Chris Taylor says "no more overtime"

Saw this over on gamespot:

Better games are made when game developers work normal hours, Gas Powered Games head and Supreme Commander designer Chris Taylor says.

“We would have made a worse game [Supreme Commander] if we had worked 14 hours a day. Dungeon Siege would have been a better game if we worked the way we did on Supreme Commander. It’s just the right way to work, and it puts fun back in the business,” he said.

“You make better games when you work regular hours. You’re more creative–when you go home at night you’re still thinking, because you’re creative people. So when you get back at the computer, you have all these ideas and you get them down anyway. So we didn’t put a bunch of stuff in our office that would make people think that they should work there until late. We don’t want people to live in the office at Gas Powered Games.”

Taylor said one of his proudest achievements about working on Supreme Commander was the minimal amount of overtime his team had to put in to finish the game. He used the example of one employee at Gas who had joined from another company where he had completed more than 1,000 hours of overtime in one year. During the last two years this new employee was at Gas, Taylor said he racked up only 150 hours of overtime.

“I’m proud of that. Because it goes against everything we’ve been taught about,” he said. “I want everyone here to be filled with inspiration to move our industry in that direction. I feel like we’re saving something with our kids and our marriages–you know, marriages actually work a lot better when you spend time with your spouse.”

So are we going to see a turn around in 14 hour days? Are they really worth it in the first place?

Everyone wants to eliminate crunch. It’s just hard to do and still meet deadlines because everything is so fucked up with the game right up to shipping.

Well, speaking for myself, if/when I ever interview for a game company position again, I will be telling them up front that large amounts of overtime are out of the question; I’ll do what I think is reasonable, and that’s it.

Admittedly, I’m not exactly planning on working in games again.

It’s just hard to do and still meet deadlines because everything is so fucked up with the game right up to shipping.

Frankly, at least over here, most bigger productions don’t have the resources they actually would require considering the scope of the respective project - mismanagment aside.

Having to compete with products that have three or four times the budget yours has, does that. The lack of manpower again results in an increased need for crunch. And most companies aren’t exactly in the position to stand up and say that they’re not going to do that anymore.

It’s also amazing how many people still seem to think that twice the amount of time spent at work results in twice the amount of work done.


Bravo for Chris. It’s like the game industry is unable to read Peopleware.

Long hours results from bad managers.

Everyone already says this…at the start of their projects. :P

We’ve done some scheduling and estimation training, but still get deadlines imposed on us from upper management. The plan is to handle that by cutting features as necessary, but I imagine that’s a lot easier for business software where features are often less loosely connected than they are for games…

Which would make Chris Taylors’ statements, after Supreme Commander was finished and shipped, that much more relevant.

Hey, it is nice to see the gaming industry finally catch on to concepts that were formed in software development 30-40 years ago. The revelation has really been overdue.

#1 reason for software project failure, unnecessary schedule pressure.

That can be a good idea as long as there is some up front structure and planning as to what gets cut if what happens, and the “risk plan” gets updated as development realities make the situation change. If the idea is just “well if we run short on time we’ll find something to cut” chances are not so good.

I don’t know gaming software, but it is true that in most business software if you take care you can craft things that a very modular making it easier to plan for the possible non-existence of some features.

If you have a huge project and your producers have zero understanding of the agile methodology, then everyone will fall into the same trap of waterfall software development. The rest of the software industry is moving away from the old, terrible style of software engineering and towards proper processes that top load design and risk assessment and minimize fiascos at the end of the cycle. This keeps you from taking up those last two months in cram sessions (which introduce tons of bugs with sleep deprived programmers and zero unit testing or peer code review) and instead you can use that time to polish and QA.

It sounds like GPG is starting to clue in on those things, and I really hope other businesses do as well. My own job we’re just starting to do it, and while there are some teething pains I think the end result will be delivering better products on tighter schedules and still being able to go home at 5-6PM every day.

We’ve had pretty solid success with cutting stuff to stay on schedule an the dev house where I work. The trick is to identify very early what needs to be cut and when it looks like things could go bad cut before too much time has been wasted on it.

There’s also a fair bit of slack built into the schedule so that helps as well.

I agree with everything Kunikos said.

Someone needs to edit together a compilation of sections of books like “The Mythical Man Month”, an more recent works like Steve McConnell’s Rapid Development and make it required reading for managers, especially those that did not rise up from development. Title it something catchy like “How not to make your project fail predictably like every other project before it”.

Putting “oh crap, we didn’t think about that” time into the schedule, and periodic risk management are both great examples of best practices. Your manager gets a cookie. ;)

Actually, what Chris Taylor’s doing isn’t all that new, all the big companies are already doing the 40 hr work week maximum. EA, Activision, and even some smaller ones like Kush Games are all doing it. Granted alot of them are doing it because of big phatty lawsuits instead of the goodness of their hearts.

Official 40h work week maximums doesn’t guarantee people are working only 40 hours a week, especially at places like EA. It just means they won’t force you to do it. Instead, what they do, is that if you don’t put in that ‘extra’, you’ll be overlooked for promotions, raises, and bonuses.

This is the reason I declared my love for Mr. Taylor a bit back when he talked about it on next-gen

If I could pick any game company to work at, I’d pick GPG.

Crunch is the bane of this industry. We lose so many people because of it. The problem is that no one ever works from a good base and adds if they have time. They all plan huge and then end up chopping in a panic and having to scramble to make it all fit together. Poor project management is a prerequisite for most producers.

Managers that build crunch into the schedule from the beginning should be fired.

Exactly. When promotions come around, who is the boss going to pick? Programmer a who works 9-5 and occasionally an hour overtime here and there, or programmer B who put in 12-14 hours a day 6-7 days a week (all other things equal).

I really wish my boss understood that 70 hours/week office time at the minimum is just wrong.