Read this part of a developer diary of the game Mafia. You couldn't believe how long hours these programmers spend on the office.


2001-10-01
Alpha
by Daniel Vavra

Summer is over, and with it, the last August days when we spent the weekends partying – the only way to relax – is over, too. Now, there is only hard work ahead of us. MAFIA is slowly making its way to the Alpha version – and this Alpha version is the greatest and most hardcore we’ve ever come across. My colleagues who are writing scripts for missions often work 40 hours straight without any sleep, and spend their weekends at work – doing anything – just to complete everything necessary before the deadline. The programmers are even worse off – they’ve been having only a couple of hours of sleep a day for the last few weeks, and otherwise they’re simply working flat out. Luckily for them, there is a Tesco nearby so they can get some food even at four in the morning. Words like embolism, “chair sores”, heart attacks, and coma are starting to take on dangerously realistic meanings :) Anybody able to exert such effort has my endless admiration (and hopefully they’ll have some bonuses as well – one cannot live on admiration only, particularly not on mine!).br>
However, all this effort is clearly visible on the game. Most of missions can be played to the end, and it’s starting to be really fun. The characters are starting to talk to us, even though they’re doing it in not-so-nice and temporary voices. Artificial intelligence is also coming to life. There are still some bugs about - sometimes nice fat ones. But it is now completely about something else than it was in the beginning. It’s shaping up into a real GAME.
A lot of work has also been done on the visual effects. For instance, in one night mission, that takes place during a storm on a deserted farm, not only does it start raining but the sky is lit with lightning, trees sway in the gale wind, the wind tears leaves off of them, and all of the sudden everything is lit by an impressive explosion.

In a stressful situation, such as this one, there are naturally some hitches, and a multitude of shortcomings come up. As far as I’m concerned, all the irrationalities and blind spots in the design are making their appearances, and the programmers have all their mistakes in a seemingly finalized thing pop up to surface, so they don’t know which to fix first. But most of these things are really small matters that can be sorted out relatively easily, and I’m not sure whether it is actually possible to avoid them. In any case, after we’d come to the conclusion that some dialogues and some parts of the plot might be just a little too complicated, we’ve decided to have a professional playwright to go through the screenplay. The first attempts didn’t quite work. The classic syndrome of no name “artists”, who first announce that there won’t be any problems and then they gradually start professing that they are not going to put their names to some “action trash” because it would damage their reputation, became apparent. It is ironic that their names are mostly unknown and even if they were known, I don’t think anyone would really read them at the end of our game. However, criticism of that kind had hit me quite hard and I began to have doubts about the quality of our small work of art!
When trying for the second time, we aimed somewhat higher and approached one of Czech film directors. If it is really so bad, I wanted to hear it from someone whom I can trust. The concerned person, whose name I’ll keep to myself for the time being because our co-operation has only just begun, had read the screenplay and after some days of tense waiting told me, that he quite liked the screenplay and was going to have a go at it. I was so relieved. Tomorrow we’ll start with editing, so I hope it’s going to work.

Today we were supposed to hand in the Alpha, but as there are still a couple of things to be done, we’re going to be a day or two late. I was having a smoke between the settings of parameters, when one of our graphics told me that some terrorists had hit the World Trade Center with a Boeing. I couldn’t believe it, but just as I got to CNN, the skyscraper was collapsing. I felt ill. As if someone had unexpectedly kicked your balls from behind. This is just not done.

Hmm. Surely I hope all the hardwork was worthwhile. Wish the game was beautiful, they made a lot money, hard workers got bonuses, time-off etc. I dont take the post-mortems and dev diaries too seriously because everybody tries to look good describing themselves - hey, look how hard we worked on this game, you ought to go out and buy it.

Personally, I have worked hard at certain times on my projects but on majority other days was doing nothing productive. Cycling, going to art galleries, spending time with friends and family etc. Sometimes other activities directly help my work. At the very least they make me feel good and probably do my work better.

I dont intend to worship hardworkers because to me that is not a lifestyle I want.

It was an excelent game IMO. One of the better games of last year. But i wouldn't be worried aobout chair sores or coma's in that type of enviornment. When your that consumed by something the your most prominent threat is the big "D". Yep deviorse... countless game programmers leaving a trail of broken families behind them.... oh dang.. better get off this soapbox. ;)

Divorce and deteriorating health are huge concerns on a project like that. But what gets me is that people pretty much expect that that's just how game projects go. It's true that there are always unknowns -- things you just can't predict -- but that level of death-march is, IMO, inexcusable: it's a failure of project management. (Though jaggu, FWIW, I find many post-mortems well worth reading. You have to infer a lot of the pain and angst that they paper over, but there are also many good lessons learned you can glean.)

I recently picked up a book btw, Under Pressure and On Time (http://www.amazon.com/exec/obidos/tg/detail/-/073561184X/qid=1045335538/sr=8-1/ref=sr_8_1/002-5145621-8222456?v=glance&s=books&n=507846), a Microsoft Press "best practices" book. Not a lot of earth-shattering stuff in there, but a ton of stuff done right. Some very clear discussions of how to avoid project development traps and then fix the ones you fall into.

I feel the culprit behind this is working to a deadline. The publisher threatens, "If you dont give us the Alpha by March 3, we pull the plug!". Whenever there is a deadline, one doesnt work well (sometimes not at all) when the deadline is far away and put needless extra hours when the deadline gets closer and closer. This used to happen all the time at university. People would put nite-outs on the night of the test and party the other nites. The ideal situation would be to study some all nites but this simply doesnt happen. Thats how the human mind works I guess. It seems to need the Carrot and Stick.

I think we should throw away the deadline altogether. It gets done when it gets done. Indies have the unique opportunity to work like this.

Only if your bank account approaches being infinitely large, and/or you're not trying to make a living from your games.

Contrary to popular (romantic) belief, constraints such as deadlines do not inhibit creativity or the ability to get things done -- used correctly (and professionally), these can enhance your creativity and productivity.

A game without a deadline has a much lower chance of ever actually being completed.

OTOH, the boat I'm very very slowly building may well end up being named "No Deadlines." That's one of the few areas in my life where I allow that indulgence.

I love reading post mortems like this. It reminds me of why I will never ever work as a programmer in the games industry. It's a mugs game.

Originally posted by Davaris
I love reading post mortems like this. It reminds me of why I will never ever work as a programmer in the games industry. It's a mugs game.

Hehe yeah, I recently got a chance to talk to a person working as a programmer in a small game dev studio. He said they're finishing up the game right now, and he has to work every day including weekends. Anyway, from talking to him I basically confirmed all the bad things I heard about the game industry. Long hours, small salaries (he said I might have to almost work for free at first if I want to get into the industry).

He also said the salary roof for a programmer is around 70K a year in general (and starts at around 30K). Of course, he said that in spite of all those things it's very fun. :) I used to think that was the most important thing in the world -- to do what I love, program games. Then slowly over the last year or two I began to realize that working insane hours for little pay, hurting my health, my social life, etc., and most importantly doing all this for someone else is not what I want. I realized that a "regular" job is much better than working in the game industry. And that your own business is better than either of them. :)

What sucks is when developers work that many hours game testers usually pull the same or even more hours to make up for it. I pulled a few 24hrs myself game testing and I wasn't even a developer :). I also noticed during those 24hrs half the crew had no idea wtf they were doing, and some individuals were straight out sleeping. This is including the leads.

So I don't really see the benefits of working 24hrs if anything past 12-16 you are basically useless. I guess it's not so bad for the company paying salary, but testers are on the hour and I bet some of the guys enjoy getting paid to sleep :).

John

Worst stint I ever had was 6 months getting about 5 hours sleep a day, for 7 days a week. Usually doing a 2 day stretch and then getting about 6-8 hours.

Other guys on the project were doing 3 days work then getting 12-14 hours sleep.

The trouble is that Xmas is a pretty hard deadline to move :), and common belief is that if you miss Xmas you miss most of your sales.

Turned out we missed Xmas (duplicators dont move like we do) and still sold extremely well.

I thought that that was too much though, we were happy to do it, but basically got NO rewards for working ourselves like that. Not even a nice big "thank you".

Worse was to come when we had to do the same thing again the next year (i.e. hit the xmas following the one we missed with the sequel).

Now, I wouldnt dream of doing those kind of hours and putting in that effort without some MAJOR benefit IN WRITING. If they want that sort of thing again, then I want to see something out of it. You can only stay naive for so long.

Guys. If someone asks you to work stupid hours like that in the games industry, theyre DEFINITELY pulling a fast one, make sure you get a very hefty bonus to pay for it all (they sure will be).

I lost the chance of buying a house, my car ran out of tax while on the road and I got fined etc. All because I was doing the death march *grind* on a project. Not worth it.

.Z.

Personally I think there's more fiction then fact in that diary. Have any of you ever tried to work 18 hrs straight? how about 24? You actually start loosing touch with reality around 25-30 hours of being awake. Halucinations start occuring... 40hrs straight work? If you believe that I have a bridge to sell you :)

While I dont doubt they expirence some hard core cruchtime.. I think the description has been embellished somewhat..

Deadlines give you focus, something to measure your progress against. Yes, with a game, where your trying to "build fun", its hard to say when it will be done, but you still have to workout a schedule and try to stick to it.

In every software engineering book I've come across, they say scheduling is hard, but, except for the smallest of projects, it has to be done. You'll probably be way off initally, but over time, and with experience, you'll get better at estimating the time required.

Also, the problem with a lot of game schedules is they're over ambitious and unrealistic. Almost every Post Mortem I've read in Game Developer Mag or at Gamasutra.com, the what-went-wrong section includes "poor scheduling" and "bitting off more that we could chew" have having to pay for it with long hours.

As far as the ole' "It will be released when its done" motto: One company I know of that still uses it is 3dRealms/Apogee (creaters of Duke Nukem), and their recent track record speaks volumes:

Duke Nukem Forever: In production since 1999, SUPPOSED to be out this year.

Prey: In development since 1997(!), no realease date in sight.

David

In the game industry it's indeed well known that you need to make long hours and the salary is poor. Not only in the USA. In Holland there was a game company where the development team also worked in the evening mostly, because of the deadline. So thats another reason why it's better to go indie, so you aren't focused on the deadline of the publishers marketing guy. You can schedule your own time.

Dan, really.. you think its abnormal to work for 18hours? I can defintiely tell you it isnt.

Yes, you DO start to hallucinate after 2-3 days awake (at one time during a project, i was talking about some fairly simple graphics code to another programmer who'd been awake the best part of 3 days, when he started calling me "butler" and talking about some rocket car. At that point I MADE him get some sleep.

We regularly during that project were working 48hour stints.

The point being, that it IS possible, is it good? no, but when youre deadline is Xmas, you cant move it. When youre young and inexperienced (and hungry to ship games) you'll do a lot to push yourself.

These days I wouldnt think of doing such mad things, especially without some MAJOR compensation. But its NOT that uncommon for unscrupulous or incompetant employers to ask for that kind of effort.

Dont think these stories are exaggerated. I am being totally truthful. Its not a stage of my life I'd like to relive.

.Z.

As far as the ole' "It will be released when its done" motto: One company I know of that still uses it is 3dRealms/Apogee (creaters of Duke Nukem), and their recent track record speaks volumes:

Another company that has this motto is Blizzard Entertainment.


Personally I think having a schedule is a good idea. However if you are producing garbage and the game doesn't sell, what is the point of reaching inflexible milestones?

Heh yeah but when they started thier new project i think they said "hey lets work on duke nukem forever" but i dont think they understand that forever was just part of the title

Originally posted by Fenix Down
He also said the salary roof for a programmer is around 70K a year in general (and starts at around 30K). Of course, he said that in spite of all those things it's very fun. :) I used to think that was the most important thing in the world -- to do what I love, program games. Then slowly over the last year or two I began to realize that working insane hours for little pay, hurting my health, my social life, etc., and most importantly doing all this for someone else is not what I want. I realized that a "regular" job is much better than working in the game industry. And that your own business is better than either of them. :)

Well, actually you can make significantly more than 70K (USD) per year programming games, but still not enough to warrant the kind of treatment that's considered "industry standard" (IMO). At times I really miss working in the "biz", which was a large motivation for striking out on my own, but I'm all but certain I'll never go back if I can't make it on my own.

Originally posted by zoombapup
But its NOT that uncommon for unscrupulous or incompetant employers to ask for that kind of effort.


And there are a lot of unscrupulous and incompetent employers out there, looking to exploit inexperienced programmers/artists/producers. However, I also believe that many people that go into games now expect this kind of treatment and therefore are willing to work ridiculous hours without making a peep (or a lot of money). And as long as colleges and universities keep cranking out fresh meat, the retail games industry will keep chewing them up and spitting them out. :(

Originally posted by zoombapup
Dan, really.. you think its abnormal to work for 18hours? I can defintiely tell you it isnt.


Heh, no I worked for Microsoft, I'm plenty familiar with 18-20 hr days. The point that I was trying to make is that they are painful hrs to work. Anyone who's worked them will attest to that. I've also stayed up for ~36hrs straight, I wasn't in any condition to function as a normal human being, let alone work :)

But I wont call you a liar, the concept of working 40 straight just reeks of all kinds of insanity...

Yeah, well, its definitely not a thing I'd recommend. As I said, my colleague was hallucinating badly after 3 days awake (on top of months of 2 days on then a few hours asleep).

Of course, go to any demo party, and theyre doing almost 5 days awake! idiots! but doing it at work is even worse.

Dont let anyone fool you into doing those sorts of hours for your normal salary EVER.

.Z.

Read this http://www.smh.com.au/articles/2002/10/10/1034061260831.html

some guy in korea died after playing on the computer for too long.

Actually i've read two seperate accounts of korean guys dying in game houses in south korea within a month. The have some crazy hard core players over there....

I'm curious to know why people are willing to work such long hours with no extra compensation. What happens if someone tries to leave after working only 8-10 hours? Do alarms go off? Does your boss try to manipulate you by implying that you aren't being a team player? Or is it nothing more than peer pressure? I.e. everyone's staying late, so whoever leaves first looks like a quitter. Do the major stakeholders and executives work crazy hours too (i.e. leading by example)? What is it that gets people to kill themselves over something they don't own?

I've only been an employee for a total of six months of my life, working for $6/hr for a retail video game store. When my shift was up, I went home. I've never experienced what it's like to work a salaried job in the gaming industry. It sounds like if you get a job like the ones described in this thread, you're still basically working for $6/hr -- you just work a lot more hours. :)

The longest hours I've worked were not in the game biz, but in the EDA industry. Other parts of the software industry have their crunch times too. I've done several 100+ hour weeks (not easy to do; in my experience most people who think they're working that much tend to be topping out at about 70-80 hours) in and out of the game industry, and I hope to never do another.

In our case, we worked that long and hard due to a combination of peer pressure, not wanting to fail, not wanting to let anyone else down, silent guilt over having mismanaged the process and/or frittered away our time earlier, disbelief that we were really more than a few hours from being done (oh sure...), and basic concern for our jobs. Alarms don't go off if you leave, but people who leave do get fired if the project doesn't ship or doesn't make it to the show floor or whatever. Or if they aren't fired, they're just relegated to an uninteresting project with no hope of promotion.

Yeah carrier advancement was a big thing at Microsoft, every 6 months you had a review where you could be allotted shares and a raise. You were ranked against your fellow co-workers and then rewarded appropriately. In my case I was working with a number of guys from India and Pakistan. Great guys, very motivated and hard working. Smart too.. however they didn't have any family or friends over here (for the most part) and were able to commit countless hours at work. I had a wife and a kid and after I while I realized I just didn't have it in me to compete with these guys.

I even tried optimizing my work so i was at 90% efficiency and worked a solid 8 hr day. I was doing as much work as the other guys and not missing any scheduled tasks, however the fact that I was around 1/2 as much as the other guys made management almost put me on a performance plan. They were more interested in the time I put in, then the work I put in. After I realized that I got out of there as fast as possible. Now I’m a contractor and loving it.

I should say that Microsoft was a great company to work for. It has this odd feature however in that the entire thing runs like a bunch of autonomous cells with no standardized processes across teams. So some teams are great to work for and others just stink. As long as you get on the right team you will have a wonderful time. Get on the wrong one and it's going to hurt....

Ugh, MS reviews... I hated that whole grading on a curve thing when I was fulltime. Contracting is so much better.

We did it out of the want to make the best product possible whilst knowing we had little time. (Naivity).

One thing that interests me, I constantly hear people say they "contract". What exactly do they contract to do? I mean, is this writing software?

How do you get into contract work?

I'd like to be able to find some flexible work in case my own development drags on a little too long (to top up the funds for instance).

.Z.

Originally posted by Dexterity
I'm curious to know why people are willing to work such long hours with no extra compensation. What happens if someone tries to leave after working only 8-10 hours? Do alarms go off?

I tested that theory myself. At the company I worked for, half the staff were fired to cut costs, leaving the rest of us to work that much harder. Not only were the salaries shockingly low, but you were "encouraged" to work around 3 hours UNPAID overtime every day. It took me about a week to realise just how pathetic this was, so I stopped doing it. I started getting to my desk for 9:00am, and leaving on the stroke of 5:00pm since these were the hours I was actually paid to work.

The rest of the team continued working crazy hours for no extra money (and this wasn't in a "crunch time"). To be honest, they seemed scared to lost their jobs if they -just- worked regular hours. After a while of constant nagging that I should work more unpaid hours "because it's what you do in the games industry!!!", I left. The nice thing was that the shock of me leaving kind of snapped them out of their almost slave like mentality, and in the following weeks half the rest of the team also left. :)

Originally posted by zoombapup
One thing that interests me, I constantly hear people say they "contract". What exactly do they contract to do? I mean, is this writing software?

How do you get into contract work?
In my case, I'm working through a contracting service. I'm signed with ArtSource, and when on contract I'm actually their employee, just working at MS until my contract is finished. Contracting is like getting short-term day jobs that last up to a year at MS. The main advantage of being a contractor is that you get overtime pay. For some managers, that means you don't work overtime, and for others, you do, but you make a lot more money in the process :)

The downtime between gigs can be a drag if you're not prepared for it; MS rules insist on 100 days off after a year's work for contract staff. I'm putting that time to use during my current break by seeing how much game I can complete in 3 months, and doing the odd freelance art to stretch my savings a little.

Yeah, there is a very real slave-like mentality in a lot of crisis-managed places.

You get used to what you know, and it cuts both ways. In the mid-80s I worked for what was at the time an enormous engineering company (Tektronix, about 25,000 employees). We worked maybe 40 hour weeks. Maybe. 2-3 hours of frisbee golf at lunch (there was an 18-hole frisbee golf course on campus) a few times a week was the norm.

And then when the company's overall brontosaurus-like agility and productivity started catching up with people and products, some folks were actually offended -- as if somehow they deserved such a cushy job.

Back on the crisis side, while sometimes you can leave at the stroke of 5pm, sometimes you can't. Sometimes the fear of losing your job is real. And reality checks here (are we being slaves, or are we justified in using our anxiety to push the product forward?) are useful.

As just one, I've talked with many C++ programmers and web developers recently. The job market for these stinks. I think maybe we're past the worst of it. But one talented programmer I know -- a guy who was a permanent contracter and who refused to work overtime a few years ago -- has had one contract in about the past two years. If asked, he would certainly work the overtime now.

So sometimes -- whether it's to further your career or keep your job -- a sustained extra effort really is called for. Deciding whether your particular situation at any given time is one of these situations (or whether you're falling into either the slave mentality or the entitlement mentality) is often a difficult call.

Originally posted by alchemist
As just one [example], I've talked with many C++ programmers and web developers recently. The job market for these stinks.

I'll second that opinion. I didn't have to burn the ships--they burned without my help.:eek:

Currently I contract though a company named Venturi Technology Partners (formerly Best Consulting). I get paid for every hour I work which I love, but gave up all my paid vacation days which I don’t love too much. I haven't ever really had a hard time finding work, all my jobs have come because I knew someone working there and they recommended me.

My company tells me that the field of available jobs is pretty small right now. I am employed for 6-month contracts. I'm currently on my second contract and Boeing doesn't have the same restrictions Microsoft does about contractors. (Probably because they haven’t had a class action lawsuit brought against them yet) The project I'm working on has a good year or so left on it so I’m not too worried.