great and I'm looking forward to try it. I'll pick the projects carefully though, as it's impossible to ship more than two products with this methodology.
PeterisP 6y ago
"The meetings were only six hours a day" ... ahhh ... (for five months straight)
solidist 6y ago
from the mod makers pov, circa after they launched in 1999: https://medium.com/super-jump/building-a-popular-half-life-mod-during-the-rise-of-counter-strike-fec6a5b9fd8f?source=friends_link&sk=6d1427b3f1d832df06bd5b07aaa456bb https://medium.com/super-jump/building-a-popular-half-life-m... Mod making was a digital cabal concept, with local leads forming fiefdoms. They were opinionated to a design/delivery. Sometimes straying to other mods, but mostly focusing on one concept at a time.
- vvanders 6y ago
  Valve also hired really heavily from the mod community. I spent a good portion of my youth in that scene and the people that they hired they really went far for(relocation, visas, etc). I think most of the leads for the more popular mods(outside of CS obviously) ended up there in one form or another if they were interested in it.
- pugworthy 6y ago
  Nice writeup. I was part of a mod team that got by Valve long time ago, and indeed it was a giddy fun time. We had no idea what we were doing, but we just created a game that we wanted to play - and it seemed to have worked.
TazeTSchnitzel 6y ago
(1999) (if the title gets edited, please delete my comment for me)
alisausaaa 6y ago
Wanna have hot-lovin' conversations? You’re on the right way! - https://adultlove.life https://adultlove.life
nmfisher 6y ago
Fascinating article. One question remains unanswered though - how did Valve afford all of this? Given Half Life was their first game, where was the money coming from to not only produce an entire game, but to then scrap it and redesign/rewrite from scratch?
- Danieru 6y ago
  Valve was started by two Microsoft Millionaires. Money wasn't an issue.
  - nmfisher 6y ago
    Even still, the initial Cabal was three engineers, a writer, an animator and a designer, so you can assume that the broader team was at least that much again. That's 12 people full-time working for 2-3 years (and the actual credits go up to 30, so it's probably even more). At $75k per FTE (in 1996 salaries), that's at least $3 million, which is quite a lot, even for two people wealthy from MSFT stock options.
    - enneff 6y ago
      Gabe Newell was in a senior role at MS for 13 years before he left to start Valve. That’s a lot of money.
    - baud147258 6y ago
      Half-Life was published by Sierra, presumably Valve received some money from them.
kevlar1818 6y ago
TLDR: Small, cross-functional teams build better products.
alisause 6y ago
I was a really bad girl. Punish me with your dick in my mouth. - https://adultlove.life https://adultlove.life
alisaus3 6y ago
Top burny busty chicks only on this site! Follow the link, and you won’t be sorry! - https://adultlove.life https://adultlove.life
anthk 6y ago
On game scripts (literature, not programming), a lot of ideas were shaped before in text adventures/interactive fiction. ZMachine-Inform/Tads have several games based on sci-fi and weird tech environments, such as Babel, or Spider and Web. And the reverse happens, too. H.P. Lovecraft->Alone In the Dark->Anchorhead.
peter_d_sherman 6y ago
>"The first theory we came up with was the theory of "experiential density" — the amount of "things" that happen to and are done by the player per unit of time and area of a map. Our goal was that, once active, the player never had to wait too long before the next stimulus, be it monster, special effect, plot point, action sequence, and so on. Since we couldn’t really bring all these experiences to the player (a relentless series of them would just get tedious), all content is distance based, not time based, and no activities are started outside the player’s control. If the players are in the mood for more action, all they need to do is move forward and within a few seconds something will happen. The second theory we came up with is the theory of player acknowledgment. This means that the game world must acknowledge players every time they perform an action. For example, if they shoot their gun, the world needs to acknowledge it with something more permanent than just a sound — there should be some visual evidence that they’ve just fired their gun. We would have liked to put a hole through the wall, but for technical and game flow reasons we really couldn’t do it. Instead we decided on "decals" — bullet nicks and explosion marks on all the surfaces, which serve as permanent records of the action. This also means that if the player pushes on something that should be pushable, the object shouldn’t ignore them, it should move. If they whack on something with their crowbar that looks like it should break, it had better break. If they walk into a room with other characters, those characters should acknowledge them by at least looking at them, if not calling out their name. Our basic theory was that if the world ignores the player, the player won’t care about the world. A final theory was that the players should always blame themselves for failure. If the game kills them off with no warning, then players blame the game and start to dislike it. But if the game hints that danger is imminent, show players a way out and they die anyway, then they’ll consider it a failure on their part; they’ve let the game down and they need to try a little harder. When they succeed, and the game rewards them with a little treat — scripted sequence, special effect, and so on — they’ll feel good about themselves and about the game."

188 points by thisismyswamp 6y ago

I think the Cabal concept is especially interesting in the wider context of Valve's "self organising" culture. It doesn't seem that this structure was designed, but was emergent.

PicassoCTs 6y ago
This structure is definitely not emergent, if left alone, human groups form little fiefdoms around local "lords", the hierarchy and bureaucracy of a feudal system emerges. To have working anarcho-syndicalism is a constant fight against this mechanism. Who decides how much anyone gets paid, and how many virtual crystal gold smileys someone gets for doing what he currently does? How many golden-smileys do you have to have on your shoulder to influence that decision?
- marcinzm 6y ago
  I remember reading that Valve's internal structure nowadays is lords of fiefdoms even if they communicate externally that it's not. The only difference from other companies is that there's no org chart saying who the lords are, you have to discover that yourself.

walrus01 6y ago

In the context of Cabal and video games, I was expecting something about Destiny 1/2

Netcob 6y ago

That's a refreshingly good example of "design by ...cabal..." actually working out pretty well. I think the key element is how the members actually had to implement their designs. If you just come up with a lazy compromise, it'll start haunting you immediately, and not just after the worker bees did their thing. I think the "cabal" responsible for USB 3.1.1.1.gen2-420gig-whatever should have done the same, with each meeting ending in every member having to explain to their grandma why she can't connect her monitor to her laptop even though the cable fits perfectly, and that it's somehow _both_ the wrong cable and the wrong laptop port.

danbolt 6y ago
I've noticed that happening a lot in my time in the games industry as well. The person making top-down design role isn't exposed to the friction that their decisions are making down the hierarchy. I think the worst part is that they're not even necessarily as invested in the product as someone else, so they're kind of structurally going along with the inefficient situation they're in.

wgjordan 6y ago

(1999)

Blikkentrekker 6y ago

> If they walk into a room with other characters, those characters should acknowledge them by at least looking at them, if not calling out their name. If only Gordon returned this courtesy. Silent protagonists are such an awful thing.

nitrogen 6y ago
You are Gordon. Did you acknowledge the NPCs by looking at them?
- Blikkentrekker 6y ago
  I realize why they do it, but it can never work because the n.p.c.'s do not react to me in return. I can talk to them, say things to them, but they will never respond to anything I say. Thus, the end result is that Gordon Freeman is a man whom many speak to, yet who never speaks back, and no one considers this to odd to say the least. — the conversations thus appear ridiculous. It worked better with Chell, because others do not that Chell does not speak back, and Chell actually canonically refuses to speak out of stubbornness, so it's amusing to see Wheatley being obviously uncomfortable from the fact that Chell does not speak back and is simply blabbering to fill the awkward silence.
  - partlysean 6y ago
    NPCs in Half-Life 2 mentioned a number of times that Gordon doesn’t say much and is a “man of few words.” I’m sure this is a nod to his silence in the first game as much as it’s a storytelling device. I’ve always liked that his character never said anything—feels like Gordon can be anyone. Same goes for the lack of cutscenes that take control of the player’s camera. You always feel like you’re looking from your own POV vs. being pulled out of the immersion.
    - Blikkentrekker 6y ago
      > NPCs in Half-Life 2 mentioned a number of times that Gordon doesn’t say much and is a “man of few words.” I’m sure this is a nod to his silence in the first game as much as it’s a storytelling device. That is an in-joke at best and does very little to rectify that n.p.c.'s have entire “conversations” with him and at no point ask as to why he's remaining quiet throughout it. — the conversations are simply completely implausible. > I’ve always liked that his character never said anything—feels like Gordon can be anyone. Same goes for the lack of cutscenes that take control of the player’s camera. You always feel like you’re looking from your own POV vs. being pulled out of the immersion. Perhaps if one go about his daily life talking to walls and being completely disinterested in what a man might have to reflect back. Whether you reply back to them or not, it always leads to an unnatural conversation where one party is talking to a wall that doesn't respond.

jimbob45 6y ago

“ Include an expert from every functional area (programming, art, and so on). Arguing over an issue that no one at the meeting actually understands is a sure way to waste everyone’s time.” This always sounds good in theory but ends up with unnecessary individuals falling asleep during meetings. I prefer either a staggered approach to meetings where the likely unnecessary people are only invited to the end of the meeting OR are simply “on-hand” for the meeting in case they need to be pulled in. I don’t care how critical the input is, if you pull someone into a three hour meeting and they only contribute five sentences, they’re not going to be happy with you.

kaba0 6y ago
It depends on the program domain, but I think it must have been quite fun designing a game as opposed to the client still doesn’t know what it wants and this API should instead do this.
danielscrubs 6y ago
They where the ones that where going to be responsible for it so how can they be unnecessary? He even said they started to say no more and more as the deadline started to loom over them.
Jare 6y ago
In most significant decisions in gamedev you want art, design and engineering (at least - QA, audio, animation and a number of others as well) to be involved. Some people from one discipline may be versed enough in another to cover the role, but the role must still be present, or there will be pain and work thrown away later on. (incidentally, the ability to understand an issue from these multiple but interrelated angles is what makes experienced people so valuable in gamedev, and their constant loss to other less crazy industries so painful)

ArtWomb 6y ago

Cabal. Design Bible. All grounded in the Single Source of Truth philosophy. With feedback from 200+ playtest sessions directly incorporated into the stream. And the result is the cohesive narrative vision in Half-Life. Contrast this with "What Went Wrong with Cyberpunk 2077" exegesis from Bloomberg this morning and I'm still not sure anyone can point to a single reason beyond "we were too ambitious". I wish this breakdown were more technical in nature. Like a good GamaSutra postmortem. I suspect tooling plays an enormous role. UE4 Blueprints are phenomenal for visual development. But does it scale to teams of hundreds of designers? Inside Cyberpunk 2077's Disastrous Rollout https://www.bloomberg.com/news/articles/2021-01-16/cyberpunk-2077-what-caused-the-video-game-s-disastrous-rollout https://www.bloomberg.com/news/articles/2021-01-16/cyberpunk...

ffhhj 6y ago
It's interesting to me how the physics bugs such as characters and object getting stuck inside walls, flying cars, failed inverse kinematics and so on have been the usual bugs for so many years now, but there isn't yet a methodology to prevent gamedevs from falling in those old known pitholes. These are very difficult to detect by QA and the way to fix them is tweaking some parameters and geometry here and there. There should be a way to supervise the physics engine, find whether future calculations will cause an invalid state on some object and choose a transformation that won't produce such conflict.
- vvanders 6y ago
  As with everything in gamedev it comes down to one thing: performance. Anything that can be faked, approximated our outright hacked in the name of perf or memory will be. The canonical example for me(which was ironically is in Half-Life) is that the AI for the grunts is surprisingly simple. If you throw a grenade they were scripted to shout "grenade" and navigate to the nearest nav node. Tons of people thought there was really complex AI behind that behavior but it was just a couple of really well telegraphed "if()" checks.
  - indigochill 6y ago

jgilias 6y ago

> If the players are in the mood for more action, all they need to do is move forward and within a few seconds something will happen. This! I remember being scared to move around when playing Half-Life the first time. Flickering lights still make an instant association with crowbars to me. Great to read a bit more in detail how they managed to make something so great!

jokethrowaway 6y ago

The Cabal sounds great and I'm looking forward to try it. I'll pick the projects carefully though, as it's impossible to ship more than two products with this methodology.

PeterisP 6y ago

"The meetings were only six hours a day" ... ahhh ... (for five months straight)

solidist 6y ago

from the mod makers pov, circa after they launched in 1999: https://medium.com/super-jump/building-a-popular-half-life-mod-during-the-rise-of-counter-strike-fec6a5b9fd8f?source=friends_link&sk=6d1427b3f1d832df06bd5b07aaa456bb https://medium.com/super-jump/building-a-popular-half-life-m... Mod making was a digital cabal concept, with local leads forming fiefdoms. They were opinionated to a design/delivery. Sometimes straying to other mods, but mostly focusing on one concept at a time.

vvanders 6y ago
Valve also hired really heavily from the mod community. I spent a good portion of my youth in that scene and the people that they hired they really went far for(relocation, visas, etc). I think most of the leads for the more popular mods(outside of CS obviously) ended up there in one form or another if they were interested in it.
pugworthy 6y ago
Nice writeup. I was part of a mod team that got by Valve long time ago, and indeed it was a giddy fun time. We had no idea what we were doing, but we just created a game that we wanted to play - and it seemed to have worked.

TazeTSchnitzel 6y ago

(1999) (if the title gets edited, please delete my comment for me)

alisausaaa 6y ago

Wanna have hot-lovin' conversations? You’re on the right way! - https://adultlove.life https://adultlove.life

nmfisher 6y ago

Fascinating article. One question remains unanswered though - how did Valve afford all of this? Given Half Life was their first game, where was the money coming from to not only produce an entire game, but to then scrap it and redesign/rewrite from scratch?

Danieru 6y ago
Valve was started by two Microsoft Millionaires. Money wasn't an issue.
- nmfisher 6y ago
  Even still, the initial Cabal was three engineers, a writer, an animator and a designer, so you can assume that the broader team was at least that much again. That's 12 people full-time working for 2-3 years (and the actual credits go up to 30, so it's probably even more). At $75k per FTE (in 1996 salaries), that's at least $3 million, which is quite a lot, even for two people wealthy from MSFT stock options.
  - enneff 6y ago
    Gabe Newell was in a senior role at MS for 13 years before he left to start Valve. That’s a lot of money.
  - baud147258 6y ago
    Half-Life was published by Sierra, presumably Valve received some money from them.

kevlar1818 6y ago

TLDR: Small, cross-functional teams build better products.

alisause 6y ago

I was a really bad girl. Punish me with your dick in my mouth. - https://adultlove.life https://adultlove.life

alisaus3 6y ago

Top burny busty chicks only on this site! Follow the link, and you won’t be sorry! - https://adultlove.life https://adultlove.life

anthk 6y ago

On game scripts (literature, not programming), a lot of ideas were shaped before in text adventures/interactive fiction. ZMachine-Inform/Tads have several games based on sci-fi and weird tech environments, such as Babel, or Spider and Web. And the reverse happens, too. H.P. Lovecraft->Alone In the Dark->Anchorhead.

peter_d_sherman 6y ago

>"The first theory we came up with was the theory of "experiential density" — the amount of "things" that happen to and are done by the player per unit of time and area of a map. Our goal was that, once active, the player never had to wait too long before the next stimulus, be it monster, special effect, plot point, action sequence, and so on. Since we couldn’t really bring all these experiences to the player (a relentless series of them would just get tedious), all content is distance based, not time based, and no activities are started outside the player’s control. If the players are in the mood for more action, all they need to do is move forward and within a few seconds something will happen. The second theory we came up with is the theory of player acknowledgment. This means that the game world must acknowledge players every time they perform an action. For example, if they shoot their gun, the world needs to acknowledge it with something more permanent than just a sound — there should be some visual evidence that they’ve just fired their gun. We would have liked to put a hole through the wall, but for technical and game flow reasons we really couldn’t do it. Instead we decided on "decals" — bullet nicks and explosion marks on all the surfaces, which serve as permanent records of the action. This also means that if the player pushes on something that should be pushable, the object shouldn’t ignore them, it should move. If they whack on something with their crowbar that looks like it should break, it had better break. If they walk into a room with other characters, those characters should acknowledge them by at least looking at them, if not calling out their name. Our basic theory was that if the world ignores the player, the player won’t care about the world. A final theory was that the players should always blame themselves for failure. If the game kills them off with no warning, then players blame the game and start to dislike it. But if the game hints that danger is imminent, show players a way out and they die anyway, then they’ll consider it a failure on their part; they’ve let the game down and they need to try a little harder. When they succeed, and the game rewards them with a little treat — scripted sequence, special effect, and so on — they’ll feel good about themselves and about the game."