A developer's thoughts on crunch - In games & beyond

Hey duders.

Cyberpunk 2077's recent delay has caused quite a stir on social media regarding crunch & the human cost of game development. As a professional software dev, I decided to write an article detailing my thoughts on the subject from that perspective.

I value the thoughts & opinions of the GB community a ton, and wanted to share the article here to hopefully get both some feedback and discussion on the go.

Here is a link to the article.

Thanks, and looking forward to hearing your thoughts =]

Excerpt(para. 51 & 52):

"The games industry chews these people up and spits them out. These workers set their intense pace, and the moment they falter from it, they will trip and they will fall. This leads to burnout. This leads to exhaustion. This leads to reconsidering their goals. This leads to them realizing their wage doesn’t reflect their work. This leads to looking at what it would be like to be a developer in a less volatile industry. And so after a year or two, they leave the game industry altogether with an acrid taste lingering in their mouth, and a few more bags under their eyes.

But you know what? There’s always another passionate kid out of college who is eager to gobble up the position they left behind."

Either from the angle that this is an attempt to expedite churn of workers to get to a point where new workers constantly have diminished expectations of QOL allowing the company can do whatever, or from the point of view that this wreckless abuse of humans is just greed to make the numbers go higher, this sucks. This is by no means new to me. It still, sucks. I was riffing on alternate reality-based humour the other night(while ignoring most of the OP's intent), and one I didn't write was one about Cyberponk(Alternate Realities are similiar but different), and how Cyberponk emerged to an astonished world who had no idea the game was being made. Hype raises funds by increasing awareness and anticipation. I'm gonna go wander and think a little on what that reality could have looked like, where the game is ready when its ready and I don't have to think about employees going through "crunch" to meet arbitrary timescales.

Yep. It shouldn't be necessary but you're also dealing in an industry where each project is uncharted territory. I think that's why planning always goes sideways. You either need phenomenal planning or a publisher willing to wait til a product is ready to push out.

@shindig: I've heard several anecdotal stories that a big problem with modern day planning is that they spend 2 years in pre-production coming up with ideas, toying around with different concepts and then are suddenly tasked with making the remaining 80% of the game in a year which is typically an impossible timeframe even with massive crunch.

This is just an uneducated opinion but games have always been insane technological feats, "it's amazing that any game gets made at all" as we like to say, many early games were developed by one or a handful of people purely out of passion, oblivious to the potential of it being a career, now we've normalized it.
Now you can have a career making games and in some ways it's easier than it was in the past, in some ways it's harder, but overall making a game is still an insane undertaking, the projects are bigger, budgets bigger and scarier, and the cost of failure has never been higher.
Seeing an artistic vision of this scale through to the end is nuts, it's like building a pyramid, i can't imagine working in this industry it sounds like a nightmare not just because of "how devs are being treated", also because of the task itself.

@cikame: @humanity: @shindig: Reading these put Halo into my mind. Different situations to be sure(Tied to hardware, size of moneypile behind publisher, etc.) and this is all outright speculation on my part, but this feels like describing different parts of the same elephant.

---

Honestly, much respect to people working under the types of conditions raised in the linked article and highlighted by the ongoing/intermittent media coverage of people retiring from working in games over said conditions. If you are one of the people who I just described, know that I wish there was more that people at this (consumer) end of things could do to not enable things like this.

Because I don't need to tell you how many pieces of clothing made in places are sold every day for you to know that trying to tell people how it's made hasn't solved the problem.

Great read! I'm another fellow software dev here about 8 yrs, but never worked in the games industry. Here are my thoughts on how to "solve" crunch.

1. Games as a service, Micro transactions or a Subscription Model (ex. Riot Games)

Before you roll your eyes and groan, this approach is common outside of the games industry. Almost ALL software coded these days are developed as subscription models: Office, Photoshop, AWS, etc. Yes it sucks to pay for a game on a regular basis but this helps keep a stream of consistent revenue to keep shareholders happy. The problem why a lot of games have tight deadlines is because game companies must prove to their shareholders that they are making money. So the CEOs are pressured to release before the end of the calendar year. And if you think fuck the shareholders, why don't all game devs just be independent, where do you think the money to pay everyone to develop AAA content will come from? Outside a miracle Kickstarter campaign like Star Citizen, most AAA devs need investment.

2. Don't advertise the game until it has gone gold (ex. Nintendo)

One of the main problems I have with CDPR is how they have frequently teased Cyberpunk 2077. There are trailers and multiple gameplay walkthroughs with commentary out there. Their game has been at multiple E3/tradeshow events. They boldly proclaimed that they would not extend the deadline again until they just did yesterday. CDPR set themselves up to an impossible standard and are now paying the price with angry and disappointed fans. The recent news of death threats on the CDPR devs is horrible to hear. A possible solution to crunch is to avoid advertising the game entirely. They could do what Nintendo does and release the game without any hype. This way CDPR has no expectations to set, no one to disappoint and they can develop the game for as long as they want.

3. Release the game on certain platforms first, or have Early Access done right (ex. Supergiant Games)

What I love about Supergiant is how smart they are with their releases. When they released Bastion, they didn't just decide to release on all platforms at the same time. They did so in segments, first they did XBox, then PC, then Mac, then PS4 and Vita. Even though Supergiant is nowhere in the same league as CDPR, they understood that releasing simultaneously on multiple platforms is an exponential QA task. Each additional platform is like a snowball effect and introduces more unforeseen variables. CDPR could have released XBox One as a console exclusive and PC for the initial launch and then slowly release on other consoles. As of this writing CDPR plans to simultaneously launch Cyperpunk 2077 on nine platforms!

In addition, Supergiant's most recent game: Hades, is the perfect example of early access done right. They started with a very polished core game and then layered new features step by step until their launch. I doubt CDPR would take this approach with Cyperpunk because it is a heavy narrative game but it is a possible solution to avoid crunch in the future.

I liked the article, and it definitely gave me some added perspective to the issue.

The only thing I can add to this discussion is that crunch has nothing to do with the quality or complexity of a game. I observe a lot of people assuming that crunch is because modern games are either more complex than games from previous generations, or that there is some necessary relationship between crunch and the quality of a game. This is totally false. The games you played on previous generations, as simple as some of them may have been, were most likely the result of crunch. This fact is doubly true in the case of licensed games. Moreover, many developers were forced to crunch on games that ended up being awful. For example, developers had to crunch for the notorious Blues Brothers 2000 for N64. I think the worst offender is Tattoo Assassins, which was also a result of crunch. Imagine having to crunch for a game that never came out, so you couldn't even list it on your resume. Here's an account from a Data East dev of the development of Tattoo Assassins from Dan Amrich's blog:

During the whole project, we were required to work 12 hours a day, 7 days a week. Our lunch and dinner were take out - paid for by the company, which was great, but the purpose was to keep us from going out and having an hour long lunch or dinner break away from our desks. Every meal was a "working" meal, meaning we would have meetings in the conference room as we ate to keep from interfering with the "real" work. If we were late to work, or left early, we would be threatened with having some of our $25,000 bonus docked. The "carrot" had become a big baseball bat we were beaten with almost daily. One guy walked out after having his bonus docked for coming in at noon, and he quit on the spot. The next day he was placated and he came back. I remember this moment vividly--working away at 3:30 a.m. one morning. I was so physically exhausted that I casually leaned over to the garbage can and vomited, after which I went right back to tapping away. My fingers never left the keyboard.

I've been working in feature animation, games, VFX and commercials for almost ten years now. I've crunched on almost every project with only a few exceptions. AMA

@martyns: Hi, tyty for volunteering like this. A question with three parts; Approximately how many hours a week would you say the job took up at the time you crunched the most/worked the most, same for time least worked/crunched and finally were they the same gig?

(Feel free to leave locations and names out, I'm just looking to talk about the practise)

@petesix0: I've worked a project that has an hour commute each way and the crunch was so shitty I got an Airbnb for a week so I could get more sleep. That was about 100hr weeks. Thankfully that was in commercials where the crunch trends to be weeks instead of months.

For the longer instances in vfx projects crunch usually involves working late (9-8 instead of 9-6) and when the push really comes a weekend day.

I've had plenty of projects that didn't crunch, from my experience it's generally something nobody wants and either happens because the production management dropped the ball, the clients throw a last minute grenade or it's a shitty company that builds it into their schedule.

I've had a pretty good time in feature animation and games though. I will do everything in my power to not work a position like the commercial gig I mentioned above. Especially as we weren't paid for the overtime.

From my perspective on the outside looking in there seems to be only one option: unions.

Crunch has been happening for years and years. Publishers and development studio management are not going to suddenly change on their own out of some newfound sense of humanitarianism. If the folks who are being crunched want change, they're going to have to do it themselves.

Hopefully, their new labor union charter includes one of my favorite quotes:

Poor planning on your part does not necessitate an emergency on mine.

- Bob Carter

@martyns: The 100 hour stuff sounds really oppressive to me and the weekends(huge sigh). I remember visiting a 2D hand-drawn animation studio that was doing commercial work and the memories of the air of resignation and commitment came to me when I read that(Not saying that's your perception of your time doing that). To be fair I think they'd been the studio that had always done the project and it was being taken and given to a cheaper studio overseas so maybe that was somewhat responsible.

---

This though, feels an indication of "Who cares, they get results" management choices made by people immunized from costs by heirarchy. Man alive, I hope this stuff somehow escapes the "We Got A Badge To Put On The Innocent" or at least moves quicker past it than things like food. Badges are just things that get paid for and are easily covered up with a sticker.

(On a personal note tyty for everything you wrote, I'd have come back to thread sooner but I-am-ha-ving-a-time.)

@bane: So many people I'm seeing are talking about unions, everywhere I'm going people are talking about unions, maybe people should get unions they seem really popular, "Unions". People should talk more about unions.

I would say the problem is less with crunch and more about the fact that they don't pay for overtime, maybe small bonuses here and there but not anything close to what they would deserve for the extra hours they put in. A studio like CDPR or Rockstar that puts out one game every 7 years, in other words you have to crunch a few week, maybe 1-2 months every 7 years, doesn't seem like an impossibility - it's maybe different for annual franchises, but those get to recycle more assets. When it comes to deadlines, my view is that they are simply a necessity in any big project, most people need to have that final concentrated push to get it done, if they keep moving the release date then that final push never happens. So it all comes down to the money - if overtime is required to make this commercial product viable, fine, just make sure your people are getting paid.

I've got 8 years of experience as a developer but I work in the financial industry. From my point of view I'd say the average dev here works maybe 40-50 hrs/week which is reasonable.

But the thing that I never understand when I hear 80+hr work weeks is how utterly unproductive you become at a certain point. It even becomes detrimental to keep working vs step away from your work and getting distance from it. Highly cognitive tasks like programming simply have a huge drop-off in quality and productivity to the point where you start producing such low quality output that you are counterproductive and make more hours for yourself where you need to fix the mess you created. And any strength of willpower will not stop that from happening - it's a biological limitation. I've witnessed this myself during weeks when I'm constantly logging into work after my kids go to bed. Your ability to creatively problem solve is severely reduced, your motivation is shot, your clarity of thinking is terrible AND this is compounded by your increasingly lack of self-awareness to recognize that....its the frog that slowly boils in the water without noticing situation.

In my experience it all starts from the top. When management goes home at 5pm everybody else should then feel like they can safely do that too. Any messaging from management that crunch is not 'required' but management is staying late everyday is horseshit. Their actions are telling their employees the real message.

I will also say I totally agree with the sentiment that the games industry perpetuates this because it has an endless supply of enthusiastic people that want to work in games. Working on a video game sounds a lot more fun than the financial software I work on. But the cost (and the pay cut) is not even close to being worth it for me. It's a shame.

@billymaysrip:

Yep, I would expect a lot of shitty games to have involved crunch. I think research and anecdotal evidence points to crunch NOT being correlated with more productivity. People are completely delusional when they treat highly cognitive knowledge work as a linear function of more time = more output. It's an antiquated Taylorism management belief that comes from when everyone worked in factories. If they do get more output that quality will be worse. Something has to give.

Sorry for the delay in responses, friends. Ironically enough, last week got super busy with work and I tend to try and avoid the computer over the weekend haha.

Thank you so much for all of the responses, feedback, and discourse - it's been great to read =] I wanted to take some time and reply to folks here! So here we go:

1000%. And, unfortunately, it is both brazen and totally "expected" these days =\ Like some folks I know say their employers treated it like a 'badge of honour' that they should wear, when really it's just kinda glorifying abusive workplace behaviours... It is exhausting, and sucks.

Totally on the bullseye with these, friends. I did bring up how estimation and stuff is a big part of the blame, but you're also right that it is super hard with an undefined scope. That's why, typically, we are able to assign things called "spikes" within a sprint. They are timeboxed and allow a developer/group of developers to spend that time purely doing research and then writing down their findings in an internal doc, generating new tasks, etc. The way the timeboxing works is, for example, let's say you assign 5 days to a spike. For an individual, that means they would have 5 days to do this research. Alternatively, you can put 5 people on it as a group effort, then that takes one "real world" day (5 people * 8 hour workday) as man-hours are a contributing factor. Either way, spikes are a big way to help alleviate the world of unknown scope. They're certainly not foolproof, and sometimes the outcome of a spike is "we need another spike on this thing we didn't know that we didn't know," but it can at least help you contextualize what needs to be done for an ambitious, new project.

Your uneducated opinion is very very very valid. Games are insane in a lot of ways, and I can't even fathom how some of the biggest AAA games still manage to come together. Typically you have people working on very disparate things (audio, visuals, combat, etc) which don't truly get "put together" until the end. And then all the bug fixing happens at the end, which is wild to me. In my current industry, we bugfix each others' stuff on the fly, because catching something early can make everything so much more efficient down the line. The fact that games have a quality pass at the very end where they try to catch & squish as many bugs as possible is some galaxy brain sometimes haha.

Also, very very yes on your second point. Making games now is such a HUGE effort, esp in the AAA space. Generating 4K assets, super high quality audio, 30+ hour storylines, etc means that not only is the expected scale bigger, the raw asset size is enormous. That means it generally takes more effort to make something which used to be simple. I mean look at CODBLOPSCW (god i love that acronym), it is like 250 GB at launch. Wtf. The amount & size of those assets people are working on are insane compared to the original BLOPS, which clocked in at 45ish GB 10 years ago.

Absolutely, I have lots of respect for those who stick it out. For a lot of them, they're driven by pure passion for the projects they're working on, and I can't help but give them a ton of kudos for that. You're right, we as consumers / audiences need to be more objective and empathetic for how the products we consume come to be. It can be a tough transition, but is ultimately way better to support companies (esp smol ones) who are compassionate to their labourers.

(reduced the text to save space in this reply)

I 100% agree with you. Especially, ESPECIALLY #3. Early Access is agile methodology put into practice tbh haha. Agile is all about the cycle of iteration, getting feedback, iteration, getting feedback, etc. By building that into the game, you're putting something out, getting feedback, and continually improving it. That way, when 1.0 eventually hits, the community has had a lot of input, and the game itself may have ideas the devs themselves may never have thought of on their own. It's win-win, allows for some income coming in during the time of development, and generally means the pace can be less frenetic. If I were making games these days, #3 would be my pick.

I do think strongly that numbers 1 and 2 are also great though. The first, yes. SaaS basically. It works, and it's proven to work. Again, iteration! Put out the MVP then improve it forever haha. And the second, also very agreed. I mean, to solve the PR issue with CDPR and Cyberpunk, IMO they never should have said "we're delaying it until December 10th." They should say "We are delaying it until 2021" (no date). That way, at the worst, you give the devs until December 2021 to get it done, and anything before that gets seen as a win to the audience. "Omg we were expecting this in Q4 but it came out in February?! Yessssss" (this was indeed mentioned in the article, but wanted to reiterate haha).

Also, high five fellow 8ish year professional dev! Hopefully you also got your start making PHPBB boards as a kid and such haha.

Yep. Quality and complexity are mutually exclusive for sure. Even the worst games tend to take people years and years to make. There's so many factors which can lead to these games coming out as bad as they do, maybe the devs finished their alpha build and the director hated it and they had to completely remake everything in a fraction of the time. Maybe they were a bunch of ienxperienced devs with no real guidance, mentorship nor leadership. Maybe the IP they had to work with (in the case of licensed games) was extremely restrictive, so they had their hands tied to the point where they couldn't make the game they had specced out. There are a ton of factors beyond the complexity of systems & visuals in play... Often times they're things we, the audience, will never see (unless a dev comes out and speaks out about it after the fact), and we are mainly left to speculate on...

Dawg. Thank you so much for your work. VFX especially - those people get treated like such fodder, and it is distressing. Much respect to you and thank you <3

That first story reminds me of my friends who work at EA. They talked about how a group of them set up an rv/trailer in the parking lot of the studio they worked at... The fact that it was normalized to the point where they had other people trying to sign up and get in on it made me feel icky, esp because some of them wanted it to be a semi-permanent thing.

Also yes, same. The grenades suck. We had one of those just last week and it's a sudden "well... Okay... We need to address this right now, and also make sure it is done in a way which doesn't bite our own ass 6 months down the line."

Generally agreed that unions help, and having them in place would definitely make things far, far more stable. I don't think it's the "only" option, but I think it would be a giant first step and would help an absolute ton haha. The problem is the ESA and other large entities in the games space. They actively push back against unions, and will not hesitate to fire anyone kicking up a stink about them.

To an extent. A lot of devs in those big studios are compensated very well, but that doesn't fix the (potentially permanent) mental impact it can have on a person. Like, for me, I got out of games where I was paid at 100% (let's say). I left for a company who paid me 75% of that, but I was INFINITELY happier. Having work/life balance was a godsend, to the point where the loss of income did not matter to me. I now make ~125% of what I did in games, have a work/life balance, and am generally comfy. A lot of my work colleagues are "recovering game developers," and also said they chose this life for the same reasons. Like yes, money makes an impact; but the human cost shouldn't be ignored.

Giving this one its own reply post haha.

Yes. I am also a dev who (now) works in the same industry. Kind of. It's adjacent. PM me if interested in particulars haha, don't want to doxx myself here (though, tbh, anyone could probably figure it out by googling my name and looking at linkedin but whatever haha)

I've put in many 50 hour weeks even in my current industry, but the thing is, no-one expects all 50 of those hours to be productive here. Like even at 40 hours a week, we're expected to maybe have 30 - 35 "productive" hours. Our brains aren't firing on all cylinders for all 40+ hours, and the fact that that's respected is such a godsend. In games, if you weren't hitting your metrics and producing and fulfilling your quotas etc, you were chastized and were at risk for losing your job.

It also reminds me of something I had to learn the hard way - thought about putting it into the article but didn't end up doing so, so here's an EXCLUSIVEfor y'all.

I learned to stop putting in overtime.

But why tho? Well, let's use some agile language which I didn't define in the article. When we estimate things, we tend to use "story points," which are numbers for effort (not time). Something could require an effort of 1, which would be relatively trivial, whereas an effort of 13 might take you an entire sprint.

Let's say at the start of a sprint you take on what you feel is a reasonable amount of work. Maybe a 3 pointer and a 5 pointer (these are all relative and subjective to the person and the team. If it turns out that 5-pointer was misestimated, and it's actually an 8 or 13 pointer, ruh roh. Suddenly you have a LOT more work on your plate, but people are still expecting you to finish it by the end of the sprint (accountability!).

So you put in overtime. You start working 10 - 12 hour days every day instead of your normal 8; but you don't really tell anyone, you just kinda do it after work hours. Now you've set up a dangerous precedent because all of your colleagues assume you can do that amount of work in a sprint, no problemo. If you stop being able to maintain that pace, it might reflect poorly on you - they'll ask "hey weren't you able to get the same amount of points done last sprint? Why can't you this time?"

So yeah, having put in that overtime, it means you've established it as a norm and you're couching yourself into putting in that much effort every week, which is ultimately self-destructive.

What's the point of all of this rabbling? This mentality is what is expected of you in the games industry. That you should be putting in as much effort outside of working hours as you can, because if you don't, then you "don't have the passion to keep going in this industry." It is, and I cannot emphasize this enough, fucked.

Regardless. High five, and thank you for the comment. Hopefully some folks are finding parts of the article & these comments educational and/or illuminating <3

Great post overall and I recognize a lot of the concepts from my school-time and you explained it well.

I can't speak much to game development beyond some smaller experiments with modding etc. However I can't help but notice that even if Ubisoft has a shitty track record of toxic assholes they sure do seem to have planning figured out pretty well for Assassin's Creed and soforth. Too bad that I have zero interest for their games except Beyond Good and Evil 2 but we don't talk about that game.

Obsidian actually seems to have learnt some lessons from their troubled history and Outer Worlds (though restricted in scope) on a new engine was impressive in the timeframe they made it. Haven't heard much regarding the quality of the DLC but I thought the base game was quite good.

Development is complicated in that it's not only cognitively taxing but also increasingly requires a team effort from several skillsets. However if you hit a consistent groove it can be creatively fulfilling so I understand why you might be enticed into spending free time on it but ultimately you will probably get exploited. Game development might not even have the option for end-to-end automated testing so Q&A must be a pain.

That's as much as I'm comfortable speaking to game development so the following is mostly me talking out of my ass about software development in general so YMMV.

I do agree that volunteering overtime is counterproductive in the long run and I try to address it as something only done for drastic necessity (or if I have an extreme interest in pushing an improvement) but I've probably hit the tipping point of punching in and out strictly for the hours paid, deadlines be damned. Paid Time Off is equally if not more important to me and I think unions (that I support and have had benefit of in my country) are good for pressuring for that but I'm not sure collective bargaining is realistic in the current landscape of projects and consultants. With time it might change though.

Personally I have never done anything productive for more than 50 hours a week and that's stretching it so when people describe going beyond that I guess they are counting at least 20% "water cooler talk". Unfortunately technology waits for noone and keeping up with frameworks, languages etc. might require at least some personal interest on the level of a hobby but I might be lucky that I find some enjoyment in it.

Ultimately I hope improved education beyond hardcore STEM subjects can help the aspirants being cycled in to understand the processes and pitfalls of software development (e.g. methodologies like waterfall, prototyping, agile) and give them the resources for planning consistently. Some of the subjects were abstract mumbo-jumbo to me in school but I'm quite glad I learned it because it seems like most of my current and previous colleagues are oblivious to it (to the detriment of work/life balance and deadline accountability that I have burned myself on as well).

EDIT: Also I'd suggest finding a better word than crunch because management is not "crunching" more work in to less time spent but in fact expanding/dilating work into more actual time spent within a shorter timespan with both short and long-term repercussions.