Rare Q&A With Rockstar Games Head Sam Houser

Paul Williams writes "Develop Magazine has posted a fascinating multi-part interview with Sam Houser, president and founder of Grand Theft Auto developer Rockstar Games. Houser is rarely quoted outside of press releases, and almost never does interviews. So, reading his frank views on things like Rockstar's critics, the creative secrets that make games like GTA IV a success, and how the developer rejects things like focus testing — a common practice at the likes of EA but an 'anathema to creativity' according to Houser — is very interesting. Houser has even written a mini biography of his career with some fun references to the Hot Coffee scandal: 'July 2005: Residue code found in San Andreas. Hackers modify it and it turns into scandal known as "Hot Coffee." Get dragged into legal nightmare, ending in trip to Washington in February 2006 to sit in front of federal trade commission staff — for nine hours.'"

Comments

Drop the script

[4D6963]

Of all things, storytelling is one of the areas that in some ways has the furthest to go.

Indeed. Please drop the whole "scripted storyline" concept and make a super fancy algorithm so that the story derives from whatever the player does and whatever happens as a consequence. It would be a tremendous paradigm shift, instead of have a linear story, or a branching storyline in which choices and such make the story branch out, every single of your action would influence more or less the future, and instead of cutting the story into missions in which you just do scripted missions until you succeed at each and every of them, let the player evolve with no safety net.

Therefore not only would you have to make complex decisions, but the way you execute these decisions would be crucial, and you wouldn't know if it was a good idea before you see it works and as long as you're there to see that there were no negative consequences.

Sure, that's a hell of a lot of R&D that would be involved, but that would kick a whole new level of arse!

Re:Drop the script

[fm6]

I agree that would certainly make for a better game. But do you have the slightest idea what you're asking? The software would have to creatively synthesis an outcome based on everything the player did. That's way beyond any "super fancy algorithm" available right now. You're basically talking about "strong" AI: software with the same level of creativity as a human.

That's an ironic suggestion, since the AI in GTA is remarkably poor. (At least in the version I played, GTA III for the PC; I suppose it might have improved in later games.) One example is the inability of NPCs to go from point A to point B if it involves any serious pathfinding. I was once in a parking lot, surrounded by cops, and none of them could get to me, because there was a low wall around the lot, and their travelling skills did not extend to "find the entrance" or "get out of your squad car and step over the wall."

Re:Drop the script

[MrMista_B]

God, I hope not. Who wants to play real life? What you're suggesting is the opposite of improving 'story' - it's the removal of story alltogether, leaving only player decisions and random 'whatever' prodecural generation.

Basically, what you're wanting is something like GTA, but without any story or direction whatsoever, just things like the cops coming after you if you get too crazy and other such things.

And, sure, that's fun, I don't really play GTA for the story anyway. But don't confuse 'sandbox game' with 'well thought out and developed storyline, characterization, and plot'. The two can live together, but removing one from the other doesn't neccessarily improve either.

Re:Drop the script

[Xtravar]

Please acquire "The Sims 2" and leave the rest of us alone who don't want to be annoyed as fuck when we find out we have to replay 9/10th of the game for not saving Billy from the Mafia in Act 1.

The only reason I play games anymore is for the storyline. Eventually, every game starts looking like the Real World and all quests look like Work. The only difference is that games HAVE A GOAL YOU CAN ATTAIN and work just drags on forever, sucking the life out of you until you die.

Re:Drop the script

[Fweeky]

Please drop the whole "scripted storyline" concept and make a super fancy algorithm so that the story derives from whatever the player does and whatever happens as a consequence

Here you go. Some assembly required. Dwarf Fortress is in many respects built to allow stories to emerge from gameplay; indeed, it's a significant part of what people find attractive about it.

It's kind of a mixture of Dungeon Keeper 2, Sim City, Nethack, The Sims, The Incredible Machine, and experimental brain surgery.

Re:Drop the script

[papabob]

This can be implemented in many ways, none of them requiring 'strong AI'. For example you make the game with a normal storyline (ascending to mafia boss and the like), and you only have to give the NPCs scripts more execution paths depending what missions you have completed before and whom did you talked to. You robbed the bank killing more than X civilians? Ok, you're a hard guy and the next NPC will give you the mission of collect the 'protection fees' from the bartenders. If instead you managed to win the drug deal without firing a single shot, well, you're perfect for this 'transport the wallet' mission. This way you've reduced a lot the "you can't continue the game untill you complete that mission" points.

Sure, it increases the work of the writers and the QA guys, but once you have the game engine polished adding new stories would be pretty easy.

Re:Drop the script

[mirshafie]

I think a completely dropped script would make for very boring gameplay. That wouldn't be storytelling at all.

Let me try to set up an analogy. I often have lucid dreams, I've had them since I were a little kid. But while there is in some ways a craze about lucid dreaming, with lots of people trying hard to achieve it nowadays, I find them boring.

Because non-lucid dreams tend to have a much richer storytelling, as if someone else dictates what will happen next and therefore I'm always taken by surprise.

The sort of complex free world that you're describing would of course be very interesting and extremely cool, but I doubt it would replace scripts. And as many other posters have pointed out, the technology needed for such a thing stretches far beyond what is available today. It's simply not within reach.

Different approaches

[4D6963]

We donâ(TM)t believe in focus testing ideas (itâ(TM)s like asking an audience what album they want to hear â" they donâ(TM)t know until they hear it!)

In art, the offer creates the demand more than the opposite. However if you're a marketing type and you want a sure shot, a product that you're sure people will want, you have to find out what the demand is.

Two different approaches really, an art-oriented risky but genre-defining approach, and a non-innovative but safe approach a couple of trains behind (i.e. the public follows the successful innovator, the EA-like company follows the public).

Re:So where have you failed the gaming community

[4D6963]

I never had the chance to play GTA IV yet but from hearing the things that are not present in the new game (i.e. airplanes, parachutes, jetpacks, etc..) and some user comments it seems that Rockstar has unfortunately stopped focusing primarily on 'fun'.

Yeah.

[Anonymous Coward]

And when I dream, I want a pony.

Re:They did a LOT of testing on Portal

[Z34107]

As others have pointed out, the Hammer editor (it comes with your Steam install) is used for making portal maps. Same tools the devs used, actually.

It's also used for making Team Fortress 2 maps, Half Life maps (and mods), Day of Defeat maps... You can pretty much make maps/mods/characters/weapons/etc for any game based on the source engine. For the most part, it's pretty intuitive, too.

Tossing softballs

[Akoman]

I hate how this interviewer softballs his questions:
"GTA IV asks the players to make a few key decisions during its story, and weâ(TM)ve seen another Take 2 game, BioShock, experiment with similar ideas. How further can that model be pushed? Is it something youâ(TM)d like to take further in future games?"

As if we haven't seen games providing two plot choices that affect game outcome in the past. Dark Messiah of Might and Magic jumps out as a very recent mediocre game that fits the bill there. The only thing exciting about Bioshock's decisions were that there were manifest benefits to selecting the immoral choice.

Railroads with junctions are not some new-fangled 'model' who's limits are waiting to be 'pushed.' It's old and stale.

Re:why 2 links

[tirerim]

Presumably because the link to the second page in TFA is in a non-obvious location -- it's only in the sidebar, not at the end (or beginning) of the actual article.