The issue of Quality of Life for game devs is back on the news: Feat. R* San Diego

[obonicus]

The lazy developers thing has weight too, though. Not every developer is lazy, but look through joker's posts for his comments about how PC ports are carried out. Someone in that whole process is lazy, and it's not the consumer. It might not be Joe Schmoe Dev, but if it's not his boss, it's his boss' boss. They're still devs.

As to poor planning being to blame, yes it is. If software companies can't negotiate contracts that aren't below what making the game would actually cost, hours-wise, if the business model is unsustainable, the solution seems not to continue on that business model. Other than the terribly weak 'games are a creative endeavor' defense, no one here has been able to explain what makes game development so unique that most of the other best-practices in software development don't apply to it.

 

[ERP]

It's the rate of change of the industry that differentiates game development. And partly the passion of the developers.
10 years ago I could build a game with 5-10 guys in a year, the model for managing that is radically different than 20-50 guys over multiple years, which is different to 300 people over multiple years.

The same is true for changing ROI's 15 years ago an ROI of 7 to 1 wasn't unusual for a game, now it's probably closer to 1.7 to 1.

Things change faster than the industry and fundamentally the people in it react, people managing products today were developing in a totally different environment 5 years ago and more often than not they don't see or understand how the difference affects development.

As to the passion, it leads to people wanting to do whatever it takes to ship a product they feel invested in. It means everyone on a team wants creative input and it's a fine line that gets progressively harder to walk as team sizes grow between making people feel involved and desgin by commitee.

One of the trends I saw before I bailed (at least in the short term) on the industry in late 2008 was people entering the industry and lasting exactly one ship cycle. Sometimes because of hours but more usually because they didn't feel they had enough input in what was being built and the investment wasn't worth it.

Most game developers don't do it for the money.

Port houses are a somewhat different beast with somewhat different set of rules. There is an enourmous differece between porting a product and building an original IP, a lot of dev houses have collapsed under the weight of trying to transition from one to the other.

 

[obonicus]

But that's a matter of shifting budgets. Within the given budgets, again, what's different between these software products? Just that the people working on them don't actually have the experience of managing very large projects? Project budgets grew too fast for prior experience to be used as a gauging tool?

I mean, most of the stuff you list is an explanation for why developers accept these working conditions, they don't really explain why project management isn't more of a discipline in game development. Even going by your earlier comments, it seems that a lot of these projects start before people even have a clear idea of what they're working on.

Mind you, I'm not saying game development houses have to be CMMI level x or use Scrum or anything silly like that but there does seem to be a reason for concern if studios that can deliver product reliably are listed as the exception.

 

[joker454]

The one studio I've heard of that doesn't crunch, and kicks everybody out of the door every day at 17:30, is Neversoft; the legend went that they shipped every year their Tony Hawk on the 30th of September to the factory, took one month of vacation, did one month of preproduction, and on the 30th of November pitched the new Tony Hawk to Activision. This is probably why Activision chose them to take care of the golden goose aka Guitar Hero.

Do you know if that's a recent change? I was last over at Neversoft in 2007 talking with some of the people there, and at the time they had was was called "hardcore hours" where the last week of every month was 9am to 9pm. The idea was to distribute crunch over the course of the project instead of all at once. Here's a quick article I found on that:

http://altgames.vox.com/library/post/neversoft-neversleep.html

So they effectively have 12 weeks of crunch per year, just spread out.

 

[ERP]

" it seems that a lot of these projects start before people even have a clear idea of what they're working on."

There is an element of this, I've always said the biggest problem with development at EA was the inability for teams to control head count. Head count is controlled by when other projects finish.

Companies go through the motions of preproduction but often learn very little, designs are approved massively over scope, there is a resistance to cutting features until too late in the project.

There is an enormous difference between running a project with 10 people and running one with even 50. It's a more complex problem in games where every dev wants to be a designer aswell.

In terms of project management methodology I've seen just about everything tried, I've seen some things work for some teams, the problem is the relative fluidity in featureset. I've worked on other large scale software projects and there is just nothing even remotely like the rate of change of features and priorities. A lot of it come downs to the difficulty understanding how a game plays from feature descriptions, how gameplay elements interact in combination, or estimating the ammount of time to polish a feature. Feature breakdowns are just inadequate to describe the actual work. In good functional teams, there is an understanding of the "tax" imposed by this, different teams deal with it in different ways, but it's dependant on the people involved and it's less common fo teams to remain intact game to game.

If you look at the companies that create consistently high quality original games, they rarely do it to timeframes, and commonly miss deadlines. Itterative improvements to existing franchises like ports are a different issue, and a lot easier to manage because feature prodution costs are understood, although I've seen that fucked up before now.

 

[Cheezdoodles]

The same is true for changing ROI's 15 years ago an ROI of 7 to 1 wasn't unusual for a game, now it's probably closer to 1.7 to 1..

This is supposed to happend in any industry where competitiveness increases

 

[Sdw]

I'm amazed that the game industry hasn't matured more, this kind of crap would never fly in other software development sectors.
And yes, it IS bad management, or rather, they get away with bad management because the devs put up with it.
Most studios have so many game project under their belt now that there should be no problem at assigning an adequately correct time budget when they start up a new project, but instead they set unrealistic goals both when talking about features/content and deadlines, and instead of speaking up, the devs seems to just silently accept unrealistic hours and pressure.

Back when I was a kid in the 80ies, I dreamed of becoming a game programmer. Many years later (late 90ies) I had the opportunity, but luckily a friend of mine had graduated a year earlier and was already working in the industry, so I had the inside info on the conditions. I decided to pass on the offers I had got another 'normal' dev job instead, and kept the game-development as a hobby. I think I made the right decision.

Another thing I don't get is the US "salary" system. It just seems retarded, it¨'s basically a contract that says:
"We, the employer agrees to pay you exactly the sum of X every month. For that sum we expect you to work an undefined amount of time that we decide as we feel like."

Eh? I mean, to me a normal contract is, "We pay you sum X and for that we expect you to do Y hours of work"

I work 40 hours a week, and if the company need me to work more (which sometimes is needed) I either get the extra time as payment, or I get the same amount time off at a later time when things have slowed down. If they want me to work saturdays or sundays, the same applies, but there is a multiplier (can't recall exactly, I think it is around 2) for each hour worked.
That way, the company can get that bit of extra work out of employees if needed for a "crunch", but they have to take into consideration that it will cost them more, therefore they gain a lot by planning ahead instead.

The system where the extra hours come at the same cost as normal hours, or even worse actually come for free, is totally mindboggling. Of course the company would want you to work more if you do it for free!

 

[Cheezdoodles]

That is not the US salary system.

That is standart for any job with high enough salary to not warrant "overtime" bonuses.

Basically they pay you x amount per year to get the job done, whatever it takes.

 

[Sdw]

That is not the US salary system.

That is standart for any job with high enough salary to not warrant "overtime" bonuses.

Basically they pay you x amount per year to get the job done, whatever it takes.

To me that is a bit strange, maybe it is just a local thing, but at least around here most "normal" employees in the software business (not talking about game dev) are on an hourly wage.
The people on fixed salary are mostly people in a more independent position, ie. they more or less decide their own hours (as you say, as long as the job they have signed up for gets done).

What these game developers seems to have is the worst of both worlds. They don't get compensated for their extra hours, but they don't have any freedom over their own time (at least it doesn't seem like that, if someone can ORDER them to work saturdays etc.)

 

[Acert93]

Big industries get big exemptions via lobbying and threats of moving. "Look, if you migrate game development out of the 'entertainment exemption' clause for salaried worker compensation we move out of California to XXXX. And you lose your next election."

 

[Silent_Buddha]

Big industries get big exemptions via lobbying and threats of moving. "Look, if you migrate game development out of the 'entertainment exemption' clause for salaried worker compensation we move out of California to XXXX. And you lose your next election."

Considering the unfriendly business atmosphere in CA with regards to taxation, regulation, and cost of services, I'm actually surprised many companies choose to remain in CA.

For example I think the only reason Micron remains a viable memory producer is that they are based in ID. If they were in CA, I'm fairly certain they would have gone bankrupt by now.

Anyways, that's getting into territory that might quickly relegate this to RPSC.

Regards,
SB

 

[ban25]

Mind you, I'm not saying game development houses have to be CMMI level x or use Scrum or anything silly like that but there does seem to be a reason for concern if studios that can deliver product reliably are listed as the exception.

I would say it's very presumptuous to assume that people are doing something terribly wrong if they slip 6 months on a 3-4 year project. Most major game projects are using Agile these days, which admittedly has plenty of its own problems. Before that, studios used XP, Gant charts in MS Project, you name it.

 

[ban25]

It's not the norm in software development, though -- not that people don't work heavy hours, but rather it seems since every game has a crunch period, then the majority of these games are poorly planned and probably mismanaged.

You obviously don't work in the valley.

 

[ban25]

As far as knowing game companies with better employee conditions, I don't know of any standouts that I can use for comparison. All the places that I know of that have shipped product all have crunch. They aren't all "EA Spouse" level of bad of course, but they all crunch. Some places like Infinity Ward do give out incredible bonuses, so that certainly takes the sting out of it! Most places though as you can imagine don't compensate the developers anywhere near that level, and many don't compensate them at all. I wish there was a normal work hours studio out there that has shipped product that one could compare with crunch heavy places, but I just don't know of any. Any one know one? I mostly just know studios in USA and Canada, maybe some people at European studios can chime in. I'm curious how work conditions are across the pond.

Working for large developers is much easier than small ones. They just have so much more in the way of resources. Being successful doesn't hurt either. Down the street, there's a studio that's struggling to survive as we speak, and the employees are crunching to produce prototypes in hopes of getting a contract before they run out of money. We'll see how it goes in the next week. I, for one, am glad not to be in that situation, having been there in the past.

 

[ban25]

Big industries get big exemptions via lobbying and threats of moving. "Look, if you migrate game development out of the 'entertainment exemption' clause for salaried worker compensation we move out of California to XXXX. And you lose your next election."

It's at-will employment, as specified in the contract you sign. It would be difficult to justify much else given the six-figure salaries. Some large publishers moved junior employees to hourly pay after the EA Spouse incident. I don't know whether that was ultimately a good move or not, since many of those guys (QA, junior engineers, etc.) now make considerably less pay when they can't get the overtime approved.

 

[Acert93]

It's at-will employment, as specified in the contract you sign. It would be difficult to justify much else given the six-figure salaries. Some large publishers moved junior employees to hourly pay after the EA Spouse incident. I don't know whether that was ultimately a good move or not, since many of those guys (QA, junior engineers, etc.) now make considerably less pay when they can't get the overtime approved.

You are the first person I know who has suggested such low level developers are making 6 figures in general...

As for the last comment, that sounds like a management assessment; most employees are pretty happy to be paid for the hours they work. If they get the weekend off and management cannot get things in over to, you know, pay you for overtime, maybe the company needs new management. Or maybe they just cannot justify, you know, paying people for the hours the really work? Or justify the hours "on the books" whereas blanket salary slaves as in the past. Every friend I have had who worked/works in the industry have the same story and it is the same reason they discourage family people for even bothering.

But I am not sure how it is a "bad" thing when management cannot cough up the money or justify the punch card numbers with willy-nilly "everyone is working 12 hours 6 days a week for 2 months". Being accountable for budgets, time and money, is a good thing.

 

[obonicus]

I would say it's very presumptuous to assume that people are doing something terribly wrong if they slip 6 months on a 3-4 year project. Most major game projects are using Agile these days, which admittedly has plenty of its own problems. Before that, studios used XP, Gant charts in MS Project, you name it.

Using buzzwords doesn't help anyone if you can't do it right. And yes, a delay of 6 months on a 3-4 year project is a big deal in other areas of software development. (Assuming not a renegotiation, of course.) I'm not sure why game development would be the exception. It means someone along the way screwed up. I understand that the scope is much more fluid with game development, but that's clearly part of the problem. You have stories of putting bodies to work on a project before anyone even has a clear idea of what the project is. How the hell does anyone mitigate risk in a situation like that?

You obviously don't work in the valley.

Obviously not. Working in the valley means bush-league project management? It may be a revelation, but other areas actually get software development down to a process. And in fact, the studios that can deliver software reliably have remarkably high reputations.

 

[MfA]

EA has a process down on their sports games, but how many game types are there which are cookie cutter to that extent? Anyway, as soon as there is a big human component things go to shit regardless of area AFAICS ... what percentage of IT projects fail again? If you only have to deal with other hardware or with small groups of trained professionals (ie. embedded software) things are a lot easier than if you have to deal with the fickle demands of gamers or office workers, also areas where you get pay'd up front for a one of a kind piece of software isn't really comparable to commodity software (no matter how much it sucks in certain ways, if there is no alternative and it works it's still going to get used).

If the game isn't fun it doesn't matter if it's on time.

 

[obonicus]

EA has a process down on their sports games, but how many game types are there which are cookie cutter to that extent?

HD games? How many are third-person shooters? How many are first-person shooters? Not to the same extent, naturally, but are you saying that all of EA's prior experience can't be used as a metric for their next projects? Which is to say nothing of sharing tech/experience between games of disparate genres. Even if every game deserves the right to develop its own custom engine (something I've argued against in the past) certainly the amount of work involved in endeavors of that nature must be known, right?

Anyway, as soon as there is a big human component things go to shit regardless of area AFAICS ... what percentage of IT projects fail again?

Oh, a huge amount, possibly 70%. Does that exclude games, though? It seems the '70% break even' number refers only to released games. The number of game projects that are actually successful might be significantly lower than 30%, going by that.

If you only have to deal with other hardware or with small groups of trained professionals (ie. embedded software) things are a lot easier than if you have to deal with the fickle demands of gamers or office workers, also areas where you get pay'd up front for a one of a kind piece of software isn't really comparable to commodity software (no matter how much it sucks in certain ways, if there is no alternative and it works it's still going to get used).

I'm not sure what you mean here. Aren't games one-of-a-kind software? In fact, games are even more one-of-a-kind in that they're not entirely fungible.

I mean, yes, software projects are incredibly difficult to get right. Which is why there's so much money invested in project management, why there's so much talk of best-practices and so much faith put in PMBOK and software maturity. But how much evidence do we have for software maturity in the games industry? I can in fact think of a few studios that are highly reliable, and can guess about a few others. But otherwise we hear about projects that seem to have started in a disastrous state!

If the game isn't fun it doesn't matter if it's on time.

Sure, but time doesn't translate to fun either. In fact, isn't it the other way around? Heavily-delayed games tend to be the messes you'd expect. And in fact, can't 'fun' be gauged in pre-production, on prototypes, before you have a full team working on some protean project with ill-defined goals? Can't teams lock down the scope before full production?

 

[MfA]

I'm not sure what you mean here. Aren't games one-of-a-kind software? In fact, games are even more one-of-a-kind in that they're not entirely fungible.

They are one of a kind, but they aren't without alternative ... because they are luxury not necessity. A one of a kind software project meant to run a factory is generally without alternative.

Areas of software development which are not commodity can't really be compared to games. If we take the grand daddy of commodity software development, Microsoft, slips are once again pretty common.

I mean, yes, software projects are incredibly difficult to get right. Which is why there's so much money invested in project management, why there's so much talk of best-practices and so much faith put in PMBOK and software maturity. But how much evidence do we have for software maturity in the games industry? I can in fact think of a few studios that are highly reliable, and can guess about a few others. But otherwise we hear about projects that seem to have started in a disastrous state!

I'm not disputing this ... I'm merely saying that in other areas working with similar constraints as game development (for instance IT and consumer software) delays are no more rare than in games development.

IT definitely has it's share of DNFs