I’d argue that a lengthy crunch pretty soon leads to less work overall being done than there would have been without the crunch. The tiredness starts to bleed over into the regular working week making those hours less and less productive. Instead I’d say that overtime allows middle managers to present the illusion of ‘we’re working as hard as we can’ to the upper management, while the reality is that less work is being done. Its all about appearances not actual ‘short-term efficiency gains’.
That leaves me with two questions. (And I really am curious here, not accusatory).
How much crunch are you talking about? All the data I’ve seen suggests that after two weeks of 40+ hours, employees become less productive, and that it takes them several more weeks to recover.
Also, if the crunch is driven by the ambition of the team and not inaccuracy in the schedule, why is it top-down and team-wide instead of on an individual basis by people who want to do a little extra?
I’m a lot more likely to believe any claim that “crunch works great for us at company X” if it actually comes from a programmer or artist than if it comes from a producer. :)
I actually earn much more as an indie than I did at Lionhead.
just saying :D
Right, but that wouldn’t be the case for me unless I created the next Bejeweled or something. :)
Crunch is never a technical or creative necessity. Crunch is how management gets more work done without more paying for more people.
That’s the cynical view, sure. Again, I’m no fan of crunch but it has it’s place in the short term.
Also, bear in mind that I’m not talking about 16 hour/7 day a week death marches here. I’m talking about 10-12 hours a day, 5-6 days a week. It isn’t insanity or anything and can be sustained for a month or so before shipping. It’s long but it’s not soul crushing.
Kyle
How much crunch are you talking about? All the data I’ve seen suggests that after two weeks of 40+ hours, employees become less productive, and that it takes them several more weeks to recover.
At the end of the project, it can be a few months of extended hours. It ramps up generally. We start with 10 hour days, and then after a few weeks throw in an 8 hour Saturday and during the last few weeks it may crank up to 12 hours a day during the week.
We get comp time afterwards and are financially compensated as well so I know my attitude will differ from those who get jack shit for crunching.
Also, if the crunch is driven by the ambition of the team and not inaccuracy in the schedule, why is it top-down and team-wide instead of on an individual basis by people who want to do a little extra?
My understanding is that it’s viewed as demoralizing to have certain people crunching while others are allowed to leave. I can see that but I don’t wholeheartedly agree with it. If I’m caught up and I want to go home, I believe I should be able to. However, in my experience there is almost always something you could be doing to move the game closer to shipping.
I’m a lot more likely to believe any claim that “crunch works great for us at company X” if it actually comes from a programmer or artist than if it comes from a producer. :)
Oh, no doubt. And you won’t hear me saying that crunch is awesome and that I love it. However, I do think it’s a necessary period at the end of a project that can last a month without adversing affecting anything or anyone.
A 6 month death march? Die in a fire.
A month of extended hours to get the game as good as it can be? OK.
If I recall correctly, the crossover point of 60 hour weeks is 8 weeks. That is to say, after 8 weeks of 60 hour weeks (and this is for manual assembly labor, not creative work), the team working 60 hour weeks has produced less total than the team working 40 hour weeks.
Henry Ford was not a humanitarian. The 40 hour work week has a century of research behind it at this point.
Reference Article: http://www.igda.org/articles/erobinson_crunch.php
And I hear the boss is a master beater!
In large-scale projects, however, the final month or two typically is where all the disparate elements are finally coming together and you’re having people going over the whole system looking for leakage at the joints.
This is also where I can see crunch as having benefit. There are a myriad number of dependencies between everybody’s work, and as a result people need to work until their pieces function properly so they don’t hold back others. There are any number of situations where Joe’s team’s extra 4 hours of work results in not wasting Sally, Bob, and Sue’s teams combined 100 hours of sitting around waiting for Joe to get his shit together.
But that depends on your outlook. For a single guy, it’s bad but mostly it’s about your personal morale. For a married guy, it’s about eating dinner at work while your wife eats dinner alone in the kitchen. That gets old very quickly, much more quickly than a month (I have no idea about your marriage, EpicBoy, but I got married because I like being with my wife). For a married guy with children, it’s soul crushing after a week. I don’t like seeing my kid for an hour in the morning, and that’s it. I want to play with him, help with homework, talk about our days, etc. - all the fun and essential evening activities my dad did with me.
It’s not just bad management at that point to ask for crunch - it’s immoral and unethical. And the reason for that is that crunch is unnecessary. There is absolutely no reason to crunch on any project - ANY project - other than mismanagement. There are always features that can be cut. There are always art passes that can be foregone. The trouble is that most creative directors view every title as a chance to make a triple-A game, and cannot possibly comprehend the concept of cutting one of their brilliant ideas. Which is fine, that’s their job. The job of the producer is to reign that in by constructing a reasonable schedule and forcing the creative director to prioritize correctly, and to keep re-visiting of assets and features to a minimum. Of course, people will read this and assume that the quality of games will decline. Some will. But others will be just fine, as designers, coders, and artists start to understand the necessity for planning and execution without the luxury of the “well, let’s spend a few months developing this feature and then throw it in the game and decide if it works” mentality that I’ve seen on a multitude of projects, in a multitude of studios.
I’d also mention that not all titles are going to be AAA. Scheduling and budgetary restraints make that a reality. But even on those titles, I’ve heard of unreasonable crunch as the team felt “motivated” to try and get a metacritic score in the 90s. That’s ridiculous. It’s not going to happen, and meanwhile you’re burning out the team, straining families, and fostering even further resentment between management and employees.
And I’d also like to address EpicBoy’s assertion that partial crunch (ie. some people working later than others) causes morale problems: too bad. If you feel you need to stay late to get your work done, or improve it to some quality level you’ve decided is necessary, more power to you. My priority is working hard during the day and getting home at a reasonable hour. I’m in this industry for the long haul, so killing myself for a couple of years to ship some game that no one will be playing six months from ship date isn’t going to be a priority for me.
Thats my experience as well from working as a developer for the last 15 years. And thats just productivity, not to mention the impact on someones life.
Yes I’ve crunched alot in my life, and the “you have to work 12h/day for the next 12 months” was one major reason why I left Funcom.
In our experience the benefit of crunch has more to do with the excitement and attitude of the team and less to do with the total volume of work hours. I tested this theory on our last project but cutting work hours and enforcing these cuts. The excitement/attitude felt the same as a crunch–we wanted to wrap things up and cram in as much as we could–but we went home at 3:30. And it worked pretty well!
Crunch also changes worker mindset from blue sky to limited resources. If you feel you only have so much time left on a project you begin to optimize your decisions. Maybe feature X requiring Y hours of production isn’t worth it, or 90% of X can be accomplished with 10% of Y.
We’re an indie data point, so maybe this changes things (we’re six people, so it’s not like we’re a single-man operation or anything).
We did this on Blush, which was an 8-week game:
http://vimeo.com/3427308 (trailer)
http://blurst.com/blush/ (play)
The weekly project breakdown was:
- Week 1, prototyping (2 people), 10-6 M-Thu
- Week 2, prototyping (2 people), 10-6 M-Thu
- Week 3, production (6 people), 10-6 M-Thu
- Week 4, production (6 people), 10-6 M-Thu (I took this week off)
- Week 5, production (6 people), 10-6 M-Thu
[B]- Week 6, production (6 people), 10-3:30 M-Thu - Week 7, production (6 people), 10-3:30 M-Thu[/B]
- Week 8, production (6 people), 10-6 M-Sat
Fridays have always been experimental days (random tests or hobby projects).
Since then we’ve permanently slashed our work hours. It’s hard to quantify exact numbers, but we aren’t losing nearly as much output as we’re gaining in personal time. The nonlinear gain of crunch goes in the other direction, too. I don’t expect we’ll raise our hours to full strength (8 hours M-Thu) anytime soon, even with our 8-week production plan for 2009.
If we can make this work other companies can, too. Yeah, we make smaller games, but we feel it’s still a big feat to pull off any of our projects in 8 weeks.
-Matthew
So, let’s say that you get twice as much done during crunch time as you do otherwise. (You don’t, of course.) You’re saying that you could completely eliminate crunch by extending your development schedules by a month or three?
The most interesting thing about that chart: Week 1 is a 40 hour week. Weeks 2-5 are 60 hour. The most productive 60 hour week was only about 30% more productive, despite being 50% longer.
So, let’s say that you get twice as much done during crunch time as you do otherwise. (You don’t, of course.) You’re saying that you could completely eliminate crunch by extending your development schedules by a month or three?
You often can’t extend the schedule. Your ship date is locked down and you have to hit it or you miss black friday/xmas. That’s something only games like GTA or Metal Gear Solid can do and still survive. Marketing is usually rolling months in advance of you shipping and if you miss the date, it’s pretty much wasted effort.
But as far as productivity goes - at the end, you’re fixing bugs. Reams and reams of stupid little bugs that are clogging up the database. Cover is broken here, coverslip isn’t available at this location, some z-fighting there, this mesh looks like it’s being lit weird, this mesh has collision and shouldn’t, there’s a little gap between these 2 meshes, this view could use some optimizing, etc. This is the stuff that can be fixed when you’re tired with little consequence. That’s why the crunch at the end is about polish and getting the game as solid as possible - it is NOT about adding features. Nobody does anything especially creative during this period … and you don’t want them to. It’s about stopping the jello from shaking and about getting the game ready to hit shelves.
Yes I’ve crunched alot in my life, and the “you have to work 12h/day for the next 12 months” was one major reason why I left Funcom.
That’s the kind of crunch that I’m happy to agree with you guys on. That’s stupid and unproductive. That will kill peoples motivation and cause them to quit.
A few months at the end of the project? That’s expected - it’s time to SHIP.
The most interesting thing about that chart: Week 1 is a 40 hour week. Weeks 2-5 are 60 hour. The most productive 60 hour week was only about 30% more productive, despite being 50% longer.
When people crunch for extended periods, a LOT of time gets eaten up with talking in the halls and surfing the web. They make their 8 hours of work extend to fill 12 hours but not really getting any more done. That’s why it’s important to manage the length of crunch. A motivated team CAN be more productive in short bursts.
This is also where I can see crunch as having benefit. There are a myriad number of dependencies between everybody’s work, and as a result people need to work until their pieces function properly so they don’t hold back others. There are any number of situations where Joe’s team’s extra 4 hours of work results in not wasting Sally, Bob, and Sue’s teams combined 100 hours of sitting around waiting for Joe to get his shit together.
Yep. It’s a period where communication is key and if you need to ask someone a question about something it’s awesome to have them there rather than have to wait until the next day.
At any rate, thanks for the responses, but I think I’m done responding in this thread. I can see idiots like NWJ starting up with insults and insinuations and it’s only going to deteriorate from here. I am not here to defend crunch. I’ve said many times that I don’t like it but I DO view it as necessary thing to ship a top shelf game.
Again, this it’s sustainable for a month or two. Beyond that, and it’s stupid.
I don’t think I’m being cynical there. I’ve worked both ends of this in my time, both dev and PM, so I’ve seen (and in few cases helped push for) crunch. I’ve never seen any crunch have ANY benefits beyond letting a team do more with less resources and adapt to deadlines in a fluid situation. From a management perspective that’s a big benefit, to be sure, but it’s purely a resource benefit.
There are valid business cases for crunch, and I won’t deny that. As you’ve noted, some employers make it worth the while of the devs to put up with work conditions that call for some crunching from time to time. But please don’t try to sell me on some rah-rah management BS that crunch is how teams achieve some creative epiphany or sharp technical focus that couldn’t have been accomplished within regular office hours if the project was more adequately resourced.
Ding ding ding, we have a winner! That’s the business case for crunch right there. That’s where management has to realize that’s is way cheaper to have Joe’s team work OT than it is to idle a bunch of other teams. And that’s okay if the compensation structure and employee expectations are built to handle the idea that crunch will be called for.
In my company crunch is expected, we know it’s coming, and when, and typically for how long, and also why. We also are paid overtime, and typically are encouraged to take vacation for a week or two afterwards.
I will say that I’ve worked here for 3.5 years now, and for the first three years I was single and this last half year I’ve been married. A couple years back I and several others basically destroyed ourselves working to make up for some poor management decisions made on the project I was on at the time. I put up with it then, but wouldn’t now.
Hell, I’m technically crunching right now - I’m just waiting for our DBA to get a damn index put on a table so I can get the rest of my shit done tonight.
The reality of crunch time: Most of it is spent surfing the net.
But please don’t try to sell me on some rah-rah management BS that crunch is how teams achieve some creative epiphany or sharp technical focus that couldn’t have been accomplished within regular office hours if the project was more adequately resourced.
I never attempted to do that.
The reality of crunch time: Most of it is spent surfing the net.
Long term crunch, yes. If that’s happening with short term crunch, you’re working with people who don’t give a shit.
I typically work 40 to 45 hours a week during most of the year but during a major crunch time I’ll work about twice that.
My productivity scales until around 60 hours a week. After that, it gets into diminishing returns. But often times, it’s not about pumping out code but rather simply increasing the communication within teams that requires so many hours.
If you’d engaged your DBAs earlier in the process for some collaborative data modeling and exploration of your usage of the data model, that index would already be there . . .
Yeah, I have strong DBA roots :)