WEBVTT - The Story of Valve Part Two 0:00:04.519 --> 0:00:13.119 Technology with tech Stuff from dot Com. Hey there, everybody, 0:00:13.119 --> 0:00:17.720 and welcome to tech Stuff. I am your host, Jonathan Strickland, 0:00:18.120 --> 0:00:22.000 here to talk about all things technological, well not all 0:00:22.079 --> 0:00:25.680 things today, because we're going to continue our story from 0:00:25.680 --> 0:00:29.400 the last episode, the story of Valve. Now, in that 0:00:29.520 --> 0:00:33.240 last episode, we looked at how to Microsoft millionaires left 0:00:33.280 --> 0:00:36.680 their stable gigs to launch a new game development company, 0:00:36.720 --> 0:00:40.400 and they called it Valve Software LLC. At that time, 0:00:40.920 --> 0:00:43.920 they use their own money to fund this company, and 0:00:43.960 --> 0:00:46.040 they hired on talent to help them come up with 0:00:46.080 --> 0:00:49.000 a new type of game, which was a story driven 0:00:49.080 --> 0:00:53.320 first person shooter. It would take them two years and 0:00:53.440 --> 0:00:56.160 the result of that work would be the game Half Life. 0:00:56.760 --> 0:00:59.639 A couple of years later, one of those two co founders, 0:00:59.760 --> 0:01:04.320 Might Harrington, would dissolve his partnership, with the other co founder, 0:01:04.400 --> 0:01:07.600 Gabe Newell, taking on charge of the company as the 0:01:07.680 --> 0:01:11.840 sole leader as it were, and leaving him a little 0:01:11.840 --> 0:01:15.559 bit lonely. So. Valve at this point had finished Half Life, 0:01:15.600 --> 0:01:19.120 which was enjoying critical acclaim as well as healthy sales. 0:01:19.440 --> 0:01:22.039 Gamers were eating it up. But as I covered in 0:01:22.080 --> 0:01:26.040 that previous episode, it didn't exactly go smoothly at the time. 0:01:26.120 --> 0:01:30.200 The development took a long while. It did, however, help 0:01:30.319 --> 0:01:34.800 establish Valves methodology for developing games. That is that within 0:01:34.840 --> 0:01:39.360 the company, employees would form cabals. Now that all sounds 0:01:39.440 --> 0:01:43.679 rather sinister, doesn't it, But the cabal structure was necessity 0:01:43.720 --> 0:01:46.840 while working on Half Life, which I should remind you 0:01:47.080 --> 0:01:49.760 was a project that Valve hit the reset button on 0:01:50.040 --> 0:01:52.640 when the original work was nearing completion. They were getting 0:01:52.640 --> 0:01:56.160 close to their original ship date and made the determination 0:01:56.200 --> 0:01:59.480 that the game just wasn't working. It wasn't fun. They 0:01:59.560 --> 0:02:02.480 developed lot of cool technology, but they had not yet 0:02:02.560 --> 0:02:05.280 figured out how to use it properly to make a 0:02:05.360 --> 0:02:10.720 really compelling game. So they wanted to actually take all 0:02:10.760 --> 0:02:13.480 that stuff they learned and turn it into a much 0:02:13.600 --> 0:02:16.920 better product. First, they thought, well, what we should do 0:02:16.960 --> 0:02:19.960 is get a game designer, someone who's brilliant in this field, 0:02:20.000 --> 0:02:24.079 like a leader, who can sweep in and rescue the project. 0:02:25.000 --> 0:02:28.320 The way Ken Birdwell put it in a blog post 0:02:28.440 --> 0:02:31.240 was that the company was looking for a person who 0:02:31.280 --> 0:02:34.160 would swing in on a on a chord, you know, 0:02:34.840 --> 0:02:38.960 sword in hand, hair swept with a wind. He or 0:02:39.000 --> 0:02:41.720 she would pick up all the disparate pieces and development 0:02:41.919 --> 0:02:44.760 and magically assemble them into the most awesome game ever. 0:02:45.440 --> 0:02:48.480 But it became clear that this was a pipe dream. 0:02:48.600 --> 0:02:52.480 There was no hero coming to the rescue. They just 0:02:52.720 --> 0:02:56.000 they couldn't find the right person who could actually make 0:02:56.040 --> 0:03:00.560 this happen. So instead, Valve employees within the company already 0:03:00.600 --> 0:03:05.400 began to form groups within that organization. They decided, well, 0:03:05.440 --> 0:03:08.359 we don't really have one person who can lead this 0:03:08.440 --> 0:03:12.920 project and turn it into something that is really compelling, 0:03:13.320 --> 0:03:15.560 but we've got a lot of talent here. Maybe we 0:03:15.600 --> 0:03:19.320 get a group of us, a multidisciplinary group, so people 0:03:19.320 --> 0:03:23.000 who are specialists in different fields to work together in 0:03:23.080 --> 0:03:28.200 a collaborative process. We can achieve what one magical superhero 0:03:28.240 --> 0:03:30.480 would be able to do. And the members of that 0:03:30.520 --> 0:03:33.520 group had a really wide range of skills that complemented 0:03:33.560 --> 0:03:36.440 one another. They would have meetings that contained a lot 0:03:36.480 --> 0:03:38.840 of brainstorming, and in those meetings the group would do 0:03:39.000 --> 0:03:42.440 a ton of work. In fact, let me quote Mr 0:03:42.480 --> 0:03:45.560 Birdwell to give you a clue about how cabals would 0:03:45.720 --> 0:03:49.600 actually get stuff done. Quote. The goal of this group 0:03:49.720 --> 0:03:52.760 was to create a complete document that detailed all the 0:03:52.840 --> 0:03:58.000 levels and described major monster interactions, special effects, plot devices, 0:03:58.240 --> 0:04:01.760 and design standards of Cabal was to work out when 0:04:02.080 --> 0:04:05.800 and how every monster, weapon and n PC was to 0:04:05.840 --> 0:04:08.840 be introduced, what skills we expected the player to have, 0:04:09.240 --> 0:04:11.840 and how we were going to teach them those skills 0:04:12.080 --> 0:04:16.600 end quote. That's that's a huge burden. You know, you're 0:04:16.640 --> 0:04:20.400 talking about every element of game design, from the mechanics 0:04:20.960 --> 0:04:23.760 as in the mechanics within the game itself, not to 0:04:23.800 --> 0:04:26.760 mention just how do those translate to what the player 0:04:26.839 --> 0:04:28.400 needs to do in order to pull it off? Like 0:04:29.400 --> 0:04:31.800 what button do you have to push in order to strafe? 0:04:31.800 --> 0:04:34.120 What button do you push to interact with your environment? 0:04:34.440 --> 0:04:37.400 And then all the elements that happen inside the game, 0:04:37.480 --> 0:04:40.320 all the stuff you encounter, all the things that make 0:04:40.360 --> 0:04:43.839 the game what it is. According to Birdwell, this process 0:04:44.240 --> 0:04:47.479 was intense and exhausting, and it would also include a 0:04:47.520 --> 0:04:51.000 second stage. Essentially, once the team had enough raw material 0:04:51.040 --> 0:04:53.880 material to work with, you know, they had brainstorm, they decide, okay, 0:04:54.120 --> 0:04:56.920 we're gonna have a level. We know that the level 0:04:56.960 --> 0:05:00.880 is going to culminate with an encounter enter with this 0:05:00.920 --> 0:05:04.920 particular type of enemy uh that has these weaknesses and 0:05:04.960 --> 0:05:08.080 nise strengths. But leading up to that, we want them 0:05:08.120 --> 0:05:11.279 to have these four or five other experiences. Plus we 0:05:11.320 --> 0:05:14.480 have to make sure that the player has the opportunity 0:05:14.520 --> 0:05:18.800 to learn the strategy that will ultimately benefit them in 0:05:18.839 --> 0:05:21.880 the battle with this big boss at the end of 0:05:21.920 --> 0:05:24.440 the level. They come up with these various ideas, they 0:05:24.440 --> 0:05:26.800 would start hashing them out and they would get really detailed. 0:05:26.839 --> 0:05:30.640 It wouldn't just be these general rules. It might be 0:05:30.760 --> 0:05:33.800 much more specific, and they would start to take all 0:05:33.839 --> 0:05:37.160 of these different disparate pieces and shape it into a narrative. 0:05:38.080 --> 0:05:41.200 They decided that the stuff that happened within the game 0:05:41.360 --> 0:05:45.800 should be location based rather than time based. So in 0:05:45.800 --> 0:05:49.159 other words, you wouldn't need to worry about missing out 0:05:49.320 --> 0:05:53.120 on a really cool story development because you were lollygagging 0:05:53.240 --> 0:05:56.440 two rooms back for too long. So you were, you know, 0:05:56.680 --> 0:05:59.280 checking out this one corner because you were convinced that 0:05:59.279 --> 0:06:01.200 that corner was going to lead you to a super 0:06:01.240 --> 0:06:03.640 secret special area. And it turns out no, it's just 0:06:03.720 --> 0:06:06.799 a corner. So you go on with your merry little life. 0:06:06.920 --> 0:06:09.280 But you missed all the great stuff because you were 0:06:09.360 --> 0:06:11.200 too busy with the corner, and the great stuff was 0:06:11.200 --> 0:06:13.480 on a timer. They did not want that to happen, 0:06:14.000 --> 0:06:17.600 so instead they decided that once you reach the appropriate 0:06:17.640 --> 0:06:21.880 location within any given level. Your arrival is what would 0:06:21.880 --> 0:06:25.360 trigger whatever the event was supposed to be, and those 0:06:25.400 --> 0:06:28.000 events could be plot developments. It could be an encounter 0:06:28.120 --> 0:06:30.359 with a monster. It could be the discovery of a 0:06:30.360 --> 0:06:34.440 new weapon. It could be uh the interaction with another character. 0:06:34.440 --> 0:06:37.520 It could be a specific special effect, whatever that might be. Now, 0:06:37.560 --> 0:06:40.760 occasionally the group would come up with ideas for really 0:06:40.800 --> 0:06:44.200 cool stuff that didn't necessarily fit in with anything else. 0:06:44.240 --> 0:06:46.840 It was kind of just a neat idea all on 0:06:46.839 --> 0:06:49.960 its own, and sometimes they would use those as transitions 0:06:50.040 --> 0:06:52.320 between areas. They'd say, well, it doesn't really fit in 0:06:52.360 --> 0:06:54.800 this level design, and it doesn't really fit in this 0:06:55.040 --> 0:06:57.480 next level design, but we could have it as something 0:06:57.520 --> 0:07:00.560 in between those two. Or some times they would just 0:07:00.600 --> 0:07:02.400 have to toss it out if they couldn't find a 0:07:02.440 --> 0:07:05.839 way for it to work within the confines of their story. 0:07:06.320 --> 0:07:08.520 They also found it helpful to occasionally just put a 0:07:08.560 --> 0:07:12.840 whole bunch of unrelated ideas into the same area and 0:07:12.880 --> 0:07:15.520 then try to figure out how they would make them 0:07:15.560 --> 0:07:19.760 all fit logically. So you might have a couple of 0:07:19.800 --> 0:07:24.400 monster designs and a new weapon, and maybe a character interaction, 0:07:24.400 --> 0:07:26.880 and all of these ideas you've had separately, and you thought, well, 0:07:26.880 --> 0:07:29.400 why don't we just throw them into this one level, 0:07:29.640 --> 0:07:31.400 see if we can make it make sense from a 0:07:31.400 --> 0:07:34.320 story aspect, and then that can help lead to some 0:07:34.360 --> 0:07:38.840 more creativity. Uh. Sometimes that didn't work. Sometimes they would 0:07:39.080 --> 0:07:42.320 end up incorporating a lot of ideas, but ultimately they 0:07:42.360 --> 0:07:45.000 would discover when they were done that whatever it was 0:07:45.080 --> 0:07:47.920 that launched them down that path was no longer part 0:07:47.960 --> 0:07:50.040 of the level. So let's say that you have this 0:07:50.120 --> 0:07:52.920 one monster design you really like, and you have decided 0:07:52.960 --> 0:07:54.960 that you're going to try and create a level around it. 0:07:54.960 --> 0:07:57.440 It might turn out that months later you've got a 0:07:57.520 --> 0:08:00.560 great level, but that monsters nowhere to be found. You 0:08:00.920 --> 0:08:04.800 don't even use it, they said. Ultimately they realized, you know, 0:08:04.840 --> 0:08:09.160 it didn't matter what the the genesis of the idea 0:08:09.400 --> 0:08:13.480 was so much that the finished idea worked for the game. 0:08:14.240 --> 0:08:16.720 And once they took that philosophy, they realized that the 0:08:16.760 --> 0:08:24.640 development process, while very exhausting, worked much much better now. 0:08:24.680 --> 0:08:27.720 At that time, Birdwell says, cabal sessions would include five 0:08:27.800 --> 0:08:30.600 or six employees, and not everyone would always have something 0:08:30.640 --> 0:08:34.000 to contribute every single day. So if you didn't have 0:08:34.320 --> 0:08:36.600 five or six, you could end up with a meeting 0:08:36.640 --> 0:08:39.880 where no one's really talking, nothing's really getting done. So 0:08:39.960 --> 0:08:42.440 they needed five or six just to kind of cover 0:08:42.480 --> 0:08:46.320 all the bases. Make sure that even if you're strapped 0:08:46.320 --> 0:08:49.640 for ideas that week, other people in the group have 0:08:49.760 --> 0:08:52.640 stuff to contribute, and it may be that three weeks 0:08:52.679 --> 0:08:56.280 later you're the person who's taking the lead because something 0:08:56.360 --> 0:08:58.680 someone else has done has sparked an idea in your 0:08:58.720 --> 0:09:00.960 head and you're able to run with it. And that's 0:09:00.960 --> 0:09:04.120 exactly what they wanted. So the cabal group for Half 0:09:04.160 --> 0:09:09.520 Life consisted of a single writer, a level designer, and animator, 0:09:09.760 --> 0:09:13.680 and then three engineers. Everyone who was in that cabal 0:09:13.760 --> 0:09:18.000 had previously worked on a product somewhere that actually shipped, 0:09:18.280 --> 0:09:20.760 which was a pre requirement of that first cabal. They 0:09:20.800 --> 0:09:24.360 wanted the group to only include people who had worked 0:09:24.360 --> 0:09:27.160 on products that actually saw the light of day. Now, 0:09:27.200 --> 0:09:29.520 that didn't mean that they had done that in game 0:09:29.559 --> 0:09:33.480 development necessarily. They might have developed software they got published 0:09:33.559 --> 0:09:37.000 somewhere else that wasn't even a game, but at least 0:09:37.000 --> 0:09:41.480 they had worked on something that published. The group would 0:09:41.559 --> 0:09:44.800 meet four days a week, and the meetings would last 0:09:44.840 --> 0:09:48.320 six hours a day, and this went on for five 0:09:48.520 --> 0:09:52.360 months when they first started the cabal system. After that, 0:09:52.400 --> 0:09:56.319 they would occasionally meet as the game continued to develop 0:09:56.440 --> 0:09:59.160 until it was finally shipped, but they weren't meeting every 0:09:59.160 --> 0:10:02.839 single day or even four days a week. Once they 0:10:02.880 --> 0:10:06.440 got through that initial brainstorming process, and according to Birdwell, 0:10:06.840 --> 0:10:10.160 the experience was so draining that you couldn't expect the 0:10:10.320 --> 0:10:12.720 people who were in the cabal to do any other 0:10:12.840 --> 0:10:17.520 work because it was just so consuming. The most you 0:10:17.600 --> 0:10:21.400 might manage is some light housekeeping stuff like answering emails. 0:10:21.400 --> 0:10:24.480 That's about as strenuous as you could get once you 0:10:24.520 --> 0:10:28.760 went through this system. Now, at the end of this process, 0:10:28.800 --> 0:10:34.400 the cabal produced a two hundred page document about Half Life. 0:10:34.760 --> 0:10:38.319 This became the guiding design behind the game. More importantly, 0:10:38.760 --> 0:10:43.360 this style of collaboration became synonymous with valves operating procedure. 0:10:43.720 --> 0:10:47.160 Cabals became the model for game development and really all 0:10:47.400 --> 0:10:52.439 projects with involved, not just game development, software development tools, patches, 0:10:52.720 --> 0:10:56.120 all sorts of stuff. All of it is centered around 0:10:56.120 --> 0:11:00.280 this concept of cabals. And it all started with Half Life. 0:11:00.720 --> 0:11:04.160 Once there was enough of Half Life to actually test 0:11:04.400 --> 0:11:08.280 Sierra because remember Sierra was the publisher for Half Life 0:11:08.640 --> 0:11:12.600 pulled in play testers to try stuff out. During those 0:11:12.600 --> 0:11:16.400 sessions that people responsible for that level would observe the 0:11:16.440 --> 0:11:20.040 players and they would take notes. So that had to 0:11:20.040 --> 0:11:23.480 be a harrowing experience to have designed something and now 0:11:23.520 --> 0:11:26.199 you're watching a stranger play through it for the first time, 0:11:26.600 --> 0:11:29.160 and you're thinking, oh, they're gonna love that thing that's 0:11:29.200 --> 0:11:30.960 around the corner, and then it turns out they never 0:11:31.000 --> 0:11:35.480 even went around the corner. That's gotta be a sobering experience. 0:11:35.760 --> 0:11:37.920 It's also how they found out if a puzzle they 0:11:37.960 --> 0:11:40.960 designed was too easy and people just breathe through it, 0:11:41.040 --> 0:11:43.760 or maybe they found out it was impossible that without 0:11:44.200 --> 0:11:47.000 the fore knowledge that there was a puzzle and how 0:11:47.040 --> 0:11:49.360 the puzzle worked, there's no real way to work it out. 0:11:50.280 --> 0:11:52.199 And it also just showed them if there were any 0:11:52.240 --> 0:11:56.240 elements that just weren't fun to play, and that would 0:11:56.240 --> 0:11:59.560 allow them to go back and refine their design for 0:11:59.600 --> 0:12:03.360 the next build of the game. Now, after Half Life 0:12:03.400 --> 0:12:07.800 was done, the cabal process continued, but it did change 0:12:07.840 --> 0:12:11.360 a bit. Birdwell says that for Team Fortress too, For example, 0:12:11.679 --> 0:12:15.440 cabals tended to include between eight and twelve people, and 0:12:15.480 --> 0:12:17.920 the meetings weren't nearly as long they were a couple 0:12:17.920 --> 0:12:21.600 of hours, not six hours a day. But the collaborative 0:12:21.640 --> 0:12:25.680 process became an important element for Valve, so much so that, 0:12:25.760 --> 0:12:28.720 according to the two thousand twelve employee Handbook, which by 0:12:28.760 --> 0:12:31.320 the way, you can actually find online and it is 0:12:31.520 --> 0:12:36.120 a pretty entertaining read, at that point, all desks at 0:12:36.240 --> 0:12:40.200 Valve were mounted on wheels that allowed him to move 0:12:40.240 --> 0:12:43.480 around the office. Even when the office got large enough 0:12:43.520 --> 0:12:47.400 to encompass multiple floors, there were freight elevators where you 0:12:47.440 --> 0:12:49.880 could wheel your desk and take it to a different floor. 0:12:49.960 --> 0:12:53.640 This would make it easier to form teams within Valve 0:12:54.160 --> 0:12:57.240 and work close by, you know, shoulder to shoulder with 0:12:57.280 --> 0:13:01.319 your teammates for as long as whatever the team project 0:13:01.520 --> 0:13:04.319 was took, and then once it was over, you could 0:13:04.760 --> 0:13:07.480 actually wheel your desk somewhere else and join a different 0:13:07.520 --> 0:13:11.320 team or form a team yourself. In fact, rather than 0:13:11.360 --> 0:13:14.000 just jumped down the chronology, it might be a good 0:13:14.040 --> 0:13:17.760 time to talk about the actual corporate culture of Valve, 0:13:17.840 --> 0:13:20.760 at least circa two thousand twelve, because that was the 0:13:20.840 --> 0:13:24.040 latest employee handbook I have at my disposal. It may 0:13:24.080 --> 0:13:26.520 be that things have changed since then, but the two 0:13:26.520 --> 0:13:29.080 thousand twelve valve. I imagine it was not too different 0:13:29.120 --> 0:13:31.560 from what today's valve is, But this gives you a 0:13:31.600 --> 0:13:37.040 way of imagining how things work inside the company itself. 0:13:37.440 --> 0:13:41.360 The two thousand twelve handbook refers to valves organizational structure 0:13:41.480 --> 0:13:46.320 as flat land, so they don't have a strict hierarchy 0:13:46.360 --> 0:13:50.280 of positions in which directors report to a president and 0:13:50.320 --> 0:13:53.480 then manager's report to the directors and employees report to 0:13:53.559 --> 0:13:57.959 managers or they're overseen by managers. Instead, it's at least, 0:13:57.960 --> 0:14:01.520 in theory, a flat line across the company, with all 0:14:01.600 --> 0:14:05.800 employees being equal. The handbook does say that Gabe Newell 0:14:06.000 --> 0:14:09.200 is the founder slash president, but he it also says 0:14:09.200 --> 0:14:12.120 that he is not quote your manager in the quote, 0:14:12.559 --> 0:14:14.839 so every employee in theory at least has the power 0:14:14.880 --> 0:14:19.920 to green light and ship products. Uh. Of course, the 0:14:20.280 --> 0:14:23.240 manual also says that out of all the people who 0:14:23.280 --> 0:14:27.400 are not your manager, Gabe Newell is the most not 0:14:28.000 --> 0:14:32.840 your manager, and it almost sounds like a double speak 0:14:32.960 --> 0:14:35.840 sort of thing, like he's not your manager, but he 0:14:35.920 --> 0:14:39.360 runs the company. M of course, that could just be 0:14:39.400 --> 0:14:44.320 my interpretation of the manual. So how does this corporate 0:14:44.320 --> 0:14:47.000 culture work in practice. Well. I have not been to 0:14:47.080 --> 0:14:51.320 Valve personally, and they are traditionally a pretty secretive company. 0:14:51.640 --> 0:14:55.720 They've had a couple of different UH folks visit and 0:14:55.880 --> 0:14:59.600 document their visits to the office, but it tends to 0:14:59.600 --> 0:15:03.760 be really controlled and usually there are multiple floors that 0:15:03.800 --> 0:15:07.080 people have no access to because there are top secret 0:15:07.120 --> 0:15:11.440 projects going on in those floors. But it sounds to 0:15:11.480 --> 0:15:15.440 me like the corporate culture Valve would be a rewarding 0:15:15.480 --> 0:15:19.720 but extremely challenging environment for a lot of people because 0:15:19.760 --> 0:15:22.320 instead of taking direction, instead of having a boss who says, 0:15:23.160 --> 0:15:26.640 this is what your job is, these are your responsibilities. 0:15:26.840 --> 0:15:29.600 Here is how we expect you to perform your duties, 0:15:29.960 --> 0:15:34.080 and we will evaluate you based on that criteria. Employees 0:15:34.120 --> 0:15:37.400 at Valve are expected to seek out projects that appeal 0:15:37.480 --> 0:15:40.800 to them and then essentially glam onto them and contribute. 0:15:40.840 --> 0:15:43.520 So you would look around at your fellow employees and say, hey, 0:15:43.560 --> 0:15:45.800 what are you working on? Oh that sounds cool, I 0:15:45.800 --> 0:15:47.920 would like to work on that too. I'm part of 0:15:47.920 --> 0:15:51.280 your team now, Or you could say what are you 0:15:51.320 --> 0:15:53.040 working on? What are you working on? You know what? 0:15:53.160 --> 0:15:55.000 It sounds to me like no one is working on 0:15:55.680 --> 0:15:58.640 this amazing idea I have so I am going to 0:15:58.760 --> 0:16:01.200 launch this as a project, and I'm going to start 0:16:01.240 --> 0:16:04.680 recruiting people to work on my own team so that 0:16:04.720 --> 0:16:07.720 we can get this done. Uh, these projects can be 0:16:08.400 --> 0:16:11.120 just about anything. There seems to be a huge number 0:16:11.160 --> 0:16:14.000 of options to choose from, and for some people that 0:16:14.120 --> 0:16:16.840 could be a little bit intimidating. I know for me 0:16:16.920 --> 0:16:19.200 it would be extremely intimidating. I mean, if you just 0:16:19.240 --> 0:16:22.600 take me to the cheesecake factory and watches I panic 0:16:22.920 --> 0:16:26.200 while flipping through that fifty page menu, you'd understand, like 0:16:26.680 --> 0:16:33.800 sometimes having unlimited options is more restrictive than it is freeing. 0:16:34.720 --> 0:16:38.200 There's actually a whole paradox about that. But if you 0:16:38.320 --> 0:16:41.200 have the aptitude, if you're the person who can thrive 0:16:41.200 --> 0:16:44.960 in that environment, you can, in theory anyway, work on 0:16:45.040 --> 0:16:48.920 any project inside Valve. The philosophy of the company is 0:16:48.920 --> 0:16:51.400 that if you're good enough to be hired, you're already 0:16:51.400 --> 0:16:54.320 on the right path. And in fact, they are very 0:16:54.360 --> 0:16:57.000 careful in the hiring process. They don't want to hire 0:16:57.040 --> 0:17:01.520 someone who is either just a generalist who is pretty 0:17:01.560 --> 0:17:04.159 good at lots of stuff but doesn't have a deep 0:17:04.160 --> 0:17:06.879 expertise in anything, Nor do they want to get someone 0:17:06.880 --> 0:17:10.480 who has an exclusive expertise in one subject and no 0:17:11.680 --> 0:17:14.720 interest or affinity for any other subject. They want to 0:17:14.720 --> 0:17:18.159 find people who have a broad range of skills and 0:17:18.240 --> 0:17:22.520 a narrow range of deep expertise. That's the sweet spot 0:17:22.600 --> 0:17:26.760 for Valve employees. And honestly, they say that they want 0:17:27.280 --> 0:17:30.520 people who can make senior level decisions because in a 0:17:30.600 --> 0:17:34.240 flat hierarchy, where you don't have a boss to report to, 0:17:34.960 --> 0:17:37.920 you are the one making these senior decisions. You might 0:17:38.080 --> 0:17:40.560 end up contributing to one project, but you could end 0:17:40.640 --> 0:17:45.600 up effectively leading another one. Now, within work groups, some 0:17:45.720 --> 0:17:50.040 hierarchies can and do emerge, but they are informal. They're 0:17:50.080 --> 0:17:55.639 not codified in the organizational chart of Valve. Typically, someone 0:17:55.680 --> 0:17:57.840 on a project will end up being a team lead, 0:17:58.080 --> 0:18:01.240 though that person is more often a repository of all 0:18:01.240 --> 0:18:04.199 the information about the project instead of being like a 0:18:04.240 --> 0:18:08.160 traditional manager. And while a group is working toward a goal, 0:18:08.359 --> 0:18:12.800 certain roles and relationships might form because just on necessity 0:18:12.840 --> 0:18:15.959 to make sure you get stuff done uh, and that 0:18:16.080 --> 0:18:19.520 might look more like a traditional company's structure, at least 0:18:19.520 --> 0:18:23.159 temporarily within the confines of that project. But all of 0:18:23.160 --> 0:18:25.520 that will fade away once the goal is complete or 0:18:25.560 --> 0:18:29.280 the project is concluded now. According to the handbook, the 0:18:29.359 --> 0:18:32.600 company also wanted to avoid the problems of crunch time 0:18:33.480 --> 0:18:35.600 because they had gone through that a couple of times 0:18:35.640 --> 0:18:38.080 by the by the year two thousand twelve when this 0:18:38.119 --> 0:18:41.960 handbook was written. Crunch is when you're coming up against 0:18:41.960 --> 0:18:45.399 a deadline and you need to make the mountain of 0:18:45.520 --> 0:18:50.320 work between you and the deadline go away. So typically 0:18:50.320 --> 0:18:53.479 this means working very long hours with an increased amount 0:18:53.600 --> 0:18:56.280 of stress. And if you've ever done a crash study 0:18:56.320 --> 0:18:59.240 course before a test, you know you're just cramming before 0:18:59.240 --> 0:19:01.600 a test. You get an idea of what this is like, 0:19:01.720 --> 0:19:06.480 only crunch is way way worse. Valves Handbook says that 0:19:06.520 --> 0:19:09.439 employee work life balance is really important and that for 0:19:09.480 --> 0:19:12.960 that reason you should only work normal office hours, and 0:19:13.000 --> 0:19:16.440 employees also have no holiday or sick leave policies, meaning 0:19:16.520 --> 0:19:18.440 that you can take as much time as you need 0:19:18.800 --> 0:19:21.200 at least in theory, as long as you're getting your 0:19:21.200 --> 0:19:24.800 work done. Whether this actually happens at the company or not, 0:19:24.880 --> 0:19:27.240 I can't say, but I can imagine why it was 0:19:27.280 --> 0:19:30.320 added in, because, after all, one of the co founders 0:19:30.400 --> 0:19:33.520 a Valve left after half life shipped and part of 0:19:33.520 --> 0:19:36.000 the reason for his leaving was because he had poured 0:19:36.200 --> 0:19:38.640 so much of his time and energy and effort into 0:19:38.640 --> 0:19:40.760 getting the game out the door that it had had 0:19:40.960 --> 0:19:44.040 a real impact on his life at that point, and 0:19:44.080 --> 0:19:47.040 no one wants to drive their talent away. So they 0:19:47.119 --> 0:19:49.639 made a formal policy saying, you know, you should just 0:19:49.680 --> 0:19:52.639 be working your office hours, and anything that's causing you 0:19:52.680 --> 0:19:55.480 to work longer hours than normal, maybe we need to 0:19:55.480 --> 0:19:57.879 take a look at it and make sure that you 0:19:57.920 --> 0:19:59.679 know the priorities are in the right spot so that 0:19:59.720 --> 0:20:03.480 you can actually have your personal time. Again, I don't 0:20:03.520 --> 0:20:05.760 know if that's how it actually shakes out in the company. 0:20:05.800 --> 0:20:08.119 It's one thing to say it in the employee handbook. 0:20:08.119 --> 0:20:10.240 It's another thing to see it play out in the 0:20:10.320 --> 0:20:13.560 actual day to day office life. But since I don't 0:20:13.560 --> 0:20:17.800 know anyone at Valve, I can't really get insider information 0:20:17.840 --> 0:20:20.199 on that. If you do work for Valve or have 0:20:20.320 --> 0:20:23.320 worked for Valve, and you have some insight into this 0:20:23.400 --> 0:20:25.680 and whether or not that in fact is the way 0:20:25.680 --> 0:20:28.760 it works within the company, or maybe it's the ideal, 0:20:29.200 --> 0:20:31.000 but that's not how things turn out, let me know, 0:20:31.200 --> 0:20:34.439 because I can tell you as a creative person on 0:20:34.480 --> 0:20:37.159 my end, most of the people I know at this 0:20:37.240 --> 0:20:41.240 company are working very long weeks, and just because you're 0:20:41.880 --> 0:20:44.640 office hours are over doesn't mean you are done working. 0:20:44.680 --> 0:20:46.720 You might be working quite a bit more, even at 0:20:46.720 --> 0:20:51.399 home or on vacation. As it turns out anyway, peer 0:20:51.400 --> 0:20:54.520 review would become the basis of employee reviews at Valve, 0:20:54.680 --> 0:20:58.119 And because there's no real hierarchy, there's no manager director 0:20:58.160 --> 0:21:01.800 that you're reporting to, you would have to rely upon 0:21:01.880 --> 0:21:05.280 your fellow employees to describe your work, and they would 0:21:05.280 --> 0:21:07.479 describe it to some people who would essentially be kind 0:21:07.520 --> 0:21:11.000 of chosen each year to be responsible for interviewing employees 0:21:11.040 --> 0:21:14.280 to find out how everyone is doing job performance wise, 0:21:15.520 --> 0:21:19.119 and compensation decisions would rest on peer descriptions of an 0:21:19.160 --> 0:21:23.119 employee's technical ability, their ability to work within a group 0:21:23.240 --> 0:21:27.320 and contribute valuable input, how much the employee actually contributed 0:21:27.320 --> 0:21:30.400 to products and development, and the employee's general output level. 0:21:30.480 --> 0:21:33.640 So you would end up having, like if you work 0:21:33.680 --> 0:21:38.000 for Valve, then your coworkers that you are in close 0:21:38.040 --> 0:21:41.120 contact with because you're all working together on a specific project, 0:21:41.440 --> 0:21:44.280 they would get interviewed by some of these employees and 0:21:44.320 --> 0:21:47.440 they would give answers like, hey, Jonathan over there, he's 0:21:47.480 --> 0:21:50.960 on your team. Uh, what's he like to work with? 0:21:51.720 --> 0:21:57.320 Did he do anything really uh? Notable in context of 0:21:57.320 --> 0:22:02.120 this particular product, you know, were his contributions meaningful? Did 0:22:02.160 --> 0:22:04.639 they help make the product better? Did they help it 0:22:04.680 --> 0:22:07.480 ship on time? All those sort of questions. The answers 0:22:07.480 --> 0:22:11.600 to that are what informs the decision to increase your 0:22:11.640 --> 0:22:15.800 compensation or not. As it turns out, and Valve really 0:22:15.840 --> 0:22:21.440 wants to reward people based upon their value to the company, 0:22:21.480 --> 0:22:24.480 but not over reward people, so it's a delicate balance. 0:22:24.520 --> 0:22:29.040 They also pride themselves on being more generous with their 0:22:29.080 --> 0:22:33.960 compensation than some other companies like Microsoft and Google. But 0:22:34.640 --> 0:22:38.680 this is important because there's no there's no career ladder, right, 0:22:38.760 --> 0:22:42.040 there's no hierarchy, so you can't be promoted. In fact, 0:22:42.119 --> 0:22:46.119 most Valve employees don't really have a position uh not 0:22:46.119 --> 0:22:49.200 not really not not a real position or even job 0:22:49.240 --> 0:22:55.320 description apart from help Valve do stuff. Some Valve employees 0:22:55.560 --> 0:22:59.400 have an effective job title because if you're ever interacting 0:22:59.440 --> 0:23:02.439 with anyone outside of Valve, typically you need to have 0:23:02.520 --> 0:23:04.560 some sort of job title to make sense of it, 0:23:04.720 --> 0:23:07.600 because anyone who's chatting with you needs to have a 0:23:07.680 --> 0:23:11.320 point of reference. So people who work at Valve who 0:23:11.359 --> 0:23:15.600 have outside facing duties, like let's say it's someone who 0:23:15.760 --> 0:23:19.080