Forgot your password?
typodupeerror
Businesses Hardware Games

Former Valve Hardware Designer Recounts Management Difficulties 224

Posted by Unknown Lamer
from the lord-of-the-flies dept.
DavidGilbert99 writes "Jeri Ellsworth has opened up about her time at games developer Valve and has hit out strongly at the so-called flatpack management structure. She says that despite Valve's claims of a democratic structure, there is a layer of powerful management in place and when she was fired she felt like she had been stabbed in the back. 'If I sound bitter, it's because I am. I am really, really bitter. They promised me the world and then stabbed me in the back.'" Develop Online has a good transcript. In the end, Gabe Newell at least let her team keep the rights to their augmented reality hardware. She also notes that she still loves Valve, but the management and bonus structure resulted in communication breakdowns at Valve's size. It does seem that a flat structure can work: Andy Wingo has been weblogging about working at Igalia and seems pretty positive about the experience.
This discussion has been archived. No new comments can be posted.

Former Valve Hardware Designer Recounts Management Difficulties

Comments Filter:
  • The standard text is The Tyranny Of Structurelessness [jofreeman.com] by Jo Freeman.

    tl;dr: if a visible hierarchy isn't allowed, an invisible one will form and bite you in the ass.

    • by Arrepiadd (688829) on Tuesday July 09, 2013 @06:04AM (#44223463)

      I haven't had the time to read the text you post, I'll try to do it later on tonight. So, this may be a bit off, I'm posting this based on your tl;dr.

      W. L. Gore and Associates [wikipedia.org] (the company responsible for Gore-Tex) can be used as a counter-example to what you/Jo say. There are no bosses (everyone is an associate) and people work in small teams. No one bites others in the ass. And the company, while not the biggest in the world or whatever, works fine and people in it seem to be happy.

      One key element seems to be the size of each of its campuses. They limit them to 150 people. More than that and what you mention starts happening. A de facto hierarchy arises and bickering ensues. But below these numbers (and this seems to be corroborated by other sources) people work as in a small community/village and peer pressure keeps everyone working nicely. Above 150 people clustering of people occurs and, while peer pressure still occurs within these groups, the problems still occur in between groups.

      So, perhaps flat structures do happen, but only in small groups because "friends" take care of their friends, but employees don't necessarily take care of other employees (especially when the employee he's supposed to take care of is his nasty boss).

      • by nine-times (778537) <nine.times@gmail.com> on Tuesday July 09, 2013 @09:54AM (#44224805) Homepage

        One key element seems to be the size of each of its campuses. They limit them to 150 people. More than that and what you mention starts happening. A de facto hierarchy arises and bickering ensues. But below these numbers (and this seems to be corroborated by other sources) people work as in a small community/village and peer pressure keeps everyone working nicely.

        I'm not so sure that's an accurate description of what happens. Under almost all situations, people will develop a de facto hierarchy. It may be fairly fluid and casual, but even among a group of 5 friends, there will usually be some kind of pecking order. Under 150 people, there's the possibility of that de facto hierarchy being managed well without formalized structures. Over 150 people, the social bonds become thin, and factions will form and compete with each other for power.

        I would tend to agree with the analysis, "if a visible hierarchy isn't allowed, an invisible one will form." It's not certain whether it will bite you in the ass. One of the obvious ways it's likely to bite *someone* in the ass is if that person believes that they're setting themselves up to be "in charge" of this non-hierarchical structure, but that person doesn't have the social power and charisma to maintain their position by informal means. That is to say, if you start a company and create a flat structure without formally putting yourself in the position of being "in charge", then you'd better be popular. Otherwise, someone else might decide to take the reigns, and they might get more support in the ensuing power struggle.

      • Limiting the size of the company is certainly going to help, but I see another problem with the way Valve was setup that might not impact W.L.Gore & Associates. According to the article Valve gives out bonuses to specific teams that work on hot projects. That sort of thing is bound to lead to infighting since everyone wants to be on the hot project and participate in the money rain when the project ships.
    • by FriendlyLurker (50431) on Tuesday July 09, 2013 @06:15AM (#44223495)

      if a visible hierarchy isn't allowed, an invisible one will form and bite you in the ass.

      ...and it will form around the worst, most manipulative personality types... which also happen to be the worst leaders.

      • by ATMAvatar (648864) on Tuesday July 09, 2013 @08:48AM (#44224215) Journal
        That doesn't sound significantly different than traditional hierarchies. Just look at most politicians and CEOs.
      • by kcitren (72383)

        ...and it will form around the worst, most manipulative personality types... which also happen to be the worst leaders.

        Actually, that's what makes good leaders. Good leaders are able to get a group to do what they want.

        • by gl4ss (559668)

          ...and it will form around the worst, most manipulative personality types... which also happen to be the worst leaders.

          Actually, that's what makes good leaders. Good leaders are able to get a group to do what they want.

          that doesn't define a good leader. that just says that he's good at leading other people to do what he wants, not that he is a good leader for the group.

    • by readingaccount (2909349) on Tuesday July 09, 2013 @06:17AM (#44223499)

      by Jo Freeman

      Thought that was worth highlighting.

    • by Trepidity (597)

      I agree structurelessness is problematic, but there are structures that work which are less hierarchical than traditional boss-and-subordinates tree-styled management structures. A common feature of Scandinavian workplaces, for example, is a set of committees with precisely specified areas of competence. It is relatively non-hierarchical but very structured and transparent: rather than informal cliques taking on different roles, formal committees with procedures take them on. Overall it works pretty well.

    • by JDG1980 (2438906) on Tuesday July 09, 2013 @09:06AM (#44224359)

      tl;dr: if a visible hierarchy isn't allowed, an invisible one will form and bite you in the ass.

      Wikipedia is a perfect example of this. Officially, there isn't supposed to be any hierarchy of editors. Administrators are supposed to be "janitors", just doing non-controversial maintenance work, and aren't supposed to have more rights than regular editors on articles. In practice, of course, it doesn't work that way, and there is a very clear hierarchy which usually remains unspoken. What you can get away with on Wikipedia depends a *lot* on whether you're an administrator, how long you've been on the site, whether you are an old friend of Jimbo's, and whether you kiss the right butt on IRC.

  • Sadly (Score:2, Insightful)

    Everytime i read "Valve" my thoughts pavlovianly go to HL3. Still not a single word about it?
    • by geirlk (171706)

      Doesn't matter.

      But where's L4D3?

      • If it's any like L4D2, they can keep it. Seriously, a horror shooter was replaced by yet another splatter shooter. What are they supposed for L4D3? Why not make it a rail shooter where all you do is keep that left mouse button pressed and rack up a few thousand kills?

    • Re:Sadly (Score:5, Informative)

      by jones_supa (887896) on Tuesday July 09, 2013 @06:27AM (#44223545)

      Everytime i read "Valve" my thoughts pavlovianly go to HL3. Still not a single word about it?

      Lately someone got snooping into Valve's Jira [valvetime.net] and some conclusions made were that HL3 was either inactive or in developmental infancy. L4D3 was advancing nicely and the Source 2 engine had huge development resources behind it.

      But who knows.

      • I wouldn't hold my breath on HL3 being in Jira meaning anything. There was virtually nothing in the HL3 category, and not even a mailing list for it, where as even the L4D3 category had that much.

        • Yeah, that was actually my thought too. It's possible that the HL3 stuff is just a dummy stub without any real plans to make development around it. If those screenshots are real, I guess we can forget getting a new Half-Life at least in the immediate years.
      • by CODiNE (27417)

        I can imagine a lack of desire to do HL3 right now. Look at how many zombie games, shows and movies are out these days. By the time you finish such a large project whatever you've done will already seem dated. If I had a large zombie project on hold I'd rather wait it out a few years til it's a bit fresh again. I think people are getting a bit bored with the vampire, werewolf and zombie stuff.

    • by mmcxii (1707574)
      Wasn't there suppose to be a third installment to HL2 first? It's been a while since I've looked into any of this but that would be a hot item for me to play. I'm pretty much out of the gaming scene but I would like to finish the HL series and would play Thief again if T4 ever came out.
      • They pretty much admitted that the episodic model was a mistake and that they'll go for HL3 directly. IF EVER. BOB SAGET!
      • by Maritz (1829006)

        Thief 4 will come out. There are some gameplay videos from E3. I'm more than a little scared that they're going to utterly fuck it up to be honest. And Garrett isn't voiced by Stephen Russell any more.

        People (Eidos in this instance) always come along and think 'yeah we can tweak the formula with this old franchise, it'll be better, more modern etc' and they're always, always fucking wrong.

        • by mmcxii (1707574)
          I thought the same thing when T3 came out but in the end it was a pretty enjoyable game. Hopefully this will be of at least that quality. I'll have to check out the videos when I get a minute. Thanks for the heads up.
  • by Anonymous Coward on Tuesday July 09, 2013 @05:08AM (#44223261)

    When you hear breathless talk about new paradigms in management social structure it's always people grasping at straws attempting to pin the tail on the contributory factors to their synergy. Good shit comes from selfless people, and selfless people attract parasites and tempt honest people in to taking advantage of the situation when their feelings get hurt.

    Frosty Piss for everyone.

    • When you hear breathless talk about new paradigms in management social structure it's always people grasping at straws attempting to pin the tail on the contributory factors to their synergy. Good shit comes from selfless people...

      I kind of agree. I think part of the problem is that people are searching for a magical formula. Bad managers like to think in terms of, "If I just do [x], then every one will work hard, there will be no conflicts, and I will get rich." They just want to know what "x" is. The problem is, "x" actually includes all of the following (plus more):

      • make good decisions
      • adapt to the situation as it changes
      • surround yourself with good people
      • be good at reading people and motivating people
      • do a good job
      • be lucky

      Sorry

      • Also, in line with the "Peter Principle," when some people are "promoted" into management they think that their job is to bark orders or to make proclamations to be followed without questions and that their title of manager makes all of their decisions right. These people enjoy the arbitrary power they think they have. I hold that a leader (manager) is a servant who's job it is to keep things out of the way of their team. It is also a leader's job to identify what motivates each team member and use those

  • by crioca (1394491) on Tuesday July 09, 2013 @05:11AM (#44223273)
    If all it takes is for one laid-off ex employee criticizing the management structure for it to be deemed not to have worked, then there's no such thing as a workable management structure.
    • by Anonymous Coward on Tuesday July 09, 2013 @05:32AM (#44223361)
      She got paid, got to do what she wanted, didn't get enough resources because the rest didn't believe in it and she couldn't convince them, then she and team got sacked but got to keep the stuff and continue with it.

      Doesn't sound that terrible to me. What other company would have paid her and let her do that?

      Maybe the sacking bit and run-up to it was done badly. But in most other companies you wouldn't even be able to do that project in the first place, much less keep the rights after you got fired.
      • by gl4ss (559668)

        well the run up happens in secret in the shadow good buddy organization.
        that's what happens in most hierarchical(on paper) systems as well. it's easier for the buddies because in theory you can't bitch about it through proper channels because there are none!

        but heck, if valve is developing games and hardware with no structure on paper at all.. they're not developing anything and just doing reactive fixing of the steam platform and buying random games to their stable. and well.. fuck, valve as a developer. e

      • by Rockoon (1252108)

        She got paid, got to do what she wanted, didn't get enough resources because the rest didn't believe in it and she couldn't convince them

        My observation is that they believed in it enough to hire Michael Abrash, who is working on the very thing that she was working on.

        Looks to me like a possible "old boys club" mentality is going on there, with Abrash being accepted into it because Newell had worked with him before (at Microsoft) and had been trying to get him to join Valve since forever. This isnt to knock Abrash because that man knows his shit, but maybe Ellsworth was considered competition to what Abrash was doing and as such "had to go

        • Abrash is critical of AR for technical reasons he has shared on his blog. She most likely was fired because they didn't believe in AR as it has significant problems to solve over and above the problems in the way of implementing good VR.

          Further VR follows a fairly obvious path from current games whereas AR requires innovation in game design and input to realize its potential.

      • by epiphani (254981) <epiphaniNO@SPAMdal.net> on Tuesday July 09, 2013 @07:50AM (#44223917)

        I actually thought this through... my first reaction was wow, that sucks. Maybe Valve isn't the utopia that people think it is.

        Then I stepped back and remembered what I've heard about Valve. You make your own decisions - and you're accountable for them. They said they had a million dollar lab, but couldn't hire anyone to do the machining. But who decided to build that lab? Did they spend a million dollars on equipment then not use it?

        A flat organizational structure doesn't mean there's no politics. It means politics are MORE important - it's harder for some team to simply burn cash, because everyone's eyes are on you. It's hugely increased freedom - but all of the responsibility that comes with it. Assuming that anyone in Valve could decide to go build a million dollar lab, what do you think would happen if it failed to get utilized?

        This is one side of the story from one person. I'm sure there's more to it than the lab, but the lab example shows a basic misunderstanding of the personal responsibility one has in a flat org structure.

      • by Darinbob (1142669)

        The difference is the perception being promoted by Valve that it does things differently. It has an employee guide that says how things always work, and this firing process didn't follow that guide. Maybe there's a bit too much naivete in thinking that this guide was law and could never be broken. Valve is using this new paradigm of management to recruit people. But the paradigm doesn't work and isn't being applied.

    • by WillKemp (1338605)

      [......] then there's no such thing as a workable management structure.

      They're certainly very few and far between. There's probably about as much chance of finding one as there is of finding an alien civilization.

      • There are workable management structures, taken at instances in time. I don't think there are "workable management philosophies" or an algorithm for creating a management structure that works in on all projects and all conditions with all personalities.

        Everyone tries to copy a formula, but it's not a formula.

    • by urbanriot (924981)
      Yep... I read this Slashdot entry and wondered, "why is the typical behaviour of a fired employee newsworthy?" Even if you had a large enough sample set, my care meter would probably stall slightly above zeor.
  • a software company. Was always going to be difficult. Seems to me they should perhaps have split into a new division - not kept it under the same roof/structure

    • by aitikin (909209)

      a software company. Was always going to be difficult.

      Works for Apple...

      • by AdmV0rl0n (98366)

        That the Apple that compartmentalised the hardware under Jobs - so members of the team did not know what they were actually building, for whom, and why?
        That Apple?

  • C64 DTV designer (Score:5, Informative)

    by Neo-Rio-101 (700494) on Tuesday July 09, 2013 @05:43AM (#44223409)

    I assume this is the same Jeri Ellisworth that designed the Commodore 64 Direct to TV unit?

    • Re:C64 DTV designer (Score:5, Interesting)

      by hamster_nz (656572) on Tuesday July 09, 2013 @06:00AM (#44223443)

      I assume this is the same Jeri Ellisworth that designed the Commodore 64 Direct to TV unit?

      Yes, uber-hacker-maker. Has a collection of self-restored electron microscopes.

      Much smarter & more creative than your average person.

    • Re: (Score:2, Interesting)

      by Anonymous Coward

      No, it's the same Jeri that made a half-working prototype typical of a college project, completed and manufactured by a German dude, who passed things back to her as salesperson. She's spent the next 9 years selling herself as more capable than she is, then whining when people get fed up with her.

      Sophie Wilson she ain't.

  • Andy Wingo has been weblogging about working at Igalia ......

    Weblogging??? Did i fall into a timewarp and end up back in the 1990s or something?

    • by kermidge (2221646)

      Say what? What's wrong with a good descriptive noun, the original in fact.

      Blog? Sounds like something you wipe yourself after.

  • It makes sense that a hierarchy (hidden or otherwise formed). It is either human nature or so ingrained in our culture that there will always be those that step up and those that submit.

    Regardless, a failure of one within such a proclaimed 'structureless' system is not necessarily a failure of the system, at that or any size.

    The yardstick to measure success or failure by is whether "HL2E3" or "HL3" or "Half Life:Eternal Wait" or whatever it is titled now ever is released.

  • I remember an Adam Curtis documentary that basically described those old 60s communes the same way. Communes were set up as completely power free institutions, places were no one would have power over anyone else and all important decisions could be made communally.

    But of course power did exist, it was just being hidden. Someone owned the land, someone had some important income maybe someone was just too damn charismatic. And so because the power was hidden, it was never confronted or addressed. There were

  • by Anonymous Coward on Tuesday July 09, 2013 @07:11AM (#44223729)

    If you look at "income", Valve is successful. Very, indeed. But money is not the only metric. It all tells us Valve got lucky developing two games aaages ago and the being the first to set up a working Digital Distribution System. That they combined it with their second (and last!) successful game was a masterstroke, but only pure luck.
    Had Half-Life 2 not been such a success (say a title in the 80s/100), Steam would not have been taken off like it did.

    If you use the metric "releases successful products" for success, Valve is working mediocre at best.

    They shovelled in a lot of cash with Half-Life and Half-Life 2 until Steam was running with full steam ahead... and that digital distribution platform is carrying them since then. After the initial phase it was a self-sustaining thing that you just need to maintain without screwing up too much. That is basically what Vale has been doing since Half-Life 2 and I ask you: What other successful projects do they have to show that we can use as proof for their successful system? You say "not much" and I agree.

    Valve seems to me very similar to 3DRealms. Both had a major success which gave them money and on that they kept running. Load words once in a while, punching their own chests how successful they are, both claim(ed) to offer a "free and creative" environment without "administrative overhead!!!1" - but both totally lack in coming up with more or better products than companies with "classical" structures. In fact, those classical structures are much more successful at chewing out successful and often high quality products.
    The difference is that Valve has Steam, a product that keeps generating revenue with Other People's Successful Games if you manage to maintain it (which is no problem with the money Valve has, it is not really requireing a lot of insight or creativity), so they can afford to be totally incompetent at creating own games (which they are).

    All Valve achieved lies in the past. And with "past" we need are quickly approaching "a decade and since then the existing stuff just has been maintained".

    They have as truly notable things
    Half Life (1998) + AddOns (1999, 2001)
    Half Life 2 (2004) + Nice AddOns that basically are TechDemos for the Engine
    Portal 2 (see below)

    That is it.
    Portal and Left4Dead they bought in (good call, but more a Publisher-Decision than actual a Develeopment-Success). Buying the right stuff requires money and one or three managers who make the right call, it's no sign your Development Hierarchy works.

    • by hibiki_r (649814)

      There's this game called Team Fortress 2, that has sold more than the Half Life series.
      Also, I'd not say that they bought portal and left 4 dead. They mostly bought the talent. Yes, that's how valve works: A whole lot of senior hires, very few entry level hires. Experienced hires tend to bring their game ideas with them. Just look at the hiring of IceFrog: Here, see a tech demo of your warcraft mod, made by our own developers. How about you make the game stand alone, and stop having to muck around with the

      • by drsquare (530038)

        Team Fortress is not a Valve creation, it's just something they bought out, like DOTA and Left 4 Dead.

    • by crashcy (2839507)
      I feel like you are measuring success by very strict terms of achieving what you want them to achieve. Yes, we'd all like to see HL3, but you're being rather flippant about the success of Portal 1 and 2, Left4Dead and L4D2, as well as Team Fortress, DOTA2, and Steam as more than just a digital distribution service. I've been on Steam since HL2, so I've watched it develop from something I hated and merely endured in order to play HL2 to something that I have on all the time. Aside from my games library, i
    • These days, Valve,s money comes from Steam. Their profits on that are stupid, like tens of millions per employee. Basically they just get to sit back and sell other people's stuff, and take a nice cut (30ish percent). As anyone who's ever had trouble with something will tell you, they have a minimal support staff, there's no phone number to call or anything, and responses take forever. Also when you really look at Steam it isn't that great. It isn't bad, but it is not some masterpiece of software engineerin

  • I don't know how many millions this project cost [theverge.com], but come on, we have rapid prototyping tools so it could've at least been made looking better than a high school science fair project.

    Add to that the fact that people refuse to wear their 3d tv glasses. Or glasses in general.

    Points to this being a smart executive decision to cut losses and move on.
    • by gl4ss (559668)

      where did you get that it was millions? at any rate, the frames for it are 3d printed(you can see from the surface of the plastic).

      practicality problems might hamper the project. maybe folks at valve weren't really into tabletop ar games. star wars chess gets old pretty quick.

      • by Rockoon (1252108)
        Abrash told them that AR wasnt ready for prime time, but that VR (always his favorite thing) was ready. He is probably right, however technology is advancing at an ever faster rate. Google seems to think AR is worth investing in, after all.
      • by csumpi (2258986)
        25 person team. $100k per person salary, just to keep it on the cheap side. Salary, per year is $2.5million, plus 25-40% benefits.

        Add to that material and equipment cost.

        Add to that rent, electricity, toilet paper.

        This project, per year, $5 million sinkhole, easily.
    • Personally I would be more inclined to use AR rather than VR glasses. VR glasses try to replace your entire field of vision and they often lag enough that it causes motion sickness. AR does not have these issues to nearly the same degree. Any prototype is going to look like that since many forms of miniaturization are lousy to use during the experimentation phase and some are indeed only doable on large scale manufacturing plants. One example is surface mounted technology.

      Yes any glasses are annoying to use

  • It does seem that a flat structure can work

    Uh, it can work at Valve. Because Valve has a LUDICROUS amount of cash. They're getting money for nothing because they have Steam. They managed to seduce the users with easy digital downloads and seduced the content owners with a promise of DRM. It's simply a better way of doing things. This is really bloody obvious, but getting that sweetspot of wooing both sides into letting you be their middleman was tight landing spot. Valve did it. And now they dominate digital distribution of gaming. Making Valve a ga

    • What you call "seduction," I call "good business." The market had a need. Valve met that need and was rewarded.

      • And if Steam would let me play my fucking games whenever I left the house or a storm took out our connection, I'd call it good too. As it is, the DRM is just a bit too much for me. Which is why I moved onto good old games [gog.com].

        • by tibman (623933)

          It has never stopped me from playing games? It just pops up a box and says that my saves won't go into the steamcloud. If the game had DRM on top of it then THAT would stop me.

  • Pretty good, if short read.

    Kinda explains why "Big Picture" is such an unholy mess. I tried using this as a replacement for my XBox. I really did.

    But every single game (including Valve one's) seemed to have a different mechanism for configuring the controller. Sometimes it was in game, sometimes it was by editting .properties files. Sometimes, I couldn't get the controller to work at all, despite steam indicating that the game had controller support.

    Even in the Big Picture interface, the menu's were

  • There are plenty of organizations with little heirarchy that work well... they're called co-ops.

    I have never heard of flat-pack, and the quick google I did found me (in addition to plenty of flat-pack kitchen stuff, and flat (apartment) management), was pure academic bs. Without some structure, trying to get anything done is like walking through mud, at best.

    From Ms. Ellsworth's post, it reminds me very much of what, in the early nineties, was called matrix management - which also sounded good, but meant, i

"Look! There! Evil!.. pure and simple, total evil from the Eighth Dimension!" -- Buckaroo Banzai

Working...