Thursday, January 27, 2005

Electronic Arts L.A. Layoffs

If you didn't hear this news yesterday, Electronic Arts laid off sixty programmers in their L.A. studio. Here's an excerpt from the Gamespot article:

EA hands out pink slips in LA
60 programmers, designers, and managers laid off at Electronic Arts' SoCal operation; studio boss Neil Young calls the move "regrettable, unfortunate, and very necessary"... According to Young, who spoke with GameSpot this afternoon, the layoffs were done in the interest of "rebalancing the teams across the entire studio." He called the cuts the "first step in transforming the studio for the future."


I'm not trying to pick on EA or Neil Young when I say this, but the phrase "transforming the studio for the future" is typical corporate doublespeak bullcrap. Studios don't get "transformed" when they're profitable--they get staff cuts because they were badly managed. Period.

Later in the day, I exchanged a series of e-mails with an employee of the EA L.A. division. This person, obviously, wishes to remain anonymous.

Here's the first e-mail I received:
I'd say at least 25% of the workforce was laid off. It is entirely because the management does not know how to utilize talent to make games, as proven by Medal of Honor: Rising Sun and Golden Eye 2: Rogue Agent. The studio wasted an amazing amount of resources on those sub-standard titles. This measure to cut costs is perhaps sensible, given that the executives don't know what to do with these people anyway.

Besides, if you spend all your money on exclusive licensing deals and advertisement, you don't have that much left to "waste" on development anyway.

The e-mail gave me an opportunity to ask a question I've always wanted to ask, Here is an excerpt of my response: "What do you think ruined those games and how early did you know that they were going to be sub-standard? I'm asking because I'm really curious as to how early in the development process something like that becomes readily apparent."

I received this response:
When it comes to the future quality of a product becoming readily apparent, it is of course different in every situation, but it is also dependant on the question, "apparent to whom?"

In the case of these two projects, it was apparent to the most experienced engineers from the very outset, due to the important engineering decisions being made by non-engineering personnel, which in the long run resulted in something like 50 thousand bugs logged, and not enough time to finish the game. Further lack of quality was driven by the lack of development skill in leadership, as demonstrated by:
- Multiple months of relative inactivity at the beginning of the project, while management meets with designers to dream up all the features which will later be cut for lack of time (presumably, designers get months of inactivity at the end of the project, since management is not yet ready to micromanage the next design phase).
- Demo-to-demo planning. ~75% of the development time (pre-finaling-phase) is spent preparing various demos. Demo hacks become ludicrous codebase during the next demo push.
- One year to make a two year game.
- "If we add online play, the hardware manufacturer will pay for advertising." Of course development time doesn’t change for this. After all we can put up a company from Texas and another from SCOTLAND in hotels across from the office for MONTHS to buy the additional man-hours we need (and extract the rest from the dev team for free!). Think of it. With no family or friends around, they have nothing to do after work or on the weekends, but work some more! I mean man-hours = product right?

And ultimately, it was clear that GE2:RA was doomed when, at the end of MoH:RS, the management didn’t say, "OK, we screwed that one up royally. Thanks for bailing us out this time.", but rather went around patting themselves on the back for the money it made (And by ‘bailing out’, I mean shipping a game, not shipping a great game. No one could have done that.).

I saw a tiny little GE2:RA postmortem that said things like "Online mode was a risk." How oblivious is that? If they want a clue, they can start with the reviews to find out what was wrong with the game, then maybe ask around about how it got that way.

Interestingly, many of the grievous management-as-engineers mistakes are being rendered impossible to make on the next go-around by the fact that they are being forced to use Renderware. But take my word for it, they’re doing their darndest.

Thanks to the anonymous source for a fascinating look inside the studio.


Site Meter