Since it worked badly we just got a worse game than otherwise. All the good stuff in the game came from the talented and creative developers, the structure just held them back.
@@xyhmo Not exactly. If you simulate a traditional vertical structure you'd most likely end up with a schlock version of Arcanum since it would've precluded Tim & Leonard's direct involvement with this project as they would be self-confined to their higher-level roles in the hierarchy. Hence why their "failure" to stay in their lane was the reason the game was anti-schlock (but a failure business wise).
I think two words got confused here: Role and Rank. My team is currently trying out no-hirachies and it is really working for us. But no hierarchies doesn't mean no roles so I'd like to "respond" to the things that went wrong in the video and tell how we are combatting this. Hire only seniors: We are currently 2 Seniors and a bunch of first-timers. We are making progress and no prejudice is in the group. I'll explain later why and how. Estimation: OK, this is the first problem. Estimation is always bad. Because it is always wrong. So we just don't. But how do we make decisions? We measure. We simply do it. measure the time it takes and then extrapolate. The results mostly are correct within a 5 to 10 percent error margin which is good enough for us to make decisions with. Deciding what to do: We have a list of ideas that are roughly ranked by how much we want them. Every time somebody completes a task, they pull the next thing from the list that they can do and start working on it. Working efficiently: Each week we retrospect on what was blocking the most or is currently the slowest part of the pipeline. Then we discuss what the problems are and how we can speed things up. This works on a macro level and we are still accelerating production speeds. Sometimes it was the coding that was blocking, sometimes it was asset creation. Almost every time the solution was a day of automation work or a 10$ plugin. On a person-to-person basis, things get discovered naturally. People notice if others are slow or they are slower than others. We combat it by grouping the fastest with the slowest if this happens (for a day or two). The slowest then has the opportunity to learn from the faster one and learn all the tricks that are needed. Sometimes it was an artist always using the mouse for everything instead of shortcuts, sometimes it was a programmer not knowing the standard library well enough. By always learning from each other, we have managed to reduce the inter-personal time differences beyond noise levels. (Took us a year to get to this point) [measured by time tracking work tasks and running a quick Python script to do statistics on the metadata. Not saying it is perfect evidence, but good enough for us.] Fluid roles: This one is tricky. Letting people just do any task they want is a recipe for chaos. The group has a consensus on who is good with that. So when somebody wants to do something that they can't do, they usually go to that person and ask for help. If they want to do a full-on switch (Programming -> UI) then we cultivate a large resource library with all the useful learning resources that the previous experts used to get to the point where they are now. Of course, this doesn't eliminate the need for practice, but you would be surprised how good programmers can make GUIs if they just read one or two books first. I have to admit this apprentice and resource system is not perfect and doesn't allow for complete switches like musik-> animation, but it reduces roadblocks in development effectively enough for us. An animator doing the UI task that's needed or a programmer doing the audio level mixing inside the game is achievable and liberating in our experience. Conflict resolution: If somebody is insisting that "that thing would be so great" then we just let them make the lowest amount of effort that can validate their point. Example: Last week the 3D artist argued that we should switch the colors up for drops that we give. The rest of the team was against this since the current colors looked fine from their point of view. So we decided to make the color change on half of our playtester's devices to look at how they interact with the game and ask them after play sessions about that (indirectly). We had results after a week and it turns out it didn't make a difference. But if it had, we would have swapped colors instantly. We are making sure that such testing can be done easily and rapidly so that even more radical ideas like a new crafting system can be tested within one or two days of work (at most). Credits: We have credits linked by git commits in specific folders. Inside the groups, we have ranked them by commit count. I'm repeating myself; this is not perfect, but it works for us. In the beginning, we thought that this meant that in combination with role-switching everyone was listed everywhere, but surprisingly, that is yet to happen. When that happens we'll figure out something like introducing a cutoff like "you need x% of the commits to be listed and more than y" but we currently don't need that right now. Press: Everyone can write what they want, but it only goes out if more than 50% of the studio agrees to what has been written. E.g the pitch deck was reviewed by 70% of the studio. Does this incur a large overhead? Not really, it took everyone around an hour and the amount of buy-in from the team that you get is in our experience worth it. That is not "the article Tommy wrote for us", that is "OUR article about OUR game". Most of the things written stay in the internal docs and never make it out. You might have noticed that the processes presented require two things from everyone involved: Courage and humility. Technical competence has been proven to be learnable in our project, but those people need to have the right character. I hope more people experiment with this. It is worth trying even if it doesn't succeed. In my opinion at least.
I notice a problem with your approach. What if your artist comes up with a really innovative and good idea that will work if it shipped but no one at the studio likes it? You have a system that takes it granted that whatever the group wants is best without question and there are no things that group doesn't know. I recommend taking outside input as well when deciding about innovative approaches. There are also people shunning the unpopular person because they don't like them even though they do good work. There is also the problem of factions forming so people who want something made into the game ask others to join them saying they will support their ideas etc...
@@dragongoddragneel7106 in any system that ideia would be ignored unless the artists is the leader, in this system more ideas are considered in the first place
Personally (and from experience) I think that virtually everyone whos worked on any sort of large scale project intuitively knows that the hierarchy of people actually doing stuff is required and (again, almost) noone has a problem with that. As you say, the definitive answer from a lead, or the ability to on-board people at a junior level just makes sense. The problems in all cases I've observed is when hierarchies form among *people who organize people*, and somehow they wind up being more important than the people actually doing the work.
That's harsh to imply, that people who organize do not "do the work". In my little team the number of people started growing recently, and my responsibilities shifted from actually doing stuff or even simply spreading vision and ideas, to managing processes, organizing calls and so on. And often it takes up to a half of my work time on the game. And I can't say I like it a lot. I'd gladly pay from my own pocket to someone who could do it instead of me and probably be more professional at it, but for now I have to do it myself, because for the money a can offer no one would do that kind of job.
Although, I see your point. The credit in terms of compensation shouldn't be overbalanced towards managers, at least when there are layoffs in the company while top management still get paid their bonuses even when projects do not perform that well
protect yourself, once you open another business things change drastically... suddenly you're not the one hiring or deciding how things are structured :P@@Noowai
@@Noowai Exactly, I'm not saying that there should be noone to do that organization, it is very useful to take administrative load off the productive teams. Also note that I'm not including project managers who are involved in the actual projects, which it sounds like your role became. I'm talking more about large businesses where you wind up with multiple layers of management and 'project management teams' who only do Jira or Excel shuffling and treat the 'art/code monkeys' as replaceable resources.
I agree with this take %100. I've worked in flat hierarchies before and it's a real challenge to get firm decisions made and move forward towards the game vison.
I know that with Valve, their flat hierarchy is supposedly a nightmare where roles still exist but now they're all unknown and hidden and it's complex social structures and cliques I don't think I could do that, there needs to be people who do the thing that they're best at when it comes to anything sufficiently complex
Human nature with no structure usually devolves into high school style cliques. It gets especially hard when people don't know each other and are not on a first name basis. Arcanum was able to work because they only had 14 people who could all talk together and fit in the same room
There must be roles because surely a group of valve fans-come-developers would make the next HL installment. Someone in there is preventing certain projects from happening or green lighting others.
@@koalabrownie This has happened several times but there are people who, though they have the same role as everyone else, they're in good with people like Gabe and nothing gets done without their say-so The roles exist but only in the anarchistic sense of there being a hirearchy Valve is a flat structure and everyone is equal, but as we all know, some are more equal than others
Thanks for sharing Tim. There are hybrid models that integrates vertical with horizontal organizational structures; ultimately you do need a nucleus to act as a brain center for decision making within the cellular structure of an organism/organization.
Pay on a flat hierarchy, manage with a hierarchy. In the lean years everyone would take a hit in pay evenly, but would make bank in the fat years and it would have that organization needed. But hindsight is 20/20. Like in everything. At least you guys tried something. More than I can say for myself.
If you pay flat, you can't hire juniors or keep people who aren't at the top of the curve. No one wants to be making the same as someone who's grotesquely less productive/competent. At one of my jobs, we had a girl with an EE degree do the simple web interface and it took her a week to add a button and have it work (with some help from a senior) in, iirc, Tomcat (yes, that long ago). If she were making the same as kernel devs, we'd have had to get rid of her, but she wasn't and there wasn't enough work to hire someone faster. It's a pathological case, and we ended up with a bad interface so I would've fired her anyway if it were up to me, but the CTO didn't have the heart to do it. Money isn't always hugely important to people in a general/absolute sense, particularly the sort of people you want, but being compensated fairly for their relative contribution very much is, and you can't even buy them off with promotions/increased responsibilities because they often just want to build things and not manage.
I swear that "ego" is a learned experience for people who want to work in the entertainment industry. Being the best specialist you can be is reward enough when you yield consistently great results and you understand what it means to be an integral member of a team.
I used to work on a modding team where everything was decided by "consensus," which no one on the team could tell you what it is short of everyone agreeing on something. As you could imagine, things got caught up in gridlock a lot, and people's motivation and effort would slowly peter out until the only people who were making decisions were the ones who were stubborn enough to keep attending meetings.
I think this bleeds into what you talked about in the "Game Development Caution" videos. Hear me out: The problems you described here, are not necessarily problems with flat hierarchies. They are problems that arise from lack of organization. Having specialized people or teams who do certain kinds of tasks has nothing to do with hierarchy, the speed at which people finish tasks is orthogonal to hierarchy and it's not necessary to switch the guys between meetings with the same publisher. Hierarchies are an organizational structure where responsibility is torn apart between higher and lower levels of decision making. I want to focus on two problems: A) Power can lead to all sorts of abuse and problems, especially if power itself becomes a goal. Decisions become "political" instead of being driven by problem solving and empathy. B) The powerless can get disconnected as they realize (or think) that their only agency is to stick to the rules that they have been given. Again, to the detriment of the actual goals of a project or team. Leadership arises naturally and is dynamic. Alice has a problem, Bob sees a solution (due to familiarity/experience for example). Alice is smart, so she grants Bob the lead for that task. They solve the problem together, Alice learns something new and praises Bob, everyone is happy. It shouldn't matter whether Bob is an intern, or a CEO. Bob should feel _empowered_ to step up. Alice should invite Bob and put her learning hat on. The problems I listed above, are stumbling blocks and barriers for this kind of dynamic leadership. Especially if you add metrics, box ticking and other stuff that distracts from open collaboration.
Congratulations on finding a unicorn named Alice. I did crisis consulting, where I would be engaged once things had clearly gone very badly wrong in an objective sense (services down or slowed to a crawl, massive ongoing losses, so my fee started looking reasonable). In my years of doing it, there was exactly 1 developer who thought "we f-cked up, this is an objective fact, now this guy with tons of experience and earning astronomical money, presumably for a reason, comes in, tells us what and how to fix, and things get objectively better, I should probably learn from him". One. Yes, this is a biased sample, but.. One. When I started doing it, I tried building consensus, teaching people things, explaining the entire thought process etc. That just p-ssed people off more, because they couldn't live with the clue differential. When I once had a clue differential, I couldn't live with it either, so I set about bridging it like a maniac by learning things. This is unicorn behaviour, as it turned out. Vast majority of people pretend it isn't happening, lie to themselves, sabotage everything, the list goes on. Competence, it turns out, is a direct result of being able to evaluate your own ability and a drive to improve; incompetent people can't do the former and refuse to do the latter, almost universally.
@@paulie-g Thanks for sharing this story! Just the job you decribe is interesting and exciting to me, even though you had to fight a lot of frustration. I appreciate your perspective on Alice being a unicorn. But isn’t the context that you wouldn’t have been called in the first place if she wasn’t? Like a psychiatrist would think that everyone is crazy or a cop would think that everyone is a criminal. What I decribed above is something that is preceded by hard work and established relationships. It’s something that has to be earned and made as a goal. Me personally, I‘m rarely achieving that ideal, but I try. In any case, I‘m fascinated by your work as a crisis consultant. Never met one, but even just the idea of it is intriguing.
IME all groups have a hierarchy, it’s just a question of how explicit it is. It’s just human social psychology, certain individual voices will carry more weight for whatever reasons. While implicit hierarchies _can_ be good, there’s absolutely no guarantee that they will be; they can be structured around charisma or bullying instead of competence, and can be oppressive to people at _all_ levels. E.g., I’ve seen “flat” hierarchies where one person is declared the “leader” (not necessarily with their consent) and then has all of the responsibility for everything dumped on them… Even in the example of Troika, it obviously wasn’t truly flat if three people had to take massive salary cuts to keep the whole thing afloat while everyone else was blissfully unaffected. Of course, formal hierarchies are often pretty dysfunctional too. But with a formal hierarchy, at least the potential exists to identify and address structural problems within that hierarchy. And it’s certainly possible to have a _relatively_ flat formal hierarchy which encourages and rewards merit at all levels. You can have a group which strives for consensus and only falls back on formal hierarchy when necessary. Heck, you can even have a workplace democracy, where everyone periodically votes on how to restructure the hierarchy. But at the end of the day, the buck stops with someone, and it’s wise to deliberately choose who that is.
Thanks for sharing Tim, this is especially useful in light of Valve's continued success and famously flat hierarchy. There are rumors that Gabe personally stepped back from day to day development because of the inability of team members to see him as equal, instead they treated his word as divine and final ultimately sidelining any attempt at feedback for a person who is heavily invested in the flat arrangement. While I'm sure it is possible to be successful long term with a flat hierarchy, I can't help but think it would only be at the hands of a very special group of people, and who would be just as or more successful under a different circumstance.
When anarchists in Catalonia and Aragon took over industry during the Spanish revolution, productivity went up as soon as they showed their bosses the door. Maybe the context has more to do with it than any very special group of very special people.
@@ReadyAimSing so a group of ideologically committed people worked together to establish a scenario that met their ideological goals, one which is almost universally atypical and you're telling this *wasn't* a special group of people?
It really seems like in order to get a flat hierarchy to work, people need to have a perfect understanding of their own limitations and must have the humility/lack of ego necessary to admit this. Basically the exact type of person who would also do well in a hierarchical structure.
No. Hierarchies are really just not workable in the real world. You convieniently left out that everyone hired has to be at the exact same skill level and that everyone has to get paid equally. The onnly reason they could even attempt this is they were brand new. 5 or 10 years down the road it would be literally impossible. If you have someone who has been working at the company for 10 years and a brand new person is hired and brought on at equal level and pay, that is not fair or reasonable. In the real world people are not equal. There are 25 year olds who know little starting out and older experts like Tim. So, tell me, which group do you never hire?
Bingo. Flat only ever works in a (small) team of A-listers who are also great people. Incidentally, based on my anecdotal experience, such teams tend to self-construct an informal hierarchy where they voluntarily give final decision power to someone (or multiple someones, one per sector of responsibility).
This is fascinating and a great look into how flat hierarchies function within indie studios and how they could be improved in the future. Your knowledge is invaluable Tim!
@@chomnansaedan4788"ants or bees" sorry bro we aren't made in the same image as ants and bees or whatever your version of God is that is conveniently more real and valid in your head than other people's versions of God, you should really read or watch more about biology or anthropology before making bizarrely baseless claims on the subject of biology and anthropology in front of others
@@chomnansaedan4788"ants or bees" sorry bro we aren't made in the same image as ants and bees or whatever your version of God is that is conveniently more real and valid in your head than other people's versions of God, you should really read or watch more about biology or anthropology before making bizarrely baseless claims on the subject of biology and anthropology in front of others
Your personal experience videos are invaluable! Many dream of making a horizontal hierarchy, however, miss the fact that everyone has their own productivity capacity and motivation, their own personal responsibility level, and, finally, many need a mentor or a leader above them to do their best! Thanks for sharing!
I work in a small town school with a somewhat flattened hierarchy and it's a life changing experience coming from industries where the opposite, top down distribution of ideas is ubiquitous and set in stone. Admittedly there is more basic structure to guide people in their roles and duties than what you describe here, but having the ability to get your concerns heard on a regular basis by an entire staff who is all pulling in the same direction is what keeps me coming back to give it my best even though the pay isn't as competitive. Feeling like you're a contributing member of a community who sees the benefits of the struggle to produce a positive result vs. feeling like a mercenary or even worse, a serf, changes your work and how you deal with coworkers for the better. I had a professor at one point sum up his experience with flat and vertical authority structures as the difference between valuing long-term growth and professional development vs. achieving efficient and viable short term results. When it's a luxury you can afford, I think experimenting with ways to flatten your workplace hierarchy is a great idea.
I did two years at a startup with a flat hierarchy. At first it was a lot of creative freedom but it quickly became a planning nightmare with no one ever knowing who had final say.
It's an interesting challenge to close the gap between people's desires to believe in perfect equality (not the social kind, the capitalist kind) and the reality that some people can literally produce considerably more than others at the same quality. Some people are simply worth more when the metric is, "getting the job done correctly and on-time." I've managed a lot of underperformers, and you know who end up with the best career outcomes? Those who acknowledge it, take responsibility, and work together with me to maximize their ability to contribute. Those who externalize by blaming others and making excuses always ended up eventually being let go.
Rather than flat hierarchies, which ignore so much about what each individual contributes to a project, I prefer a system that respects the real contributions of each individual. I spent some time at a company with a near 1:1 ratio of producers/project management to designers/developers/artists and it was clear that producers/project management were compensated at rate far beyond what made sense considering their contribution. Some producers/project managers were there with us on those nights when we just had to crunch through to hit a deadline, and we all loved them for it. The team gets to know who coasts and whose backs are coasted on. All of this to say I wish that there was more transparency about compensation vs contribution. When you're doing a rundown at the end of a project, and you see quality contributions across several domains all over the place from one member of the team vs a few contributions, even high quality, here and there from another, it isn't hard to understand why one might be compensated more than the other, and most folks on the team aren't going to argue that that's unfair. I never really thought of flat hierarchies the way you're describing them. I just want to avoid having some BS chain of command I have to traverse to get the answers I need or help someone in a different domain. I want to go to the person I know has the answer or who can point me in the right direction. If I see somewhere I can contribute, I don't want to ask a manager to forward a message to a lead to give a suggestion to an artist. I just want to go to the artist. I really hate the slowdown that can come from a deep, siloed hierarchy. If I wanna chat with the programmers, I don't want to be locked out of doing so because I'm in the art dept.
Huh? I've never worked at a company where staff couldn't talk to each other and had to communicate through managers. Managers are there to make important decisions,resolve conflicts, approve your vacation time, do annual reviews,, etc. If staff are working on something and have questions there is always ways to directly communicate in my experience. Some people do hide behind their managers, but that is pretty rare.
"from each according to his ability, to each according to his needs" It was a nice experiment to have, and to hear about! I conclude that hierarchy is not a bad thing, direction, leadership, are needed for organization, it is a matter of logistics. But being Hierarchical that can be bad thing, when a structure needs to enforce aggressively the verticality of it. Even collective companies and movements need a head, not to have exclusive power of decision, but to have a trusted final word, and a proactive interest in making sure all are doing well, and know what they are doing, to the best of their abilities. Everyone being given the same amount of time and work and expecting the same results is utopic, idealistic. For good or bad, we are different, and we become different from ourselves too with time and situation. But everyone can be very good at something of theirs, and best using that (while not chipping away other interests) is the way to go for projects. With that, you are totally right, Tim! While far from your amount of experience, I've seen in games I worked on how it is not good to make everything a group decision, or to let everyone meddle with one another's work, that is for their department/group to do, and their lead to settle if there is an impasse. Especially when the whole team is not made-up of people who know each other. At last, I must say that doing the experiment from scratch, inside a system that hinders in every way it can a collective owning and effort (and ideologically is very individualistic and hierarchical system), will be a Sisyphus effort. Everything, from the ideas implanted on everyone from the culture, to the tangible corporate structure, rules and culture as well, is vying against that. Even so, I think we can all appreciate the effort at Troika for trying it, and congratulating them on the masterpiece they delivered while doing the experiment!
That quote in this context is interesting because it implicitly rejects both the equality of ability and the equality of needs, otherwise it wouldn't need to mention either. And it sounds like Troika tried to do this but didn't want to let go of the idea of determining rewards by the same metrics as preexisting hierarchies, and they tried to resolve the contradiction by simply making sure that everyone's abilities in the company were on the exact same level. And obviously that didn't work, because it can't.
So here's the thing. Re-read the phrase "From each according to ability, to each according to need" carefully. Nowhere in that maxim does it talk about "equality," about everyone being the same. In fact, it rejects and refutes that whole idea. "From each according to ability" implies quite clearly that people have different skills and different abilities. But "to each according to need" separates their "performance" from their ability to secure the basic needs of life. Just throwing that out there for you to think about. And, full disclosure. If you didn't figure it out by now, I'm a communist :)
It really was not fair or equal though. The three owners were on the hook for the rent and the contracts. If things didn't work out they could have lost their houses and savings while the employees would have walked away scott free to their next jobs. There is risk in new ventures, often owners need to put up a lot that they may lose which is why they make a lot more than employees if things suceed. Employees never stand to lose anything more than loss of work, they have no actual risk.
@@PaulRGauthier Yeah, so in practice advanced Marxist socialism failed to produce a classless society. New classes formed on the basis of things not related to economic relations. It's possible to culturally enforce 'wants' being much closer to 'needs' than in capitalist societies, but in practice that turned out to require prohibiting people from leaving the country and concealing information about comparative living standards, even in the presence of some hardcore social engineering. It's a nice notion, but frankly we're past that in modern industrialised economies - they are well in surplus, so solutions to scarcity problems are a bit anachronistic.
The experience is really valuable to understand the problems faced by the idea. That's the best kind of information if one wants to try something similar in the future.
One of my favorite videos so far. Since I'm interested in both game development and anarchism, loved to hear about an actual flat hierarchy experiment in the industry. Even though it failed in the end I believe such attempts are meaningful to create a way towards better forms of work relationships.
It doesn't surprise me that a flat structure doesn't work well; but I think the problem that it is trying to address is valid. Hierarchies and especially top down dictatorial ones can be problematic. Yes, its possible to have a 'benevolent dictator' for a time; but a team structure that requires one doesn't normally protect itself from a non-benevolent or authoritarian dictator (and are dictators really benevolent or are they just charismatic). Having a flat or just a more democratic structure; you aren't going to avoid problems, you are just going to trade problems for different ones.
This is super interesting! I feel like a bit of hierarchy is essential as you've proven; but my problem is a lot of businesses take hierarchy as an opportunity for corruption and nepotism. A bunch of people get shoe-horned into management roles that don't need to exist and basically are there just to devalue the people who are actually doing the work. So to me a flat hierarchy can still have a lead, or maybe a 'senior' person who can make a final call and keep order. It's a way to avoid the all to common situation where if you need approval, have a question, need help, or anything you have to ask your lead, who has to ask their manager, who has to go to HR, who then has to go to upper, who then have to go to the real in-charge people who have no idea what you need or what's even a good response and frankly don't care anyway.
It's the difference between shallow (often and confusingly called flat) hierarchy and no hierarchy at all. Hard to make the last one work with humans since we are naturally hierarchical, but I completely agree with you that leadership is what makes hierarchy work, not positional/imposed authority.
@@JavierBonnemaison "we are naturally hierarchical" -- Sounds like something Jordan Peterson might say, though some of the greatest minds of our time think him very wrong about that among a few other things, perhaps especially his confusion of individualism with iindividuation. (It's an understandable confusion that is even found in the writings of luminaries such as R.W. Emerson at times, but is a confusion nonetheless.) Strictly hierarchical organization is not a "force of evolution," but is based on a dying worldview whereas Helena Norberg-Hodge, et alia, is quite correct when she says, "We human beings didn't evolve for a hyper-individualized, competitive way of being. For most of our time on this planet, we lived in collaborative, intergenerational communities." We know this intuitively, if in no other way, and indigenous communitites are generally exemplary of it. Some might say Helena's, among others, is a "feminine" sensibility, but it's actually a sensibility we all possess to one degree or another. And anyone who has worked within a strict hierarchy knows the feeling of being treated as "just another cog in the machine." The aforementioned worldview may be prevalent and predominant, but is in no way natural. As Deborah Frieze notes in her TED talk, "Somewhere along the way, maybe around the time of Isaac Newton, we got confused about how life works. We convinced ourselves that the world was causal, linear and predictable and, so, we began to treat our bodies, our communities and our ecologies as though they were machines." This "techno-logical" wordview or "mental-rational structure of consciousness" (in Jean Gebser's terms) pervades every aspect of our lives and human endeavor from business to education to government to healthcare and reductionism is rampant to the point that even some scientists, e.g. Adam Frank and Marcelo Gleiser, are decrying it in their respective fields. An earlier comment by alexpetrovich85 mentioned "hybrid", integral models of organization we're beginning to see emerging here and there. Personally, I think such "models" deserve our attention and nourishment as the modern era naturally passes away and a new era is born.
@@lrinfi I don't think it's debatable that we are hierarchical. You have to do some serious cherry-picking to come up with an example of a properly functioning society that existed which didn't organize itself hierarchically. If you can even come up with any. Most of what you're saying goes against observable reality. We are collaborative and competitive, and we've always been, including in tribes and between tribes, and we'll always be, because all Life is collaborative and competitive. And most, if not all mammals exhibit some degree of hierarchical and individualistic/tribal nature. Humans are not outliers. We act according to our nature, and the proof is everywhere around us. No amount of pseudo-philosophy can argue against observable reality.
@@lrinfi TED talks are not peer-reviewed. Hierarchy in humans is a broad consensus in evo- and *-psych as well as anthropology. "X might agree with this, therefore it is wrong" is a logical fallacy, regardless of how odious you think X is.
i remember some 6ish hour podcast with John Carmack, him takling how this didnt worked because some ppl work more, some less adn where theres no boss to call the shots it turns into everyone doing his thing if they feel that the collective decision does not fit them
The problem is that there's always the two extremes. You have the typical big company structure where those at the top make 100x the amount of those at the bottom and while they have more responsibility they don't have to do anywhere near as much heavy lifting as those at the bottom and when the shit hits the fan because of decisions made by those at the top it's always those at the bottom who are let go so it's responsibility without much consequence. Likewise this extreme was never going to work either for all the reasons outlined by Tim. There's a happy medium which rarely gets implemented because the obstacles are either the greed from those at the top or on the other end people exploiting the good nature of owners/leaders who try something different which leads to poor work, a lack of direction and poor work ethic. It's a very difficult balance to strike and I applaud these guys for trying something different but any successful project will need clear leaders, delegation and square pegs in square holes.
In general dev I really like having a product owner who decides how the product will be but doesn't involve themselves in the code and a lead dev to make final decisions on code related matters rather than a project manager who has to decide everything. It's flatter but you still have someone who can say yes or no.
Obviously, the history of experiments in human organization without any form of hierarchy is long and goes far beyond the game industry. I'm sure part of the reason people react with "but it would've worked if you did XYZ" is because the subject matter overlaps too much with their opinions about wider societal organization (and thus draws the conversation towards more emotionally intense beliefs).
Ooo, there is so much good stuff you go over in this video. Among lower-level employees, many don't want a higher level of responsibility. They may enjoy working, but they don't want the responsibility if something goes wrong or have to make any difficult or important decisions. Conversely, others who want to lead may be horrible leaders. Customers, vendors, etc may feel most comfortable interacting with a particular individual. At my current job we sometimes have customers who would rather talk to me than the sales/customer support team. Problem is, I can't do any of the help or contract changes that the sales team could. Not to mention, every minute I spend talking to a customer is a minute I'm not making the products the company will deliver to the customer. I can explain technical details that the sales team don't know. There are other people that want to talk to me because they think they can get a better deal doing that. I can't give them the discount they want. Hell, I would charge them more it was up to me. You have the mommy/daddy issue when you have people at the same decision-making level. People will ask one, and if they don't get the response they want, they go ask the other.
I don't think I would do a completely flat hierarchy if I had a company myself, but there are experiments that I want to do to incentivize developers, such as flexible hours and still get a part of the sales revenue, and a loose team-based approach to development, wherein developers organically form teams to get things done.
Yeah, tried this as well when me and a couple of friends/colleagues created a company doing VR games. It worked about just as well. Not that people didn't have skills, interesting things to say across the board, or passion, but it went down exactly the same. I think if anything, everybody who's naively optimistic about flat hierarchies should try it to gain the understanding of why hierarchies work, and why it's more fun in the end. It strengthens you as an individual, and it strengthens the team if everyone is willing, but it is not a sustainable model for coherent games. Not that we were a team of 15 people when we were at max capacity at the time, but personally, I think even with 3-4 people, hierarchies are still preferable. It's easier to do overlap, and own certain areas, especially the broader skillsets you all have, but having someone with the final word is always great, to lead the vision of a project.
When everyone is responsible for some decision, it actually means that no one is responsible. So yes, totally agree -- structured hierarchy, which implies clear responsibility for a particular area of the project -- is a way to go.
Thank you for sharing this! I seem to be noticing a lot more people discussing alternative business and work structures these days, especially after the pandemic, so I'm sure this will be helpful for some upcoming companies.
Working for a start up that had recently turned into a huge corporation, I experienced various levels of hierarchy. By far, less hierarchy is better. You need to take into consideration the massive morale boost of giving everyone the respect of being equal. Direction is needed, not management.
Love each and every one of these insightful videos. Just wanted to note that even the purest anarchist societies have hierarchies on the job / hunt. The thing about them is that they fall away after the work is done. Culturally, we're a while away from it (you can try building any kind of structure under capitalism, but you're still drawing from a pool of people raised by capitalism and operating under the rules that ultimately reinforce it [bank loans, publishers, property laws]) I think democracy in the workplace would be a better way to go then having a flat hierarchy. It wouldn't be perfect, nothing is, but being able to elect managers or specialists (i.e. communication) based on their performance (and recall them) might create a more fair environment for everyone. It would probably lead to a lot of problems too, but we have democracy in public life and most would agree that it's preferable to the tyranny of kings. A huge cultural shift and boost in education would be required to get this kind of workplace running, but that's the dream. Sorry for getting political here, I just found this video particularly interesting! Getting people to do the same amount of work can be challenging as people can just be jerks or slow. Speaking as the latter type, I'm particularly intimidated by work made by legendary figures in the industry such as Tim et al., but these videos have inspired me to keep trying no matter what. Have a great day everyone!
7:55 You *can't* say that? "No. I'm pushing back on you, publisher." Why do I think you *can*, but it's perhaps not considered advisable in the present industry environment and atmosphere. After all, the publisher just might decide to say, "Well, then, we're not publshing your game. Nyeh." (Or pull some other such juvenile attempt at "power play".) Any alternatives to bowing to the whims of the major publishers on the table anywhere?
Interesting that Arcanum was made under a flat structure, I wonder if the flat hierarchy contributed to the "everything and the kitchen sink" feature set. Still one of my favourite RPGs ever, love the setting especially. I imagine whether a flat company structure is viable depends heavily on the people involved, the financial strength of the company, etc. The most famous example of a flat structure gaming company is Valve and it has been working well for them for decades, so it is possible.
I'm not sure I would say its working well. They barely make games anymore, if it wasn't for Steam then I have a feeling it would have fallen apart since there isn't much guidance for what to do so people have a hard time getting together to actually make a product. Probably a big part of the reason that Half-life 3 is a meme.
@@Worgen33 You're right, it's because that flat structure that Valve barely did anything, as I understand, Half Life Alyx very much required to abandon that in favor of a hierarchy to get anything done. Also the lack of a formal structure doesn't mean there's a no a de facto one. Back in 2013 Jeri Ellsworth said it very much is like a high school, the popular guys are the ones that have all of the power.
@@Worgen33While I love the Half Life series as much as anyone, I'd rather take a pile of cancelled projects over the sort of soulless yearly release that gets pushed out the door on time regardless of quality. I used to consider Far Cry one of my favourite series before FC3 came along, I don't want to see Half Life go through that.
@@AndrewFullertonExcept that Valve has enough money to not need to rely on sequels. They can make more original IPs if they want, but with a flat hierarchy they can't get enough motivation to team up and even do that. I think the Deck is the most motivated they have been in a very long time.
Thank you so much for this video Tim, a lot of fascinating ideas & results to chew over! I'm taking all the notes on this one lol. Glad i got to learn about this secondhand and not first hand because a flat hierarchy is something right up my super idealistic alley lol
We tried a flat hierarchy system, too, with the group I first tried to make commercial games with. It didn't work out at all, never released anything but a couple of jam games.
A totally flat hierarchy would be difficult with more than, like, five people, but there might be some hybrid approaches worth trying. Perhaps one day ill get to try it out. I love the idea, but perhaps there are more lessons to learn from it than practical applications.
Hey Tim! At 11:32 you mentioned random modules shooting events at each other. What techniques do you like to use to share information between modules in code without tight coupling?
I think there are many examples of this actually working, but its quite difficult to get off the ground. Mondragon,valve,etc. While history for the most part has been pretty terrible to more co-op organized orgs i think in the long run and how the world currently is this may actually be the superior way to do things. Being an early adopter of new systems is always difficult.
Valve is a rentier operator of a service that has had a quasi-monopolistic position until recently (arguably even now; the lengths to which Epic is going to try to compete demonstrate how high the barriers to entry are). So they can almost do whatever they want in terms of internal management because they aren't subject to normal market discipline. I think they have a very low number of employees for what they are doing, but even so, from the outside the way they operate Steam seems very stagnant and inefficient (e.g. buggy or poorly-documented APIs persisting over long periods). I doubt their success can be attributed to a lack of internal hierarchy, and I think the main reason it persists is network effects keeping them at the top of an external hierarchy.
@@globalistgamer6418Yep. It's like if one location of a co-op coffee shop was pulling the revenue of every Starbucks in the country. They could have whatever structure they wanted; there's too much money coming in for them to fail anyways.
Often the advantages and disadvantages of a co-op don't stop broad market effects or don't stop a bad company from doing badly. And they ignored the positive effect it has on its workers. A key example of this is a study done on how co-ops handled 2008 and found that they were hit harder and earlier but then recovered faster. Workers voluntarily furloughed themselves to keep the co-op going which allowed them to recover and survive. But plenty also decided not to. Just like how some traditional businesses better prepared for such a crises and acted fast to keep things afloat. Plenty of co-ops went under but so did plenty of traditional businesses and what is often ignored is that workers in co-ops are better off, hence them willing to furlough themselves even when the future is unclear and it might be for nothing. Having a say in your employment offers you better conditions and pay and in return employees are more loyal and do more to keep things going. Research has shown that at the very least co-ops are better for the employees and either don't effect productivity/performance or make it slightly worse/better. The advantage of co-ops will always be the fact they better distribute wealth to their employees which make the community wealthier, which then makes the co-op more stable during down turns since workers and communities are more likely to bail them out but I've gone on long enough.
Valve is a flat hierarchy only on the surface. Beneath the surface it is basically a meritocracy. People who matter dictate what gets done, everyone else has to go along with it.
A funny thing about the human mind is that we can believe anything; nevertheless, there are other aspects of human nature. We are indeed hierarchical animals, yet our strength lies in reciprocity and cooperation. I don't think rigid and deep hierarchies work well, but neither does full egalitarianism. The ideal balance for a particular group of people is very much context specific and dynamic, but it will always be closer to the middle than to the extremes. After all, this is why humans are so adaptable.
Man, the alphabetically listed credits is an incredibly subversive and very cool thing though. I've only seen it one other time, and that is for Drakengard 3. A game that was definitely made is a hierarchical structure, but those credits just completely floored me. Even the company itself had to wait until S to be named. And the lead designer was almost the last person in the list.
I heard Resident Evil 7 did a similar until the middle of development thing expect they had roles and hierarchy its just that almost anyone could pitch ideas middle in development they switched to a more assembly line work flow that game went fast only took them two or three years to make
I wonder if this "failed" structure is, in the end, a big contributing factor to what made Arcanum so great. The secret sauce, if you will. Don't get me wrong, I love VtM: Bloodlines as well (and I admit, I haven't played Temple of Elemental Evil... yet), but there's just *something* about Arcanum. Is it buggy? Hell yes. But now that I understand the work structure behind it, it's dawning on me that maybe Arcanum is so special, because your structure back then was more akin to a passion project of a bunch of friends getting together to make something great, instead of just a bunch of people being hired to do a specific job. As you've said - people tried many hats on, and that makes me think that perhaps Arcanum is so great specifically because every tiny bit of work that went into it, was work that people *wanted* to do, not just did for a paycheck. Thank you for sharing. You call this structure a failure, and perhaps in business terms it was, but it certainly led to something amazing.
thats really cool insight you shard. wsh you guys had had the time and resources to investigate how to do flat hierarchies or similar models that try to flatten power dynamics in the workplace, e.g co-op models.
I apologize if this is something I am simply unaware of, but have you had any experience in developing TTRPG’s, or a desire to? What are you thoughts on the most recent iterations of D&D and Pathfinder?
Hi Tim, Valve is famously a flat heirarchy company, and they've been doing it for a while now. They released a PDF outlining how this flat system works, no idea if you have seen that or not. Do you think they faced the same issues that you have faced? Or did they do some things that in the right way that made it work? From our side (the players/consumers), we see the benefits of this system (if a dev is passionate about something, they can spend more time than they need to on something and make it very good/crafted with love) but we also see the cons of this. Famously, projects/features get neglected despite having a large playerbase. Team Fortress 2 for example has been in the Top 10 most played games on steam for a long time, yet has bugs, riddled with bots, stopped receiving official content, and may have either 1 or 2 devs working on it (or 0). Because their devs can work on whichever project they want, but every dev also feels like working on the next "big" thing at their office. Were there any benefits that you enjoyed from having a flat hierarchy at Troika that made the game better for the players or was it overall a negative thing?
The "problem" with Tim's structure was that it tried to be both flat AND open. You can have flat when you keep everyone restricted to a specific set of tasks and make all tasks equally valuable. You can have open when you have a hierarchy in place to keep people from wasting time and doing things they shouldnt, but also to approve and encourage people to grow their natural talent they show. Both simultaneously is a trap as they found out. Still really happy that the Troika owners put the idea of employee value forward and really ran with it. Even if it was a mess, it was a glorious one.
I don't see Valve as a broadly replicable system - it's easy to run a commune when you can only handpick the most intelligent .0001% of the world's population to populate it and you never have to worry about funding/project deadlines since you have guaranteed income from your monopoly on the gaming distribution network.
You realize Valve has been around since 1996 and Steam wasn’t the first thing they did, and it didn’t really take off immediately when it was released in 2003, when it was just a platform for valve games anyway? And the first thing Valve DID do was a game that killed dozens of competitors on arrival and God knows how many companies just tried to copy them for years after that? And they did it again with the sequel?
@@OlStinky1 Exactly. Valve can basically do whatever shit they want now, as long as it doesn't undermine Steam. Games, software, hardware, profile backgroinds (lol).
I think this thought is interesting because in the macro politics it's people under financial stress that tend to advocate for flatter hierarchy. I have no statement beyond that it's just food for thought.
I've long wanted to participate in a co-operative democratic structure where there is a hierarchy, but there were structured means for people lower on the hierarchy to vote on the ratification of higher level decisions and positions. For example, people would be nominated as creative director, art director, lead programmer, for a project or for a two year term, and they would have the authority to make decisions at that level. They would perhaps hold veto power over team-wide ratification votes. I have many more specific ideas of how to integrate balances into this, but it's a little much for a youtube comment 😄 There would of course be issues with new problematic dynamics in the workforce, but it would give lower level employees more say in what they're doing when they felt strongly about something, while also giving operational structure and clearly defined roles to make day to day work flow better.
Yeah, Tim and the guys are the kind of dev that I would root for their success. I kinda wish Tim born a bit later so we can have him still leading more projects today. I would love to play more games from him. Heck I want to work with him 😁
Valve is a famously fully flat company. Valve is also famous for being a brutal place to work, full of cliques, projects that get stalled or killed due to office politics, and using stack ranking to "motivate" people. If it wasn't for the endless money printer that is Steam, the company would have imploded years ago operating the way it does.
It sounds like ego was the focus on this experiment, which could have been its downfall. Instead of a flat hierarchy, it could have been a collective. Roles could have been filled appropriately while money was distributed evenly.
@@BobArctor Tim described that the flat hierarchy meant you could grab any hat at any time and give it a go. A proper collective is more realistic about what a business needs, and doesn't worry about hurting people's feelings by plugging them in where they belong. But the pay is even. Tim said that Troika wasn't just even pay, but also even roles (hence the word hierarchy). It just sounds like it was a symptom of everyone (rightfully) feeling wronged by their previous employers, so they decided that they would distance themselves from that form of business in every way possible. It's an understandable reaction, but sadly one that ended up being a learning opportunity and not a full-on success story. Of course I wasn't there so what do I know? I'm just going off of what Tim had to say about it.
I think the right way to try a flat hierarchy is to have the participants do what they are good at 80% of the time and 20% flat hierarchy stuff. And the team have to be balanced from the start, those 80% should pretty much be enough to get the project done within deadline and the 20% is more of a educational part of your time and not focused on deliverables.
I think it's pretty telling that even Valve - the most famous flat heirarchy studio has recently started to move to a more traditional structure. Flat heirarchy sounds really great in concept until you get into all the problems Tim mentioned here...
To be fair...thers flat hirarchy and theres FLAT hirarchy, and this sounds like Nevada desert flat. There are definitely more nuanced versions out there, with something like only 2-3 layers for example, that do in fact work.
a flat hierarchy is likely only optimal in low growth environments, it can generate creativity, but hinders growth, in theory it should tend to have one sane person managing everything amid the chaos - some guy who works in quantitative finance and game theory
It's easy to interpret worker self-management as design by committee and ritualized bikeshedding, but I think, in reality, it can take a substantial amount of time and resources to figure out how to run a worker coop: where leadership roles make sense, how to have them without sacrificing workplace democracy, how to make those positions accountable and recallable, etc. It's not trivial, but a lot of these things have partial answers, owing to a century of anarchists, conscious and unconscious, actually doing them and seeing what works. What you describe sounds like an early, naive attempt at exactly the right idea. Maybe if Troika had the time and money to learn from it, it would have gone a different way.
I've been trying to explain this to my leftists friends (I'm on the left, but not a leftist). Having run a business for 15 years, I know that there's so many things that the average person doesn't understand about what it takes to make a business works. And if the owners are putting in more resources than the rest, its' already not an equal situation. Most people don't want that responsibility.
I went from flat to top-down. The flat was WAY better in acting fast and creating a product iteractivelly, we always had roles, we had people with veto power to balance things, but even with those in place it didnt work over 50 employees. Dividing work ALWAYS fail, ALWAYS! The churn rate of employees was 50% PER YEAR! The average tenure of employees was 1.5yrs Its hard to do flat and manage it, very hard...
It's work. It mystifies me when people and organisations bend over backwards to try and make it like normal social life, happy funtime. It's awkward and never works (because you'd never socialize with half the people you work with). You can (and should) enjoy your work and appreciate your workmates but it's still a job. It's just so much easier to get s*t done when everyone knows the plan and where the buck stops.
If you read Polybius, it's clear he thinks that the reason the Romans were able to conquer the entire Mediterranean is due to how well-organized and hierarchical not only their military was, but the whole rest of their society.
I don't understand why you would even want a flat hierarchy. Most of the problems you listed are pretty obvious a priori. What problems did you expect a flat hierarchy to solve and were they solved when you tried this?
I think the error is assuming that "everybody receiving equal pay and having the same amount of work" is a flat hierarchy. If you have this arrangement and it was imposed by 2 or 3 people in the organisation and the others had no say in these and other decisions, it's just a typical not-flat hierarchy, with just two layers. At least as far as I understand it, a flat hierarchy is an organisation where everybody has equal say & vote, where everyone is up to speed of the company "health" situation and major external relations, and where ideally big important decisions are made by consensus or at least with a high portion of the members. These decisions might imply that not everyone will be compensated equally (except perhaps regarding benefits) but rather in function of the work being done and/or other criteria reached by consensus. It might imply that there are designated roles by (past AND present) merits instead of no roles or interchangeable at the whim of anybody. It also implies that these decisions CAN BE CHANGED if it can be shown they're not working. Also, AFAIK, there are some kinds of company structures that allow everybody to be owners of the company in equal parts (or another kind of arrangement) and have one or more delegates to be legal and commercial representatives of said company.
I was also part of a small company that tried this. There were only four of us, and it still didn't work. Groups of people just work better with clearly defined roles and structure. Perfectly egalitarian hierarchy is a lovely pipe dream but it's just not the way humans are wired.
I think though that there is a mindset that goes with this - just because someone is a lead and someone is a junior their contributions should be weighted by their position, in that sense most reasonable project based hierarchies are totally egalitarian. As has been mentioned in other comments the 'unfairness' comes when you wind up with management layers who are compensated more for essentially putting brakes on communication lines and shuffling around paper.
Could you tell us stories about the development of Tyranny? I really liked that game and think it had a lot of potential, it was a great experience to play through! (and finding other paths/branching storylines in next playthroughs) :)
Applying this micro economic infinitesimal failure across a broad swath of economic theory like this comment section is doing is far from “based”. Plus, see how employee owned companies work.
It sounds like you discovered that while the assumption is a hierarchy is a means of boxing people in, it also allows people to stay protected in their box. When you don't have the ability to say "we want one of the coders on this" you have to say "we want someone who's actually good at coding on this" and suddenly everyone is effectively stat ranked by everyone in the studio on every task that needs doing. Sounds exhausting. There's no reason you can't have many of the benefits of a flat hierarchy such as people wearing multiple hats when they already have their one primary hat nobody can take off them. Very interesting video.
Holy cow! I would have loved this. And equal pay for everyone including the owners -- INCLUDING THE BACKEND! Fuck. I'd have quit a job and moved for such a thing. Because I'm an obsessed, work-a-holic weirdo who loves to wear many hats and is very entrepeneurial. I used to get in trouble at large corporations a lot for going outside my lane and often doing others jobs better than them before I got tired of getting yelled at. Most corps want easily replacible cogs. Anyway, this experimen would have been fun.
Hi, Tim! Great videos, so cool you've decided to create a channel and share the experiences! You've probably been asked this a million times, but as a long-time fan of Troika and Arcanum, I have to ask: is there ANY chance of Arcanum 2, at some point?
![KSI - Dirty [Official Visualiser]](http://i.ytimg.com/vi/VoDh1h1dPUE/mqdefault.jpg)
It's funny and appropriate that a game that touches class struggle was made this way.
If it's any consolation, that organizational "failure" produced a masterpiece of gaming art.
Even though you didn't specify which game, I think I know you're talking about Arcanum
@@BuzzKirill3Dhe said at arcanum is the game made with that structure
@@jtaco4101 fuck, I wasn't paying eonugh attention
Since it worked badly we just got a worse game than otherwise. All the good stuff in the game came from the talented and creative developers, the structure just held them back.
@@xyhmo Not exactly. If you simulate a traditional vertical structure you'd most likely end up with a schlock version of Arcanum since it would've precluded Tim & Leonard's direct involvement with this project as they would be self-confined to their higher-level roles in the hierarchy.
Hence why their "failure" to stay in their lane was the reason the game was anti-schlock (but a failure business wise).
I think two words got confused here: Role and Rank.
My team is currently trying out no-hirachies and it is really working for us.
But no hierarchies doesn't mean no roles so I'd like to "respond" to the things that went wrong in the video and tell how we are combatting this.
Hire only seniors: We are currently 2 Seniors and a bunch of first-timers. We are making progress and no prejudice is in the group. I'll explain later why and how.
Estimation: OK, this is the first problem. Estimation is always bad. Because it is always wrong. So we just don't. But how do we make decisions? We measure. We simply do it. measure the time it takes and then extrapolate. The results mostly are correct within a 5 to 10 percent error margin which is good enough for us to make decisions with.
Deciding what to do: We have a list of ideas that are roughly ranked by how much we want them. Every time somebody completes a task, they pull the next thing from the list that they can do and start working on it.
Working efficiently: Each week we retrospect on what was blocking the most or is currently the slowest part of the pipeline. Then we discuss what the problems are and how we can speed things up. This works on a macro level and we are still accelerating production speeds. Sometimes it was the coding that was blocking, sometimes it was asset creation. Almost every time the solution was a day of automation work or a 10$ plugin. On a person-to-person basis, things get discovered naturally. People notice if others are slow or they are slower than others. We combat it by grouping the fastest with the slowest if this happens (for a day or two). The slowest then has the opportunity to learn from the faster one and learn all the tricks that are needed. Sometimes it was an artist always using the mouse for everything instead of shortcuts, sometimes it was a programmer not knowing the standard library well enough. By always learning from each other, we have managed to reduce the inter-personal time differences beyond noise levels. (Took us a year to get to this point) [measured by time tracking work tasks and running a quick Python script to do statistics on the metadata. Not saying it is perfect evidence, but good enough for us.]
Fluid roles: This one is tricky. Letting people just do any task they want is a recipe for chaos. The group has a consensus on who is good with that. So when somebody wants to do something that they can't do, they usually go to that person and ask for help. If they want to do a full-on switch (Programming -> UI) then we cultivate a large resource library with all the useful learning resources that the previous experts used to get to the point where they are now. Of course, this doesn't eliminate the need for practice, but you would be surprised how good programmers can make GUIs if they just read one or two books first. I have to admit this apprentice and resource system is not perfect and doesn't allow for complete switches like musik-> animation, but it reduces roadblocks in development effectively enough for us. An animator doing the UI task that's needed or a programmer doing the audio level mixing inside the game is achievable and liberating in our experience.
Conflict resolution: If somebody is insisting that "that thing would be so great" then we just let them make the lowest amount of effort that can validate their point.
Example: Last week the 3D artist argued that we should switch the colors up for drops that we give. The rest of the team was against this since the current colors looked fine from their point of view.
So we decided to make the color change on half of our playtester's devices to look at how they interact with the game and ask them after play sessions about that (indirectly).
We had results after a week and it turns out it didn't make a difference. But if it had, we would have swapped colors instantly. We are making sure that such testing can be done easily and rapidly so that even more radical ideas like a new crafting system can be tested within one or two days of work (at most).
Credits: We have credits linked by git commits in specific folders. Inside the groups, we have ranked them by commit count. I'm repeating myself; this is not perfect, but it works for us. In the beginning, we thought that this meant that in combination with role-switching everyone was listed everywhere, but surprisingly, that is yet to happen. When that happens we'll figure out something like introducing a cutoff like "you need x% of the commits to be listed and more than y" but we currently don't need that right now.
Press: Everyone can write what they want, but it only goes out if more than 50% of the studio agrees to what has been written. E.g the pitch deck was reviewed by 70% of the studio. Does this incur a large overhead? Not really, it took everyone around an hour and the amount of buy-in from the team that you get is in our experience worth it. That is not "the article Tommy wrote for us", that is "OUR article about OUR game". Most of the things written stay in the internal docs and never make it out.
You might have noticed that the processes presented require two things from everyone involved: Courage and humility. Technical competence has been proven to be learnable in our project, but those people need to have the right character.
I hope more people experiment with this. It is worth trying even if it doesn't succeed. In my opinion at least.
Amazing insights! Thank you.
Ok, I'm copying this to my notes so that I always have reference to that!
I notice a problem with your approach.
What if your artist comes up with a really innovative and good idea that will work if it shipped but no one at the studio likes it?
You have a system that takes it granted that whatever the group wants is best without question and there are no things that group doesn't know.
I recommend taking outside input as well when deciding about innovative approaches.
There are also people shunning the unpopular person because they don't like them even though they do good work.
There is also the problem of factions forming so people who want something made into the game ask others to join them saying they will support their ideas etc...
@@dragongoddragneel7106 in any system that ideia would be ignored unless the artists is the leader, in this system more ideas are considered in the first place
@@devforfun5618 it becomes game made by a committee from that approach
Personally (and from experience) I think that virtually everyone whos worked on any sort of large scale project intuitively knows that the hierarchy of people actually doing stuff is required and (again, almost) noone has a problem with that. As you say, the definitive answer from a lead, or the ability to on-board people at a junior level just makes sense. The problems in all cases I've observed is when hierarchies form among *people who organize people*, and somehow they wind up being more important than the people actually doing the work.
That's harsh to imply, that people who organize do not "do the work". In my little team the number of people started growing recently, and my responsibilities shifted from actually doing stuff or even simply spreading vision and ideas, to managing processes, organizing calls and so on. And often it takes up to a half of my work time on the game. And I can't say I like it a lot. I'd gladly pay from my own pocket to someone who could do it instead of me and probably be more professional at it, but for now I have to do it myself, because for the money a can offer no one would do that kind of job.
Although, I see your point. The credit in terms of compensation shouldn't be overbalanced towards managers, at least when there are layoffs in the company while top management still get paid their bonuses even when projects do not perform that well
protect yourself, once you open another business things change drastically... suddenly you're not the one hiring or deciding how things are structured :P@@Noowai
"Interview with a Boomer CTO in 2023" these are the people your company will be managed by xD @@Noowai
@@Noowai Exactly, I'm not saying that there should be noone to do that organization, it is very useful to take administrative load off the productive teams. Also note that I'm not including project managers who are involved in the actual projects, which it sounds like your role became. I'm talking more about large businesses where you wind up with multiple layers of management and 'project management teams' who only do Jira or Excel shuffling and treat the 'art/code monkeys' as replaceable resources.
I agree with this take %100. I've worked in flat hierarchies before and it's a real challenge to get firm decisions made and move forward towards the game vison.
I know that with Valve, their flat hierarchy is supposedly a nightmare where roles still exist but now they're all unknown and hidden and it's complex social structures and cliques
I don't think I could do that, there needs to be people who do the thing that they're best at when it comes to anything sufficiently complex
Human nature with no structure usually devolves into high school style cliques. It gets especially hard when people don't know each other and are not on a first name basis. Arcanum was able to work because they only had 14 people who could all talk together and fit in the same room
cliques are hell.
There must be roles because surely a group of valve fans-come-developers would make the next HL installment. Someone in there is preventing certain projects from happening or green lighting others.
@@koalabrownie This has happened several times but there are people who, though they have the same role as everyone else, they're in good with people like Gabe and nothing gets done without their say-so
The roles exist but only in the anarchistic sense of there being a hirearchy
Valve is a flat structure and everyone is equal, but as we all know, some are more equal than others
@@sandwich2473 So pretty much animal farm
Thanks for sharing Tim. There are hybrid models that integrates vertical with horizontal organizational structures; ultimately you do need a nucleus to act as a brain center for decision making within the cellular structure of an organism/organization.
Pay on a flat hierarchy, manage with a hierarchy. In the lean years everyone would take a hit in pay evenly, but would make bank in the fat years and it would have that organization needed. But hindsight is 20/20. Like in everything.
At least you guys tried something. More than I can say for myself.
If you pay flat, you can't hire juniors or keep people who aren't at the top of the curve. No one wants to be making the same as someone who's grotesquely less productive/competent. At one of my jobs, we had a girl with an EE degree do the simple web interface and it took her a week to add a button and have it work (with some help from a senior) in, iirc, Tomcat (yes, that long ago). If she were making the same as kernel devs, we'd have had to get rid of her, but she wasn't and there wasn't enough work to hire someone faster. It's a pathological case, and we ended up with a bad interface so I would've fired her anyway if it were up to me, but the CTO didn't have the heart to do it. Money isn't always hugely important to people in a general/absolute sense, particularly the sort of people you want, but being compensated fairly for their relative contribution very much is, and you can't even buy them off with promotions/increased responsibilities because they often just want to build things and not manage.
I swear that "ego" is a learned experience for people who want to work in the entertainment industry. Being the best specialist you can be is reward enough when you yield consistently great results and you understand what it means to be an integral member of a team.
I used to work on a modding team where everything was decided by "consensus," which no one on the team could tell you what it is short of everyone agreeing on something. As you could imagine, things got caught up in gridlock a lot, and people's motivation and effort would slowly peter out until the only people who were making decisions were the ones who were stubborn enough to keep attending meetings.
I think this bleeds into what you talked about in the "Game Development Caution" videos. Hear me out:
The problems you described here, are not necessarily problems with flat hierarchies. They are problems that arise from lack of organization. Having specialized people or teams who do certain kinds of tasks has nothing to do with hierarchy, the speed at which people finish tasks is orthogonal to hierarchy and it's not necessary to switch the guys between meetings with the same publisher.
Hierarchies are an organizational structure where responsibility is torn apart between higher and lower levels of decision making. I want to focus on two problems:
A) Power can lead to all sorts of abuse and problems, especially if power itself becomes a goal. Decisions become "political" instead of being driven by problem solving and empathy.
B) The powerless can get disconnected as they realize (or think) that their only agency is to stick to the rules that they have been given. Again, to the detriment of the actual goals of a project or team.
Leadership arises naturally and is dynamic. Alice has a problem, Bob sees a solution (due to familiarity/experience for example). Alice is smart, so she grants Bob the lead for that task. They solve the problem together, Alice learns something new and praises Bob, everyone is happy. It shouldn't matter whether Bob is an intern, or a CEO. Bob should feel _empowered_ to step up. Alice should invite Bob and put her learning hat on.
The problems I listed above, are stumbling blocks and barriers for this kind of dynamic leadership. Especially if you add metrics, box ticking and other stuff that distracts from open collaboration.
Congratulations on finding a unicorn named Alice. I did crisis consulting, where I would be engaged once things had clearly gone very badly wrong in an objective sense (services down or slowed to a crawl, massive ongoing losses, so my fee started looking reasonable). In my years of doing it, there was exactly 1 developer who thought "we f-cked up, this is an objective fact, now this guy with tons of experience and earning astronomical money, presumably for a reason, comes in, tells us what and how to fix, and things get objectively better, I should probably learn from him". One. Yes, this is a biased sample, but.. One. When I started doing it, I tried building consensus, teaching people things, explaining the entire thought process etc. That just p-ssed people off more, because they couldn't live with the clue differential. When I once had a clue differential, I couldn't live with it either, so I set about bridging it like a maniac by learning things. This is unicorn behaviour, as it turned out. Vast majority of people pretend it isn't happening, lie to themselves, sabotage everything, the list goes on. Competence, it turns out, is a direct result of being able to evaluate your own ability and a drive to improve; incompetent people can't do the former and refuse to do the latter, almost universally.
@@paulie-g Thanks for sharing this story! Just the job you decribe is interesting and exciting to me, even though you had to fight a lot of frustration.
I appreciate your perspective on Alice being a unicorn. But isn’t the context that you wouldn’t have been called in the first place if she wasn’t? Like a psychiatrist would think that everyone is crazy or a cop would think that everyone is a criminal.
What I decribed above is something that is preceded by hard work and established relationships. It’s something that has to be earned and made as a goal. Me personally, I‘m rarely achieving that ideal, but I try.
In any case, I‘m fascinated by your work as a crisis consultant. Never met one, but even just the idea of it is intriguing.
IME all groups have a hierarchy, it’s just a question of how explicit it is. It’s just human social psychology, certain individual voices will carry more weight for whatever reasons.
While implicit hierarchies _can_ be good, there’s absolutely no guarantee that they will be; they can be structured around charisma or bullying instead of competence, and can be oppressive to people at _all_ levels. E.g., I’ve seen “flat” hierarchies where one person is declared the “leader” (not necessarily with their consent) and then has all of the responsibility for everything dumped on them… Even in the example of Troika, it obviously wasn’t truly flat if three people had to take massive salary cuts to keep the whole thing afloat while everyone else was blissfully unaffected.
Of course, formal hierarchies are often pretty dysfunctional too. But with a formal hierarchy, at least the potential exists to identify and address structural problems within that hierarchy. And it’s certainly possible to have a _relatively_ flat formal hierarchy which encourages and rewards merit at all levels. You can have a group which strives for consensus and only falls back on formal hierarchy when necessary. Heck, you can even have a workplace democracy, where everyone periodically votes on how to restructure the hierarchy. But at the end of the day, the buck stops with someone, and it’s wise to deliberately choose who that is.
Thanks for sharing Tim, this is especially useful in light of Valve's continued success and famously flat hierarchy. There are rumors that Gabe personally stepped back from day to day development because of the inability of team members to see him as equal, instead they treated his word as divine and final ultimately sidelining any attempt at feedback for a person who is heavily invested in the flat arrangement. While I'm sure it is possible to be successful long term with a flat hierarchy, I can't help but think it would only be at the hands of a very special group of people, and who would be just as or more successful under a different circumstance.
When anarchists in Catalonia and Aragon took over industry during the Spanish revolution, productivity went up as soon as they showed their bosses the door. Maybe the context has more to do with it than any very special group of very special people.
@@ReadyAimSing so a group of ideologically committed people worked together to establish a scenario that met their ideological goals, one which is almost universally atypical and you're telling this *wasn't* a special group of people?
It really seems like in order to get a flat hierarchy to work, people need to have a perfect understanding of their own limitations and must have the humility/lack of ego necessary to admit this. Basically the exact type of person who would also do well in a hierarchical structure.
No. Hierarchies are really just not workable in the real world. You convieniently left out that everyone hired has to be at the exact same skill level and that everyone has to get paid equally. The onnly reason they could even attempt this is they were brand new. 5 or 10 years down the road it would be literally impossible. If you have someone who has been working at the company for 10 years and a brand new person is hired and brought on at equal level and pay, that is not fair or reasonable. In the real world people are not equal. There are 25 year olds who know little starting out and older experts like Tim. So, tell me, which group do you never hire?
Bingo. Flat only ever works in a (small) team of A-listers who are also great people. Incidentally, based on my anecdotal experience, such teams tend to self-construct an informal hierarchy where they voluntarily give final decision power to someone (or multiple someones, one per sector of responsibility).
This is fascinating and a great look into how flat hierarchies function within indie studios and how they could be improved in the future. Your knowledge is invaluable Tim!
flat hierarchies would work with Aliens who aren't made in the image of God, or ants or bees.
@@chomnansaedan4788"ants or bees" sorry bro we aren't made in the same image as ants and bees or whatever your version of God is that is conveniently more real and valid in your head than other people's versions of God, you should really read or watch more about biology or anthropology before making bizarrely baseless claims on the subject of biology and anthropology in front of others
@@chomnansaedan4788"ants or bees" sorry bro we aren't made in the same image as ants and bees or whatever your version of God is that is conveniently more real and valid in your head than other people's versions of God, you should really read or watch more about biology or anthropology before making bizarrely baseless claims on the subject of biology and anthropology in front of others
Your personal experience videos are invaluable! Many dream of making a horizontal hierarchy, however, miss the fact that everyone has their own productivity capacity and motivation, their own personal responsibility level, and, finally, many need a mentor or a leader above them to do their best! Thanks for sharing!
I work in a small town school with a somewhat flattened hierarchy and it's a life changing experience coming from industries where the opposite, top down distribution of ideas is ubiquitous and set in stone. Admittedly there is more basic structure to guide people in their roles and duties than what you describe here, but having the ability to get your concerns heard on a regular basis by an entire staff who is all pulling in the same direction is what keeps me coming back to give it my best even though the pay isn't as competitive.
Feeling like you're a contributing member of a community who sees the benefits of the struggle to produce a positive result vs. feeling like a mercenary or even worse, a serf, changes your work and how you deal with coworkers for the better.
I had a professor at one point sum up his experience with flat and vertical authority structures as the difference between valuing long-term growth and professional development vs. achieving efficient and viable short term results. When it's a luxury you can afford, I think experimenting with ways to flatten your workplace hierarchy is a great idea.
I did two years at a startup with a flat hierarchy. At first it was a lot of creative freedom but it quickly became a planning nightmare with no one ever knowing who had final say.
It's an interesting challenge to close the gap between people's desires to believe in perfect equality (not the social kind, the capitalist kind) and the reality that some people can literally produce considerably more than others at the same quality. Some people are simply worth more when the metric is, "getting the job done correctly and on-time."
I've managed a lot of underperformers, and you know who end up with the best career outcomes? Those who acknowledge it, take responsibility, and work together with me to maximize their ability to contribute. Those who externalize by blaming others and making excuses always ended up eventually being let go.
Rather than flat hierarchies, which ignore so much about what each individual contributes to a project, I prefer a system that respects the real contributions of each individual. I spent some time at a company with a near 1:1 ratio of producers/project management to designers/developers/artists and it was clear that producers/project management were compensated at rate far beyond what made sense considering their contribution. Some producers/project managers were there with us on those nights when we just had to crunch through to hit a deadline, and we all loved them for it. The team gets to know who coasts and whose backs are coasted on.
All of this to say I wish that there was more transparency about compensation vs contribution. When you're doing a rundown at the end of a project, and you see quality contributions across several domains all over the place from one member of the team vs a few contributions, even high quality, here and there from another, it isn't hard to understand why one might be compensated more than the other, and most folks on the team aren't going to argue that that's unfair.
I never really thought of flat hierarchies the way you're describing them. I just want to avoid having some BS chain of command I have to traverse to get the answers I need or help someone in a different domain. I want to go to the person I know has the answer or who can point me in the right direction. If I see somewhere I can contribute, I don't want to ask a manager to forward a message to a lead to give a suggestion to an artist. I just want to go to the artist. I really hate the slowdown that can come from a deep, siloed hierarchy. If I wanna chat with the programmers, I don't want to be locked out of doing so because I'm in the art dept.
Huh? I've never worked at a company where staff couldn't talk to each other and had to communicate through managers. Managers are there to make important decisions,resolve conflicts, approve your vacation time, do annual reviews,, etc. If staff are working on something and have questions there is always ways to directly communicate in my experience. Some people do hide behind their managers, but that is pretty rare.
Interesting experiment. Arcanum turned out amazing so despite your issues the team created something legendary.
"from each according to his ability, to each according to his needs"
It was a nice experiment to have, and to hear about!
I conclude that hierarchy is not a bad thing, direction, leadership, are needed for organization, it is a matter of logistics. But being Hierarchical that can be bad thing, when a structure needs to enforce aggressively the verticality of it. Even collective companies and movements need a head, not to have exclusive power of decision, but to have a trusted final word, and a proactive interest in making sure all are doing well, and know what they are doing, to the best of their abilities.
Everyone being given the same amount of time and work and expecting the same results is utopic, idealistic. For good or bad, we are different, and we become different from ourselves too with time and situation. But everyone can be very good at something of theirs, and best using that (while not chipping away other interests) is the way to go for projects. With that, you are totally right, Tim! While far from your amount of experience, I've seen in games I worked on how it is not good to make everything a group decision, or to let everyone meddle with one another's work, that is for their department/group to do, and their lead to settle if there is an impasse. Especially when the whole team is not made-up of people who know each other.
At last, I must say that doing the experiment from scratch, inside a system that hinders in every way it can a collective owning and effort (and ideologically is very individualistic and hierarchical system), will be a Sisyphus effort. Everything, from the ideas implanted on everyone from the culture, to the tangible corporate structure, rules and culture as well, is vying against that.
Even so, I think we can all appreciate the effort at Troika for trying it, and congratulating them on the masterpiece they delivered while doing the experiment!
That quote in this context is interesting because it implicitly rejects both the equality of ability and the equality of needs, otherwise it wouldn't need to mention either. And it sounds like Troika tried to do this but didn't want to let go of the idea of determining rewards by the same metrics as preexisting hierarchies, and they tried to resolve the contradiction by simply making sure that everyone's abilities in the company were on the exact same level. And obviously that didn't work, because it can't.
So here's the thing. Re-read the phrase "From each according to ability, to each according to need" carefully.
Nowhere in that maxim does it talk about "equality," about everyone being the same. In fact, it rejects and refutes that whole idea. "From each according to ability" implies quite clearly that people have different skills and different abilities. But "to each according to need" separates their "performance" from their ability to secure the basic needs of life.
Just throwing that out there for you to think about.
And, full disclosure. If you didn't figure it out by now, I'm a communist :)
It really was not fair or equal though. The three owners were on the hook for the rent and the contracts. If things didn't work out they could have lost their houses and savings while the employees would have walked away scott free to their next jobs. There is risk in new ventures, often owners need to put up a lot that they may lose which is why they make a lot more than employees if things suceed. Employees never stand to lose anything more than loss of work, they have no actual risk.
@@PaulRGauthier Yeah, so in practice advanced Marxist socialism failed to produce a classless society. New classes formed on the basis of things not related to economic relations. It's possible to culturally enforce 'wants' being much closer to 'needs' than in capitalist societies, but in practice that turned out to require prohibiting people from leaving the country and concealing information about comparative living standards, even in the presence of some hardcore social engineering. It's a nice notion, but frankly we're past that in modern industrialised economies - they are well in surplus, so solutions to scarcity problems are a bit anachronistic.
The experience is really valuable to understand the problems faced by the idea. That's the best kind of information if one wants to try something similar in the future.
One of my favorite videos so far. Since I'm interested in both game development and anarchism, loved to hear about an actual flat hierarchy experiment in the industry. Even though it failed in the end I believe such attempts are meaningful to create a way towards better forms of work relationships.
It doesn't surprise me that a flat structure doesn't work well; but I think the problem that it is trying to address is valid. Hierarchies and especially top down dictatorial ones can be problematic. Yes, its possible to have a 'benevolent dictator' for a time; but a team structure that requires one doesn't normally protect itself from a non-benevolent or authoritarian dictator (and are dictators really benevolent or are they just charismatic). Having a flat or just a more democratic structure; you aren't going to avoid problems, you are just going to trade problems for different ones.
This is super interesting! I feel like a bit of hierarchy is essential as you've proven; but my problem is a lot of businesses take hierarchy as an opportunity for corruption and nepotism. A bunch of people get shoe-horned into management roles that don't need to exist and basically are there just to devalue the people who are actually doing the work. So to me a flat hierarchy can still have a lead, or maybe a 'senior' person who can make a final call and keep order. It's a way to avoid the all to common situation where if you need approval, have a question, need help, or anything you have to ask your lead, who has to ask their manager, who has to go to HR, who then has to go to upper, who then have to go to the real in-charge people who have no idea what you need or what's even a good response and frankly don't care anyway.
It's the difference between shallow (often and confusingly called flat) hierarchy and no hierarchy at all. Hard to make the last one work with humans since we are naturally hierarchical, but I completely agree with you that leadership is what makes hierarchy work, not positional/imposed authority.
@@JavierBonnemaison "we are naturally hierarchical" -- Sounds like something Jordan Peterson might say, though some of the greatest minds of our time think him very wrong about that among a few other things, perhaps especially his confusion of individualism with iindividuation. (It's an understandable confusion that is even found in the writings of luminaries such as R.W. Emerson at times, but is a confusion nonetheless.)
Strictly hierarchical organization is not a "force of evolution," but is based on a dying worldview whereas Helena Norberg-Hodge, et alia, is quite correct when she says, "We human beings didn't evolve for a hyper-individualized, competitive way of being. For most of our time on this planet, we lived in collaborative, intergenerational communities." We know this intuitively, if in no other way, and indigenous communitites are generally exemplary of it. Some might say Helena's, among others, is a "feminine" sensibility, but it's actually a sensibility we all possess to one degree or another. And anyone who has worked within a strict hierarchy knows the feeling of being treated as "just another cog in the machine."
The aforementioned worldview may be prevalent and predominant, but is in no way natural. As Deborah Frieze notes in her TED talk, "Somewhere along the way, maybe around the time of Isaac Newton, we got confused about how life works. We convinced ourselves that the world was causal, linear and predictable and, so, we began to treat our bodies, our communities and our ecologies as though they were machines." This "techno-logical" wordview or "mental-rational structure of consciousness" (in Jean Gebser's terms) pervades every aspect of our lives and human endeavor from business to education to government to healthcare and reductionism is rampant to the point that even some scientists, e.g. Adam Frank and Marcelo Gleiser, are decrying it in their respective fields.
An earlier comment by alexpetrovich85 mentioned "hybrid", integral models of organization we're beginning to see emerging here and there. Personally, I think such "models" deserve our attention and nourishment as the modern era naturally passes away and a new era is born.
@@JavierBonnemaison thank you for replaying! I think this is what I'm looking for. I've never heard the term 'shallow' hierarchy before!
@@lrinfi I don't think it's debatable that we are hierarchical. You have to do some serious cherry-picking to come up with an example of a properly functioning society that existed which didn't organize itself hierarchically. If you can even come up with any.
Most of what you're saying goes against observable reality. We are collaborative and competitive, and we've always been, including in tribes and between tribes, and we'll always be, because all Life is collaborative and competitive. And most, if not all mammals exhibit some degree of hierarchical and individualistic/tribal nature. Humans are not outliers.
We act according to our nature, and the proof is everywhere around us. No amount of pseudo-philosophy can argue against observable reality.
@@lrinfi TED talks are not peer-reviewed. Hierarchy in humans is a broad consensus in evo- and *-psych as well as anthropology. "X might agree with this, therefore it is wrong" is a logical fallacy, regardless of how odious you think X is.
i remember some 6ish hour podcast with John Carmack, him takling how this didnt worked because some ppl work more, some less adn where theres no boss to call the shots it turns into everyone doing his thing if they feel that the collective decision does not fit them
The problem is that there's always the two extremes. You have the typical big company structure where those at the top make 100x the amount of those at the bottom and while they have more responsibility they don't have to do anywhere near as much heavy lifting as those at the bottom and when the shit hits the fan because of decisions made by those at the top it's always those at the bottom who are let go so it's responsibility without much consequence. Likewise this extreme was never going to work either for all the reasons outlined by Tim. There's a happy medium which rarely gets implemented because the obstacles are either the greed from those at the top or on the other end people exploiting the good nature of owners/leaders who try something different which leads to poor work, a lack of direction and poor work ethic. It's a very difficult balance to strike and I applaud these guys for trying something different but any successful project will need clear leaders, delegation and square pegs in square holes.
In general dev I really like having a product owner who decides how the product will be but doesn't involve themselves in the code and a lead dev to make final decisions on code related matters rather than a project manager who has to decide everything. It's flatter but you still have someone who can say yes or no.
Obviously, the history of experiments in human organization without any form of hierarchy is long and goes far beyond the game industry. I'm sure part of the reason people react with "but it would've worked if you did XYZ" is because the subject matter overlaps too much with their opinions about wider societal organization (and thus draws the conversation towards more emotionally intense beliefs).
Well said
Ooo, there is so much good stuff you go over in this video. Among lower-level employees, many don't want a higher level of responsibility. They may enjoy working, but they don't want the responsibility if something goes wrong or have to make any difficult or important decisions. Conversely, others who want to lead may be horrible leaders. Customers, vendors, etc may feel most comfortable interacting with a particular individual. At my current job we sometimes have customers who would rather talk to me than the sales/customer support team. Problem is, I can't do any of the help or contract changes that the sales team could. Not to mention, every minute I spend talking to a customer is a minute I'm not making the products the company will deliver to the customer. I can explain technical details that the sales team don't know. There are other people that want to talk to me because they think they can get a better deal doing that. I can't give them the discount they want. Hell, I would charge them more it was up to me. You have the mommy/daddy issue when you have people at the same decision-making level. People will ask one, and if they don't get the response they want, they go ask the other.
"When three people set out on a journey, they should appoint one of them as their leader." -Hadith
Valve is allegedly flat hierarchy, but if that's the case why hasn't a puddle of employees made Half Life 3 yet? There must be someone saying no.
I don't think I would do a completely flat hierarchy if I had a company myself, but there are experiments that I want to do to incentivize developers, such as flexible hours and still get a part of the sales revenue, and a loose team-based approach to development, wherein developers organically form teams to get things done.
I've always wondered how Valve makes it work, though I guess it helps to hold a near-monopoly on digital PC game sales.
Ca-ching! Exactly! It works for them because they have a money printer. Aka, it doesn't actually have to work and they can be wasteful.
Yeah, tried this as well when me and a couple of friends/colleagues created a company doing VR games. It worked about just as well. Not that people didn't have skills, interesting things to say across the board, or passion, but it went down exactly the same. I think if anything, everybody who's naively optimistic about flat hierarchies should try it to gain the understanding of why hierarchies work, and why it's more fun in the end. It strengthens you as an individual, and it strengthens the team if everyone is willing, but it is not a sustainable model for coherent games. Not that we were a team of 15 people when we were at max capacity at the time, but personally, I think even with 3-4 people, hierarchies are still preferable. It's easier to do overlap, and own certain areas, especially the broader skillsets you all have, but having someone with the final word is always great, to lead the vision of a project.
When everyone is responsible for some decision, it actually means that no one is responsible. So yes, totally agree -- structured hierarchy, which implies clear responsibility for a particular area of the project -- is a way to go.
When everyone is responsible, it actually means no one is responsible - exactly! Well said!
Troika is my biggest what if in gaming.
Thank you for sharing this! I seem to be noticing a lot more people discussing alternative business and work structures these days, especially after the pandemic, so I'm sure this will be helpful for some upcoming companies.
Working for a start up that had recently turned into a huge corporation, I experienced various levels of hierarchy. By far, less hierarchy is better. You need to take into consideration the massive morale boost of giving everyone the respect of being equal. Direction is needed, not management.
Love each and every one of these insightful videos.
Just wanted to note that even the purest anarchist societies have hierarchies on the job / hunt. The thing about them is that they fall away after the work is done.
Culturally, we're a while away from it (you can try building any kind of structure under capitalism, but you're still drawing from a pool of people raised by capitalism and operating under the rules that ultimately reinforce it [bank loans, publishers, property laws]) I think democracy in the workplace would be a better way to go then having a flat hierarchy.
It wouldn't be perfect, nothing is, but being able to elect managers or specialists (i.e. communication) based on their performance (and recall them) might create a more fair environment for everyone. It would probably lead to a lot of problems too, but we have democracy in public life and most would agree that it's preferable to the tyranny of kings.
A huge cultural shift and boost in education would be required to get this kind of workplace running, but that's the dream.
Sorry for getting political here, I just found this video particularly interesting! Getting people to do the same amount of work can be challenging as people can just be jerks or slow. Speaking as the latter type, I'm particularly intimidated by work made by legendary figures in the industry such as Tim et al., but these videos have inspired me to keep trying no matter what.
Have a great day everyone!
All you needed to do was one group assignment during highschool to know this would be a bad idea.
7:55 You *can't* say that? "No. I'm pushing back on you, publisher."
Why do I think you *can*, but it's perhaps not considered advisable in the present industry environment and atmosphere. After all, the publisher just might decide to say, "Well, then, we're not publshing your game. Nyeh." (Or pull some other such juvenile attempt at "power play".)
Any alternatives to bowing to the whims of the major publishers on the table anywhere?
Interesting that Arcanum was made under a flat structure, I wonder if the flat hierarchy contributed to the "everything and the kitchen sink" feature set. Still one of my favourite RPGs ever, love the setting especially.
I imagine whether a flat company structure is viable depends heavily on the people involved, the financial strength of the company, etc. The most famous example of a flat structure gaming company is Valve and it has been working well for them for decades, so it is possible.
I'm not sure I would say its working well. They barely make games anymore, if it wasn't for Steam then I have a feeling it would have fallen apart since there isn't much guidance for what to do so people have a hard time getting together to actually make a product. Probably a big part of the reason that Half-life 3 is a meme.
@@Worgen33 You're right, it's because that flat structure that Valve barely did anything, as I understand, Half Life Alyx very much required to abandon that in favor of a hierarchy to get anything done. Also the lack of a formal structure doesn't mean there's a no a de facto one. Back in 2013 Jeri Ellsworth said it very much is like a high school, the popular guys are the ones that have all of the power.
@@Worgen33While I love the Half Life series as much as anyone, I'd rather take a pile of cancelled projects over the sort of soulless yearly release that gets pushed out the door on time regardless of quality. I used to consider Far Cry one of my favourite series before FC3 came along, I don't want to see Half Life go through that.
@@AndrewFullertonExcept that Valve has enough money to not need to rely on sequels. They can make more original IPs if they want, but with a flat hierarchy they can't get enough motivation to team up and even do that. I think the Deck is the most motivated they have been in a very long time.
Thank you so much for this video Tim, a lot of fascinating ideas & results to chew over! I'm taking all the notes on this one lol. Glad i got to learn about this secondhand and not first hand because a flat hierarchy is something right up my super idealistic alley lol
Love you Tim!
We tried a flat hierarchy system, too, with the group I first tried to make commercial games with. It didn't work out at all, never released anything but a couple of jam games.
10:31 "It just worked...." -Timothy Howard
A totally flat hierarchy would be difficult with more than, like, five people, but there might be some hybrid approaches worth trying.
Perhaps one day ill get to try it out. I love the idea, but perhaps there are more lessons to learn from it than practical applications.
Hey Tim! At 11:32 you mentioned random modules shooting events at each other. What techniques do you like to use to share information between modules in code without tight coupling?
I think there are many examples of this actually working, but its quite difficult to get off the ground. Mondragon,valve,etc. While history for the most part has been pretty terrible to more co-op organized orgs i think in the long run and how the world currently is this may actually be the superior way to do things. Being an early adopter of new systems is always difficult.
Valve is a rentier operator of a service that has had a quasi-monopolistic position until recently (arguably even now; the lengths to which Epic is going to try to compete demonstrate how high the barriers to entry are). So they can almost do whatever they want in terms of internal management because they aren't subject to normal market discipline.
I think they have a very low number of employees for what they are doing, but even so, from the outside the way they operate Steam seems very stagnant and inefficient (e.g. buggy or poorly-documented APIs persisting over long periods). I doubt their success can be attributed to a lack of internal hierarchy, and I think the main reason it persists is network effects keeping them at the top of an external hierarchy.
@@globalistgamer6418Yep. It's like if one location of a co-op coffee shop was pulling the revenue of every Starbucks in the country. They could have whatever structure they wanted; there's too much money coming in for them to fail anyways.
Often the advantages and disadvantages of a co-op don't stop broad market effects or don't stop a bad company from doing badly. And they ignored the positive effect it has on its workers. A key example of this is a study done on how co-ops handled 2008 and found that they were hit harder and earlier but then recovered faster. Workers voluntarily furloughed themselves to keep the co-op going which allowed them to recover and survive. But plenty also decided not to. Just like how some traditional businesses better prepared for such a crises and acted fast to keep things afloat. Plenty of co-ops went under but so did plenty of traditional businesses and what is often ignored is that workers in co-ops are better off, hence them willing to furlough themselves even when the future is unclear and it might be for nothing. Having a say in your employment offers you better conditions and pay and in return employees are more loyal and do more to keep things going. Research has shown that at the very least co-ops are better for the employees and either don't effect productivity/performance or make it slightly worse/better. The advantage of co-ops will always be the fact they better distribute wealth to their employees which make the community wealthier, which then makes the co-op more stable during down turns since workers and communities are more likely to bail them out but I've gone on long enough.
Valve is a flat hierarchy only on the surface. Beneath the surface it is basically a meritocracy. People who matter dictate what gets done, everyone else has to go along with it.
A funny thing about the human mind is that we can believe anything; nevertheless, there are other aspects of human nature. We are indeed hierarchical animals, yet our strength lies in reciprocity and cooperation. I don't think rigid and deep hierarchies work well, but neither does full egalitarianism. The ideal balance for a particular group of people is very much context specific and dynamic, but it will always be closer to the middle than to the extremes. After all, this is why humans are so adaptable.
Man, the alphabetically listed credits is an incredibly subversive and very cool thing though.
I've only seen it one other time, and that is for Drakengard 3. A game that was definitely made is a hierarchical structure, but those credits just completely floored me. Even the company itself had to wait until S to be named. And the lead designer was almost the last person in the list.
I heard Resident Evil 7 did a similar until the middle of development thing expect they had roles and hierarchy
its just that almost anyone could pitch ideas middle in development they switched to
a more assembly line work flow
that game went fast only took them two or three years to make
I wonder if this "failed" structure is, in the end, a big contributing factor to what made Arcanum so great. The secret sauce, if you will. Don't get me wrong, I love VtM: Bloodlines as well (and I admit, I haven't played Temple of Elemental Evil... yet), but there's just *something* about Arcanum.
Is it buggy? Hell yes. But now that I understand the work structure behind it, it's dawning on me that maybe Arcanum is so special, because your structure back then was more akin to a passion project of a bunch of friends getting together to make something great, instead of just a bunch of people being hired to do a specific job. As you've said - people tried many hats on, and that makes me think that perhaps Arcanum is so great specifically because every tiny bit of work that went into it, was work that people *wanted* to do, not just did for a paycheck.
Thank you for sharing. You call this structure a failure, and perhaps in business terms it was, but it certainly led to something amazing.
valve didnt fail
thats really cool insight you shard. wsh you guys had had the time and resources to investigate how to do flat hierarchies or similar models that try to flatten power dynamics in the workplace, e.g co-op models.
I apologize if this is something I am simply unaware of, but have you had any experience in developing TTRPG’s, or a desire to? What are you thoughts on the most recent iterations of D&D and Pathfinder?
Happy Halloween, Tim!
It's quite funny that alphabetical order, a totally reasonable choice, still inadvertently puts Anderson, Boyarsky and Cain at the top.
Haha, I did notice this too and was amused.
This was a great video thanks 👍
Hi Tim, Valve is famously a flat heirarchy company, and they've been doing it for a while now. They released a PDF outlining how this flat system works, no idea if you have seen that or not. Do you think they faced the same issues that you have faced? Or did they do some things that in the right way that made it work? From our side (the players/consumers), we see the benefits of this system (if a dev is passionate about something, they can spend more time than they need to on something and make it very good/crafted with love) but we also see the cons of this. Famously, projects/features get neglected despite having a large playerbase. Team Fortress 2 for example has been in the Top 10 most played games on steam for a long time, yet has bugs, riddled with bots, stopped receiving official content, and may have either 1 or 2 devs working on it (or 0). Because their devs can work on whichever project they want, but every dev also feels like working on the next "big" thing at their office. Were there any benefits that you enjoyed from having a flat hierarchy at Troika that made the game better for the players or was it overall a negative thing?
The "problem" with Tim's structure was that it tried to be both flat AND open. You can have flat when you keep everyone restricted to a specific set of tasks and make all tasks equally valuable. You can have open when you have a hierarchy in place to keep people from wasting time and doing things they shouldnt, but also to approve and encourage people to grow their natural talent they show. Both simultaneously is a trap as they found out.
Still really happy that the Troika owners put the idea of employee value forward and really ran with it. Even if it was a mess, it was a glorious one.
I don't see Valve as a broadly replicable system - it's easy to run a commune when you can only handpick the most intelligent .0001% of the world's population to populate it and you never have to worry about funding/project deadlines since you have guaranteed income from your monopoly on the gaming distribution network.
You realize Valve has been around since 1996 and Steam wasn’t the first thing they did, and it didn’t really take off immediately when it was released in 2003, when it was just a platform for valve games anyway? And the first thing Valve DID do was a game that killed dozens of competitors on arrival and God knows how many companies just tried to copy them for years after that? And they did it again with the sequel?
Depends on who you ask how well Valve's setup works as well. PeopleMakeGames had a video where they interviewed some employees.
@@OlStinky1 Exactly. Valve can basically do whatever shit they want now, as long as it doesn't undermine Steam. Games, software, hardware, profile backgroinds (lol).
I think being under financial stress is a particularly bad time to try a flat structure and probably contributed a fair bit to the failure.
I think this thought is interesting because in the macro politics it's people under financial stress that tend to advocate for flatter hierarchy.
I have no statement beyond that it's just food for thought.
I've long wanted to participate in a co-operative democratic structure where there is a hierarchy, but there were structured means for people lower on the hierarchy to vote on the ratification of higher level decisions and positions. For example, people would be nominated as creative director, art director, lead programmer, for a project or for a two year term, and they would have the authority to make decisions at that level. They would perhaps hold veto power over team-wide ratification votes. I have many more specific ideas of how to integrate balances into this, but it's a little much for a youtube comment 😄
There would of course be issues with new problematic dynamics in the workforce, but it would give lower level employees more say in what they're doing when they felt strongly about something, while also giving operational structure and clearly defined roles to make day to day work flow better.
Yeah, Tim and the guys are the kind of dev that I would root for their success. I kinda wish Tim born a bit later so we can have him still leading more projects today. I would love to play more games from him. Heck I want to work with him 😁
I think its great that ya'll were trying to innovate at the company organization level. Wild takeaways.
Loved Arcanum and Vampire, I was Sad Troika had to close
I've always thought that a flat hirearchy doesn't make a lick of sense. You need people to be in charge and to lead. It's just chaos without it.
Valve is a famously fully flat company. Valve is also famous for being a brutal place to work, full of cliques, projects that get stalled or killed due to office politics, and using stack ranking to "motivate" people. If it wasn't for the endless money printer that is Steam, the company would have imploded years ago operating the way it does.
It sounds like ego was the focus on this experiment, which could have been its downfall. Instead of a flat hierarchy, it could have been a collective. Roles could have been filled appropriately while money was distributed evenly.
You've literally just reworded the flat heirarchy they tried and called it collective instead, as if that would fix any of the inherent issues.
@@BobArctor Tim described that the flat hierarchy meant you could grab any hat at any time and give it a go. A proper collective is more realistic about what a business needs, and doesn't worry about hurting people's feelings by plugging them in where they belong. But the pay is even. Tim said that Troika wasn't just even pay, but also even roles (hence the word hierarchy). It just sounds like it was a symptom of everyone (rightfully) feeling wronged by their previous employers, so they decided that they would distance themselves from that form of business in every way possible. It's an understandable reaction, but sadly one that ended up being a learning opportunity and not a full-on success story. Of course I wasn't there so what do I know? I'm just going off of what Tim had to say about it.
Sounds very similar to the same as Valve’s production style is. I wonder how they solve/mitigate these issues.
I think the right way to try a flat hierarchy is to have the participants do what they are good at 80% of the time and 20% flat hierarchy stuff. And the team have to be balanced from the start, those 80% should pretty much be enough to get the project done within deadline and the 20% is more of a educational part of your time and not focused on deliverables.
It reminds me of those commune experiments in the US. None worked without radically changing the way each commune worked i.e. bringing hierarchy back.
A look inside the development of games and the many ways teams can work.
I think it's pretty telling that even Valve - the most famous flat heirarchy studio has recently started to move to a more traditional structure. Flat heirarchy sounds really great in concept until you get into all the problems Tim mentioned here...
You know what else is flat? Water. Water finds its level.
That's what Motion Twin is doing too.
You guys ran it like Octavian and those guys, a triumver- thing...
To be fair...thers flat hirarchy and theres FLAT hirarchy, and this sounds like Nevada desert flat. There are definitely more nuanced versions out there, with something like only 2-3 layers for example, that do in fact work.
a flat hierarchy is likely only optimal in low growth environments, it can generate creativity, but hinders growth, in theory it should tend to have one sane person managing everything amid the chaos - some guy who works in quantitative finance and game theory
I think it's really cool you tried!
It's easy to interpret worker self-management as design by committee and ritualized bikeshedding, but I think, in reality, it can take a substantial amount of time and resources to figure out how to run a worker coop: where leadership roles make sense, how to have them without sacrificing workplace democracy, how to make those positions accountable and recallable, etc. It's not trivial, but a lot of these things have partial answers, owing to a century of anarchists, conscious and unconscious, actually doing them and seeing what works. What you describe sounds like an early, naive attempt at exactly the right idea. Maybe if Troika had the time and money to learn from it, it would have gone a different way.
I've been trying to explain this to my leftists friends (I'm on the left, but not a leftist). Having run a business for 15 years, I know that there's so many things that the average person doesn't understand about what it takes to make a business works.
And if the owners are putting in more resources than the rest, its' already not an equal situation. Most people don't want that responsibility.
I went from flat to top-down. The flat was WAY better in acting fast and creating a product iteractivelly, we always had roles, we had people with veto power to balance things, but even with those in place it didnt work over 50 employees. Dividing work ALWAYS fail, ALWAYS! The churn rate of employees was 50% PER YEAR! The average tenure of employees was 1.5yrs
Its hard to do flat and manage it, very hard...
I worked on a project where there was a structure but one guy private messaged everyone telling them what to make and basically nothing got done.
It's work. It mystifies me when people and organisations bend over backwards to try and make it like normal social life, happy funtime. It's awkward and never works (because you'd never socialize with half the people you work with). You can (and should) enjoy your work and appreciate your workmates but it's still a job. It's just so much easier to get s*t done when everyone knows the plan and where the buck stops.
If you read Polybius, it's clear he thinks that the reason the Romans were able to conquer the entire Mediterranean is due to how well-organized and hierarchical not only their military was, but the whole rest of their society.
A creative project needs a head-honcho that gives the direction, and can call a veto if it drift off course. Otherwise the project is directionless.
I don't understand why you would even want a flat hierarchy. Most of the problems you listed are pretty obvious a priori. What problems did you expect a flat hierarchy to solve and were they solved when you tried this?
I think the error is assuming that "everybody receiving equal pay and having the same amount of work" is a flat hierarchy. If you have this arrangement and it was imposed by 2 or 3 people in the organisation and the others had no say in these and other decisions, it's just a typical not-flat hierarchy, with just two layers.
At least as far as I understand it, a flat hierarchy is an organisation where everybody has equal say & vote, where everyone is up to speed of the company "health" situation and major external relations, and where ideally big important decisions are made by consensus or at least with a high portion of the members. These decisions might imply that not everyone will be compensated equally (except perhaps regarding benefits) but rather in function of the work being done and/or other criteria reached by consensus. It might imply that there are designated roles by (past AND present) merits instead of no roles or interchangeable at the whim of anybody. It also implies that these decisions CAN BE CHANGED if it can be shown they're not working.
Also, AFAIK, there are some kinds of company structures that allow everybody to be owners of the company in equal parts (or another kind of arrangement) and have one or more delegates to be legal and commercial representatives of said company.
But has there been a game company that uses the chain that binds...
I was also part of a small company that tried this. There were only four of us, and it still didn't work. Groups of people just work better with clearly defined roles and structure. Perfectly egalitarian hierarchy is a lovely pipe dream but it's just not the way humans are wired.
@@DawnRazor-eb9ow uh... no they have not?
@@AnvilOfDoom"Nazism is a valid approach because people have fought in a war over it" like wtf is this guy even saying lmao
I think though that there is a mindset that goes with this - just because someone is a lead and someone is a junior their contributions should be weighted by their position, in that sense most reasonable project based hierarchies are totally egalitarian. As has been mentioned in other comments the 'unfairness' comes when you wind up with management layers who are compensated more for essentially putting brakes on communication lines and shuffling around paper.
absence of hierarchy is in no way antithetical to roles, rules or structure
Very interesting. What made you guys attempt this? What was the catalyst for the decision to attempt it?
So, as Charles said: "All men are not created equal."
Wasn't expecting Tim to disprove Communism in this episode. Unintentionally based.
Could you tell us stories about the development of Tyranny? I really liked that game and think it had a lot of potential, it was a great experience to play through! (and finding other paths/branching storylines in next playthroughs) :)
I tell some stories about Tyranny here: ruclips.net/video/KZTkg50ihsU/видео.htmlsi=boO-RBhmIEViGB__
@@CainOnGames Thank you! New here, just recently discovered your videos, but I like it and how you approach game design. :) Best regards!
Applying this micro economic infinitesimal failure across a broad swath of economic theory like this comment section is doing is far from “based”.
Plus, see how employee owned companies work.
It sounds like you discovered that while the assumption is a hierarchy is a means of boxing people in, it also allows people to stay protected in their box. When you don't have the ability to say "we want one of the coders on this" you have to say "we want someone who's actually good at coding on this" and suddenly everyone is effectively stat ranked by everyone in the studio on every task that needs doing. Sounds exhausting. There's no reason you can't have many of the benefits of a flat hierarchy such as people wearing multiple hats when they already have their one primary hat nobody can take off them. Very interesting video.
Holy cow! I would have loved this. And equal pay for everyone including the owners -- INCLUDING THE BACKEND! Fuck. I'd have quit a job and moved for such a thing. Because I'm an obsessed, work-a-holic weirdo who loves to wear many hats and is very entrepeneurial. I used to get in trouble at large corporations a lot for going outside my lane and often doing others jobs better than them before I got tired of getting yelled at. Most corps want easily replacible cogs. Anyway, this experimen would have been fun.
Hi, Tim! Great videos, so cool you've decided to create a channel and share the experiences!
You've probably been asked this a million times, but as a long-time fan of Troika and Arcanum, I have to ask:
is there ANY chance of Arcanum 2, at some point?
It is probably doable if you hire for a flat hierarchy, rather than imposing one on people not used to it.