Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 22, 2014
arrowPress Releases
October 22, 2014
PR Newswire
View All





If you enjoy reading this site, you might also want to check out these UBM Tech sites:


Ex-Valve hardware expert shares uncommon look inside the company
July 8, 2013 | By Kris Ligman

July 8, 2013 | By Kris Ligman
Comments
    54 comments
More: Console/PC, Production, Business/Marketing, Video



Jeri Ellsworth is a woman with a considerable resume. An entrepreneur and self-taught computer chip designer, she's best known for developing the Commodore 30-in-1 emulator and for her technical articles for building DIY electronics. She was also among Valve's early 2012 hardware hires, but was fired in February of this year.

Now in a new video interview with Jenesee Grey of the Grey Area Podcast (above; beginning at 11:15), Ellsworth paints her former employer in a complicated light, casting doubt on the "flat management" structure Valve boasts, saying that the company is in actuality more hierarchical than it appears.

"I should frame all of this with I have a lot of friends at Valve and there are lots of great people there," Ellsworth says at the outset of the interview. "My view is not one hundred percent true for all the different groups in there."

"We've all seen the Valve handbook, which offers a very idealized view. A lot of that is true. It is a pseudo-flat structure, where in small groups at least... you are all peers and make decisions together."

"But the one thing I found out the hard way is that there is actually a hidden layer of powerful management structure in the company."

Ellsworth refers to this 'hidden' management with loaded terms, saying it "felt a lot like high school." "There are popular kids that have acquired power, then there's the troublemakers, and then everyone in between," she continues.

She also states that Valve's then-burgeoning hardware team was vastly undersupported. Because there isn't a lot of 'official' structure, Ellsworth says, employees gravitate toward the most "prestigious" projects, leaving teams like hers "starved for resources."

It was not for lack of money, says Ellsworth. "We had a machine shop with millions of dollars of equipment in it and couldn't hire a machinist for $40,000 a year to manufacture machine parts for it. Because they were worried that bringing in a machinist would hurt their precious culture."

"We were understaffed by about a factor of one hundred," she says. "We were having a difficult time recruiting folks -- because we would be interviewing a lot of talented folks but the old timers would reject them for not fitting into the culture."

Asked if these troubles signal that Valve is pulling out of hardware development, Ellsworth remarks simply, "There were five of us working on this project, and all of us were canned on the same day."

"If I sound bitter, it's because I am," Ellsworth acknowledges. "I am really, really bitter... I don't think [Valve's structure] works. Give people complete latitude with no checks and balances, it is human nature that they will minimize the work that they do and increase the control that they have."

Gamasutra has reached out to Valve for comment but has not yet received a response.


Related Jobs

Zynga
Zynga — Chicago, Illinois, United States
[10.21.14]

Senior Software Engineer (Front End)
Harmonix Music Systems
Harmonix Music Systems — Cambridge, Massachusetts, United States
[10.21.14]

Senior Product Manager
Harmonix Music Systems
Harmonix Music Systems — Cambridge, Massachusetts, United States
[10.21.14]

Web Developer
Cloud Imperium Games
Cloud Imperium Games — Santa Monica, California, United States
[10.21.14]

Marketing Director










Comments


Kujel s
profile image
I've never been a fan of valve, in fact I'm a critic and this new information only strengthens my opinion of the company.

Freek Hoekstra
profile image
that is fine, but one must accept that this is not the most unbiased of sources.
a disgruntled employee that was fired because the project she was working on was cancelled.
and that has a problem with the management structure (the culture) that has been working well for them for years. maybe they are over protective, that may be but I can see how going against this would be seen as being abrasive...

a company is not evil for firing people, if they are not adding value or are not needed anymore.
not saying what she says isn't the case, but if the project did indeed get canned, then it is only logical that the hardware people would no longer be required.
all I'm saying is it seems a little one sided from her end as well.

D Scheffler
profile image
Ok so a self-taught computer chip designer, and DIY electronics genius is complaining because she can't hire a machinist? With all the information, training manuals programs, and resources from machine tool companies online, in libraries, and at community colleges available she couldn't figure out how to machine her own part? Yep, now I know why she and her team where fired.

I'm no expert, but I can program a machine tool path and cut my own parts, or make the molds to make my own plastic parts. I figure by now valve has a 3D printer too, and you probably don't even need to go through all the machining stuff. But even if you did it is not much different than making a computer wireframe model. I mean if you can't figure out X, Y, Z, or polar coordinates are you really the sort of expert Valve is paying you to be.

I'm not saying Valve is without fault as they really should have tried to utilize her obvious talents in a more productive way, but I don't think she was entirely blameless when her project was not continued.

Bob Johnson
profile image
No one driving the bus at Valve. No wonder they do things at a snails pace. Their stranglehold on the pcgaming digital store market helped them work on whatever too without having to actually release anything and make money on it.

But what have they done the past 6 years? Portal 2 is all I can think of. And it was a small game not unlike the first.

Karl Schmidt
profile image
Constant updates to Team Fortress 2, L4D1, L4D2, CS:GO, Source Engine Mac and Linux support, Steam for Mac and Linux, DOTA2... http://en.wikipedia.org/wiki/List_of_games_developed_by_Valve

Lance Thornblad
profile image
Kinda silly - they're not a huge team. I wouldn't measure a company's success by the number of products they put out. Personally, I think the stuff they put out is first rate and that's more important.

Piotr Podgorski
profile image
Last 6 years? Here we go... Releases according to Wikipedia:

2007 -- Half Life 2: Episode Two, Portal, Team Fortress 2
2008 -- Left 4 Dead
2009 -- Left 4 Dead 2
2010 -- Alien Swarm
2011 -- Portal 2
2012 -- Counter Strike: GO
2013 -- Dota 2

A game per year. I don't know a lot of studios with this kind of track record. In fact, those who come close to it usually keep making the same game every year -- be it sports or generic shooters.

Not to mention what Karl Schmidt wrote.

I know it looks like Valve's been doing nothing because they haven't released Half Life 3 but... the world doesn't end there, you know?

Bob Johnson
profile image
OK i did forget all about L4D and its sequel. Not sure I would count those as 2 full games, but anyway that was the one I was missing.

Dota2 isn't officially out yet. Updating TF2 I guess is something. But if you first bought the game 6 years ago it isn't so much. :)

And I was talking 6 years as in the 6 years since Orange Box came out.

Portal 2 was decent but I mean come on basically Portal. It wasn't exactly a full game either.

Maybe it is just the way they do their work. They seem to release small bits and work on them. I have heard about Dota2 for forever. Might just be me, but I think I'm over it before it officially comes out. lol.

And maybe they are smaller than I think and most of their staff is working on the steam store. I guess that's why EA can come up with Origin and have it look better already.

Never cared for the black Steam UI and store layout.



Karl Schmidt
profile image
@Bob Johnson

Have you played TF2 recently? The frontend has been completely redone, there are lots of new maps and game modes, countless balance tweaks, a loot system with customization and trading, Mac and Linux ports, and this one 'little' thing: an entire system supporting an economy of users creating content and selling that content to other users. These are all significant updates.

Bob Johnson
profile image
@Karl

Yeah my son has played it recently. I'm pretty aware of what is going on there. Just that I got that game 6 years ago. :) I liked it. But I'm not making it a career or anything. So the move to F2P and various balance tweaks don't move me. And thus contribute to my view that Valve hasn't done much in 6 years.

Tyler Martin
profile image
@ Bob Johnson - No offense, but it feels like there's a bit of special pleading going on with you. Essentially moving the goal posts if you're unfamiliar with the term.

"Valve do things at a snails pace. What have they done in the last six years?"
"A whole bunch of stuff actually."
"Well sure, but most of them shouldn't count because reasons, and I can't make my point otherwise."

I mean, does a game need to be something other than multiplayer, or have a ten plus hour single player campaign to count? Is making a sequel to the original Portal that's twice as long, adds a myriad of new game mechanics, co-op mode, and ready support for easy creation and distribution of levels something that doesn't qualify Portal 2 as anything more than a small project like the original? What mystical criteria does a game need to meet before it's enough of a game for you to count it as such?

And even if we accept the suggestion that these are small titles and don't really count as full games, why is a small title every year less valuable than a large big budget title every 3-5 years? Should the amount of time players spend with it count for nothing?

And slight nitpick which is separate from my point above really, Dota2 isn't in beta anymore.

Samuel Green
profile image
Piotr's list is great but what is there to complain about really? Yes they take a while to develop games but pretty much never released a less than Triple A quality game experience (and I don't mean just AAA visuals). Worth it.

Bob Johnson
profile image
@Tyler

Hey I've softened my position as I said earlier. They have done some stuff. I did forget about L4D.

See before I forgot about L4D I was like what have they done that is new. Yeah ok worked on TF2 F2P and tweaked it and let the community make digital goods for it. Great if you still playing that game 6 years later and are into buying digital costumes and what not. But not so great if you aren't making it a career.

And then i had Portal 2 down which was alright. They couldn't expand enough on the first to merit a sequel.

And that's what I had down. And so I was like 6 years since Orange Box and that's it?

But now remembering L4D and the sequel softens my stance a bit. :) I still don't say those are 2 full $50 games though. Not sure those are even one big game. :) And I think it is legitimate to point that some of these games are basically studios they bought in the meantime which only leads back to what has Valve been doing since the OrangeBox?

So that's where I'm coming from. I guess I also thought VAlve was a bit bigger studio. But it seems most of their personnel work on steam and various research projects. I still hold the view that their store provides them with too much income so they can take their time working on projects and games. And not have to release anything. Sort of like how the cable company works. or how MS works in the OS/OFfice space.

I mean come on Dota 2? I'm already over it. I mean I played Dota eons ago. So long ago I forgot what game was modded to get us Dota. Wait WC3 right? :)

And yes ok Valve is saying Dota2 isn't in beta anymore but they also aren't ready for everyone to show up? Smells like beta to me still. lol. Unless we have a new state of software development. Gamma anyone?

edit: I am 2233 in line in the queue for Dota2.

marty howe
profile image
Hi Bob, maybe they're working on a game in secret. Half Life 2 took 5 years, and is one of the best action games ever created.

I'm assuming (and hoping) that they're making Half Life 3, inside that 6 years you're mentioning :)

Ramin Shokrizade
profile image
The video is long, and the discussion of Valve starts at 11:15.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

John Szczepaniak
profile image
I read the Valve handbook that circulated a while back (or at least excerpts). It sounded like the worst place in the world to work. I would run screaming from such a work environment. The book painted it like some kind of lawless chaotic void, where your pay would fluctuate depending on how much effort they thought you were putting into things, but there was no actual structure to work within. (I could be remembering this wrong, it was a while ago, but still... Bleurgh!)

Richard Eid
profile image
I guess you saw through it, too. I interpreted it as propaganda. The salary part you mentioned also had me wondering who would enjoy that.

TC Weidner
profile image
while I think a flat structure is great, and actually can work well in a small tight knit tech group/company, the problem lies in the fact that I dont think it scales well at all. That "management" structure simply will not work well after a certain amount of people, money,ego are introduced into the equation.

What works in a 24 person company likely is not going to work in a company of 240 etc

Lars Doucet
profile image
I'm pretty sure the problem with hierarchical organizations is that they *also* do not scale well.

Lance Thornblad
profile image
@ Lars - Then what does scale well?

Lars Doucet
profile image
Decentralized networks of individuals operating at human-scale enterprises.

http://en.wikipedia.org/wiki/Distributism

If Valve has gotten too big to operate effectively at its current scale, it should spin off new "cells" rather than trying to keep everything under one roof.

Matt Mihaly
profile image
And yet every single large organization in America is hierarchical from the Post Office to the US Military to Walmart.

Kujel s
profile image
@Lars Doucet thank you for that link on Distributism, I found it a most interesting read.

Jakub Majewski
profile image
Lance - the military, and especially the navy. Does that sound crazy? Well, if you think about it, the navy offers a great compromise between rigidity and independence. Naval commanders had total autonomy onboard their ships, and often weren't even able to communicate with their base until their mission was completed. It goes without saying, military structures are designed specifically to scale well - for example, if you look at the German navy of the 1930s, you'll notice that it was pretty tiny (due to treaty restrictions) but also designed to allow for rapid expansion as soon as the possibility presented itself.

Lars Doucet
profile image
You'll find a classic tenet of business theory (c.f. PeopleWare) is that organizations are more effective when you let the people at the bottom actually make some decisions and think for themselves. A light hierarchy which encourages "self-rule" among its lower divisions is similar in practice to a decentralized, human-scale network.

Highly vertical organizations that bureaucratize thinking and decision making have been shown to promote inequality, inflexibility, and ultimately, catastrophic sudden collapse (as recently witnessed in the financial crisis).

Lance Thornblad
profile image
Jakub - The military is about as hierarchical as it gets.

Lewis Pulsipher
profile image
In a serious war (such as WW II) that hierarchy in the military is much less noticeable. Peacetime military is the near-opposite of wartime military.

But a hierarchy is necessary, so that one person can order another to do something that's likely to get the second one killed, and he'll do it. Nothing like that in normal business.

In general, anytime an organization gets so big that not everyone knows everyone, and everyone knows the owner, you start to have cliques and politics that interfere with good business.

Kevin Albers
profile image
+1

Yes...that was a great read! It's interesting to see how the Distributism approach, promoted back in the late 1800's and based on even earlier ideas from the Middle Age, is quite applicable to modern society.

I'm happy with a computer, small music studio, software etc rather than '3 acres and a cow', but the idea is the same, and is very inline with the current Indie trend in game development.

Eben Sullivan
profile image
I apologize if this sounds a bit cynical, but anyone who read the handbook and believed it wholesale has likely never had a real job, or is viewing the industry through incredibly rose tinted glasses. That said, I think that some of the ideas contained within the handbook are very interesting (such as the "t-shaped" employee concept), but it is definitely a concentrated (and highly polished) effort to push the image that Valve wants the public to consume.

Lance Thornblad
profile image
Clearly that image is supposed to attract certain people to the company. I don't see anything wrong with that, but I can also see why people would call it propaganda.

I agree that it is silly to buy into it completely. As organizations grow, it gets more and more difficult to maintain a flat structure. I've always assumed that there is authority, when it is required.

I can also see people gravitating to the "prestigious" projects, as Ellsworth said. But she admits to being bitter - that certainly affects her perspective.

Sean Francis-Lyon
profile image
It has been a while since I read that handbook, but I don't recall anything in it that I should not believe. The central idea is no formalized power structure, which in my mind is the same as saying that there is an informal power structure.

Lance Thornblad
profile image
Listened to the entire interview. My impression is that the hardware program was simply never taken that seriously at Valve. It was always a relatively low priority - an experiment. It is not surprising that a former employee would be bitter about that.

Was her group really the hardest working people at Valve? Was she as "abrasive" as they said? Who knows, but how many company owners would let their former employees have the project they'd paid for? People get fired all the time with nothing to show for it.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Roberto Bruno
profile image
It's also entirely possible than what she came out with wasn't really that interesting and at some point they had to conclude that "this isn't leading anywhere"? But I noticed a lot of people are too busy throwing jabs at Valve to consider the idea.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Maria Jayne
profile image
I like Valve, I like them a lot as a customer, I'll say that now with over 100 games on Steam and nearly 9 years as an account holder.

It's entirely possible their structure isn't optimal or efficient, but when you ask people who have been fired what they think of the business practice at the company that fired them, I think very few of them would have positive words to say.

The mention of employees choosing not to work on the project and instead opting to work on alternative projects seems to be telling of the project progress/scope. When you allow people to choose what they want to do, you get the most honest assessment of what people think.

If anything, more should be made out of the fact Valve gave her the project she was working on, no doubt they had spent money on it, to just give it away says two important things. Firstly that Valve are confident it wasn't as good as they wanted and secondly, that they didn't take the more normal stance of shoving it in a basement somewhere and forgetting it exists.

The lady is now free to hire that Machinist, and her hundred or so staff and go and poke the Occulus Rift guys right in the eye...or not.

Bob Johnson
profile image
Yeah probably the reason for all of this wasn't a project they wanted to keep pursuing.

But probably frustrating to be talked into joining a company and then not be given what you perceived to be promised. Tons of support to realize your vision. And then you find out it isn't the utopia they promised either in terms of structure.

Casey Dockendorf
profile image
Jeri's experience working with Valve sounds very similar to the accounts of other people I have met who worked there. Some people thrive in an environment like that, others don't. I think it just depends on people's work styles and personalities.

All in all, I think her description of the Augmented Reality work she is talking about sounds incredible, and I hope this all comes to fruition. I personally feel like the bulk of people looking to invest in new virtual game experiences are more likely to flock to Augmented Reality than the Occulus rifts Virtual Reality space (I personally have no desire to play a game that completely blocks me out from the world so maybe I am a little biased).

I look forward to the Kickstarter campaign! ;)

Chris Dias
profile image
"human nature that they will minimize the work that they do and increase the control that they have."
You say that like it's a bad thing.

Sheldon Laframboise
profile image
Usually the most bitter people are the ones that get fired... I would take a lot of anything said with a grain of salt. Professionally she should just move on and look for other work with a bit of grace. Turning this into a press event is just going to look bad on her for any potential future employers.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Jed Hubic
profile image
How do you get a cliquey culture from their handbook? It describes an environment they want, you know what you're getting into. A flat hierarchy doesn't mean everyone is always equally important, it means everyone has the opportunity to pursue avenues they think bring value. It's up to the individual to make sure what they do falls in line with a company's vision, and if the company isn't on board you either work harder to make a case you're passionate about or you move on to the next project. Not saying this is best for everyone and maybe I've misjudged Valve's intention, but that's how I see it (probably an inbred view I'm sure).

Sheldon Laframboise
profile image
A little on the offensive there Matthew? What I'm saying is: She lost her job, all we hear is one side of the story. Her saying things like "We were actually the hardest working bunch in the company." (Something along those lines) reeks of a completely biased statement - there are also a few others.

Someone may have a different view of their work ethic, quality of work and even not really jiving with the company atmosphere can cause enough friction to lose a job. This is especially true in a flat hierarchy.

The fact is, she lost her job. It sucks but it does unfortunately happen. It has happened to me and many of my friends. Would she take her job back if it was offered to her? Maybe, but probably not. Doing something like this puts a stigma on on her head and the truth is, many future employers will go: "So if I hire this talented person, am not satisfied with her work or for some other reason need to let her go.. I can expect bad press? No thanks."

So no, I don't exactly have a bleeding heart for the situation. This was done in completely unprofessional form.

Jonathan Murphy
profile image
I completely disagree, Sheldon! How are we supposed to know a good work environment from a bad one?! DETAILS! I don't want people afraid to speak their minds. As long as it's factual we gain as an industry. A person should be measured by their track record, and not what they say in the moment. If we measured Cliff, Warren, and David Jaffe under those standards, they'd never make games.

Duncan McPherson
profile image
The funny thing about us is that we're primates. As primates, we're social creatures that will naturally develop hierarchies in those social structures. These hierarchies may assume one of several patterns, but they are hierarchies all the same. We are not so complicated or self-aware that we can successfully avoid the same hierarchical patterns.

You can see the same patterns emerge again and again among indie game dev teams, on web forums, in theatre troupes, in bands, and so on. The more wealth and influence a group begins to wield, the more important a dominant hierarchical paradigm becomes.

In my career, I've been a part of a few companies that have made similar claims. "We're not political," or, "we just get into groups and put our best efforts towards projects that matter to us." In my experience, those companies tend to be the most political and the most internally competitive. What you end up with is a bunch of fiefdoms, and -- generally speaking -- you compromise your output.

That said, Valve has had enough of the right successes to prosper. And, perhaps, their fiefdoms are willing -- maybe eager -- to ally from time to time in order to achieve greater collective goals. If that's so, that's admirable. That is truly uncommon. This interview doesn't give enough information one way or the other to confidently make that judgment.

Their handbook was a slick production, though. Because of that, it probably shouldn't be taken too seriously. Its "leak" seemed oddly intentional. It read like an ad. Maybe I've seen one too many ARGs, but it read like material that was trying to target an audience and to make a very specific type of impression among that audience. It didn't really read anything at all like an actual employee handbook... what with the lack of sections detailing things like military leave policy, maternity/paternity leave, harassment policies, travel policies, insurance policies, etc. Surely at least some of those things matter, even in a "flat" corporation!

Given that Valve is staffed by people (more or less), what are the odds they would be able to keep to their own lofty ideals? Are they, to refer to Hempel's Paradox, a white raven among game developers? They'd certainly like us to believe so, and that itself is worth questioning.

Llaura Mcgee
profile image
I think you're exactly right about social structures Duncan. I recommend people check out, the admittedly a bit sensationalist but very interesting, All Watched Over By Machines of Loving Grace by Adam Curtis which looks at techno-utopianism. Specifically related to this story it looks at flat hierarchies, such as, in the 2nd episode, the hippie communes of the late 60's, early 70's which collapsed at the hands of very similar problems.

Although I do think this whole thing is a bit of a non-story and rather unfair to both Valve (this structure works for them, and this hardware team was a case of a group not playing along) and Keri, who clearly wasn't seeking such press (only a portion of a really interesting interview). I do think firing without a heads up, contradicting the manual, was very unfortunate though.

I wonder where Abrash's focus on VR fits into all this as well...

Kain Shin
profile image
Thank you for this, Jeri. This is extremely valuable insight

Jed Hubic
profile image
Semi interesting podcast for sure, but working in a company with a flat hierarchy myself, I love it but it's not for everyone. Also I'm not trying to make assumptions but anyone we ever let go in that environment was usually never surprising. Valve obviously found success in that model but the caveat which was sort of addressed about people working on visible projects, is actually a big positive in ways as it forces people to constantly think about the value they bring to the company as opposed to getting caught up in your own world and losing sight of any sort of vision (that fact can definitely blind side certain people).

Vijay Srinivasan
profile image
I really agree to the point of 'hidden powerful management structure' I am also a firm believer of the saying 'Too many cook with spoil the broth'. When a company like Valve is pioneering something like for instance the digital game service or for that matter any of their games, there has to be a group, in this case the leadership team making strategies and decisions. Not everything can be handed over to the groups - that will only create more chaos.The concept of independent teams within a org is practiced by quite a few studios. Look at the guys at Supercell- their CEO admitted that they have this concept of independent cells with complete autonomy and he is the least powerful CEO.What ever said, end of the day you are looking at a distinctive studio always in forefront of innovation in what ever they produce. And the ultimate question is that 'are the employees at Valve happy ?' if yes,none of these are going to matter. I am sure a lot of us are hoping that Half life 3 will surface one fine day- I am looking forward to that announcement from Gabe at least !

WILLIAM TAYLOR
profile image
So basically Valve is like everything in life. Some people are outgoing and become popular and gain power. People become friends, a group of friends become a clique, and those cliques eventually obtain power through their size and influence. And it's not total anarchy with no management whatsoever.

I wonder what percentage of people working in the games industry never had a job prior to it or experience working in a group (like not even middle school project) or organization? A lot of things they seem surprised or shocked about about come off as, "umm, welcome to the real world."

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Matt Cratty
profile image
As with anything, the truth is almost certainly somewhere in the middle.

D Scheffler
profile image
I just had another thought about the machinist issue. How about she contact a local school with a machinist program and see if they would like to do a student project to make her parts for her. Students with the winning parts or designs would get credit on steam. Now that is the kind of creative leadership to get things done valve was looking for.


none
 
Comment: