Sign in to follow this  

Grand Theft Splinter Cell

This topic is 4530 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Quite a strange title, but hopefully the reason for it will become apparent. GTA's success and addictivity (is that even a word?) comes from it's sandbox/toolkit design. Heres a world, heres some rules that govern the world, go and do this... without restricting the method of accomplishing a certain task (too much), the player is free to attempt the task in any way they see fit. "I have to go and ice this heavily guarded guy, should i run him over, snipe him, go in all guns blazing?" As the Neo eloquently put it "the problem is choice". Although in this case it's not so much a problem as an opportunity. The downside to this is that the player could fail through no fault of their own. A car could pull out just as your lining up to squish the guy etc. but i'm sure your all familiar with that. ;) so, onto splinter cell. wouldnt it be better to employ the same design ethic as GTA? Heres a building, just there. There are certain ways in, which can be selected by the player. HALO parachute onto the roof, climb through a window, pick the front door, stealth kill the guards and turn of the survailance etc etc. From there you have to make you way to floor X. you could take the lift, climb up the lift shaft, scale the outer walls of the building etc etc. I think you know where i'm going with this. Would that be a improvement? Does a game like SC need a structured route and scenarios to make it enjoyable? does any game? Is this (as another thread put it) "the real future of gaming"?

Share this post


Link to post
Share on other sites
The main problem is that you, as the designer, need to think of every conceivable method the player could use to accomplish the goal. GTA characters were pretty much drones. They didn't react specifically to anything you did. In a game where the ideas have been already planned out for you, the dev team is able to provide very specific decisions and interactions from the NPCs.

It's not always a problem. If you're just wanting to break into an area and steal something, then generic AI will do pretty well. But it prevents the designer from giving a unique experience to that one job. It will end up being almost exactly like some other mission to break in and blow something up, except the building is different, guards are in different places, and other little details. If the player has specific routes, the designer is free to add any amount of dialog and details specific to the choices the player makes.

I like the idea though. I prefer generic AI over a set course. When I get bored with games, one of my favorite things to do is to mess the game world up by trying to do things that I'm not supposed to be doing.

Share this post


Link to post
Share on other sites
I'm actually 1.7 years into development of my game "Gang War" ( Website ) which is a lot like what you're saying. I'm going to enter it into the Independent Games Festival this year...and I've signed a publishing deal for worldwide distribution after that.

You start in a randomly generated city and you need to reach one of several "victory" conditions in order to win (kill all other gangs, take a certain % of the city, etc.) You play against up to 6 other gangs. The real cool part about this game is the fact that you can give orders to your gangsters like in any RTS game...but you can take control of any of your gang members at any time, and complete missions from the henchman's point of view...using any means necessary...but having to deal with the real world consequences and latter strategic impacts of such actions....legal ramifications,death,etc.

I've modeled the interiors of every building in the city, and it's completly open ended.

I've got a lot of great ideas for the game...if you want more information/screens check out my developers journal here at GameDev ( Developers Journal )

- Dan

Share this post


Link to post
Share on other sites
So basically Thief, which provided a relatively open path to victory in each level (the levels where designed more like structures, rather then a linear set of corridors and rooms to be passed in sequence).

(for reference I played Thief before Splinter Cell, resulting in Splinter Cell initially feeling almost like a slightly more interactive copy of the old FMV games)

Share this post


Link to post
Share on other sites
Yep I am with Jiia on this. It would be great but it would be very hard to do well. Good AI is hard enough to do when you know exactly which route the player will have to come. Give the player all that freedom and it starts to fall apart.

Share this post


Link to post
Share on other sites
Very good point Jiia.

Although the AI in GTA is very drone like, it does display some semblence of thought. And many games go far beyond rudimentary behaviour of GTA's NPC's (RE4, Half Life etc). From what I've seen of Splinter Cell games, the NPC behaviour is not exactly the pinnacle of AI. (admittedly, ive not played much splinter cell, but it struck me as a good example of how to apply sandboxes to other genres)

I guess what my mine gripe about the design of certain games is the linearity of them. Escpecially when other games have shown the benefits of a more open game style. Even if at certain points the gameplay is bottlenecked into certain set pieces, the majority of the world could still be very much a sandbox design.

It may take more time and effort in terms of design, testing and coding, but we should be moving away from the "do this thing this way, otherwise you will fail/die" design.

Quote:

one of my favorite things to do is to mess the game world up by trying to do things that I'm not supposed to be doing.


ever thought about being a bug tester? ;) (assuming your not one already!)

EDIT: A few people posted before i completed mine so i'll adda little here.

dgreen02: That sounds like an awesome idea. I'll be sure to check your site.

Michalson: having never played Theif, i now feel quite envious. but if Theif can do it, why not others too? The idea i get from SC is that the designers want to take you through their lovely little game they've made. "come, look at these set peices I've made for you to do!"
"Well, actually, i want to see whats in this room/window..."
"Hell no! stick to the path!"

Share this post


Link to post
Share on other sites
Just for the record, I wasn't using Splinter Cell as a reference. I actually haven't had a chance to play it yet. I've just noticed that stale plots and engine-driven AI usually come free with a sandbox environment. Giving the player real freedom of choice, as in the player is able to do things that even the designer didn't realize, means sacrificing some elements. But I think it's worth it. I much prefer that style of game.

Share this post


Link to post
Share on other sites
Quote:
Original post by Multiverse
but if Theif can do it, why not others too? The idea i get from SC is that the designers want to take you through their lovely little game they've made. "come, look at these set peices I've made for you to do!"
"Well, actually, i want to see whats in this room/window..."
"Hell no! stick to the path!"


I'm actually psyched to see more and more people open to the idea of open ended gameplay environments. But the tension is always going to come from those who like structured play and those who like unstructured play. Structured usually goes with story, and mission-based story games rule the roost at the moment.


Share this post


Link to post
Share on other sites
Video games are like children: They need some structure, but give them too much and you make them really boring.

Remember that old platformer you used to play? You know, where you had to jump at Juuuust the right time, and then jump over a barrel and immediately duck under a rolling rock? Perfect example of going too far with structure.
On the other hand, who wants a game that lets you drive your car to the store, mix your own explosives with common household cleaners, and then go completely demolish the level you were supposed to sneak through? Well, maybe you would like that, but it would get boring after you blew up your third or fourth building.

Open-ended gameplay requires just enough structure to force you to do something, but not enough to make you feel like you have to do x right after y before you shoot z. See the platformer example above.
Even GTA encourages you to go on missions, and provides cars, guns, and explosives. If it was realistic, you would get your ass kicked by rednecks when you tried to pull them from their cars, the cops would come as soon as you jacked somebody, and you would be killed or arrested for at least a few years.
Maybe it's a good thing games aren't always completely realistically open-ended?

Share this post


Link to post
Share on other sites
Quote:
Original post by Meagermanx
If it was realistic, you would get your ass kicked by rednecks when you tried to pull them from their cars, the cops would come as soon as you jacked somebody, and you would be killed or arrested for at least a few years.
Maybe it's a good thing games aren't always completely realistically open-ended?

I'm still waiting for that GTA release. My most beloved moments are while battling it out with the police. When I go back and play the series now, I never go on a mission without at least four stars to spice things up. I don't think you could add enough realism to GTA. In my eyes, it would just keep getting better.

Share this post


Link to post
Share on other sites
Consider Deus Ex (the original, the sequel didn't do this as well). It was basically a set of missions connected in a linear fashion. However, there were generally a lot of different ways to complete each mission, and even some cross mission results based on what you did (for example, if you keep Paul alive). You could use stealth kills, hack computers, raid storage rooms from lock picks (a little weird, but fun), or just go in shooting. Lots of variety, but at the same time there was plenty of story (it was quite a good story) between and in the missions, at points where you ended up regardless of how you got there.

tj963

Share this post


Link to post
Share on other sites
I don't think players (read: I) would be too delighted by a totally open-ended stealth-action game like you describe. Wavinator ran into this problem in his design a few times, and so maybe he can shed more light than I can, but I really think that in order for a game like Splinter Cell to work, you need to have levels with reasonably clear puzzles built into them.

When I come into a situation in Splinter Cell, I know that the guards will all have patrol routes that they stick to, usually spanning only a few yards, and always stopping just short of seeing around some corner. That's required. If they peeked into likely hiding spots, the game would be ridiculously hard, and you'd wind up shooting your way through it. Let's face it: With semi-intellingent (or even semi-random!) guard behavior, there's no way anybody would get through that game without some serious firepower. Why do they never look up? Why do they give up on shattering glass when the bottle itself isn't a ninja? Why do they explore solo rather than call in backup, like real security professionals would? The answer to all of these questions is the same: It's a spy game, and those guys are spy food.

If you build a genuinely open-ended world, you'd have to populate each building and dockyard and stadium with genuinely scripted spy food in order for your spy players to get through the game without starving for fun. Really, what are the odds you'll stand any chance of actually creeping through a bona fide high-security facility without tripping an alarm or getting accosted by MPs? The only way to make it possible is to make it semi-realistic--that is, two security guards for the whole mall, one of which is asleep and the other of which is peeing. That's insanely boring.

"Fun" sneaking games are characterized by huge populations of ragingly incompetent "guards". Why are sixteen dudes patrolling the sewer, ine every six yards? Why is there a guy on the roof, staring out into space and walking over to "get a cup of coffee" every 11.7 seconds?

Spy games need to be high-tension obstacle courses. GTA is a high-tolerance sand box.

A compromise may be possible, but a pure fusion can't work.

Share this post


Link to post
Share on other sites
Quote:
Original post by Meagermanx
On the other hand, who wants a game that lets you drive your car to the store, mix your own explosives with common household cleaners, and then go completely demolish the level you were supposed to sneak through? Well, maybe you would like that, but it would get boring after you blew up your third or fourth building.


I'm being maybe a bit petty, but with a point: Freeform action without consequence isn't what I'd define as open-ended. Open-ended gets a terrible rap because of the hordes of people who either get lost or don't pick their own goals. But that's I think the fault of the design, not the player. The player needs to know that the world reactions and escalates. Open ended gameplay isn't letting you blow up a building four or five times because you feel like it, that will lead to boredom. Rather, it's seeding the world with goals, obstacles, and changes to the basic rules that force you to keep updating your strategy.

Quote:

If it was realistic, you would get your ass kicked by rednecks when you tried to pull them from their cars, the cops would come as soon as you jacked somebody, and you would be killed or arrested for at least a few years.
Maybe it's a good thing games aren't always completely realistically open-ended?


But why did you raise realism? It has nothing to do with open-endedness.

Share this post


Link to post
Share on other sites
Quote:
Original post by Iron Chef Carnage
Wavinator ran into this problem in his design a few times, and so maybe he can shed more light than I can, but I really think that in order for a game like Splinter Cell to work, you need to have levels with reasonably clear puzzles built into them.


Well, I suffer from only having watched Splinter Cell being played, most of my experience was with Thief. But I think you nailed the main challenge of this kind of design, in that there's going to be a certain encounter frequency needed to make the game fun. In a GTA-style game, you choose the encounter frequency just by walking up to someone. In a Splinter Cell, the guards and threats choose it.

So this makes me wonder: How much can you change Splinter Cell before it no longer has whatever vital essence that draws people to it? What if it was easier to know where people are? What if there were more options for diversion and distraction? Not being a player, I can only guess how these might change the game.

Quote:

When I come into a situation in Splinter Cell, I know that the guards will all have patrol routes that they stick to, usually spanning only a few yards, and always stopping just short of seeing around some corner. That's required.


Patterning. Right. But I wonder if there are gameplay ways of giving you better tools and non-lethal attacks? Again, I don't know if the only fun of that game can come solely from picking off targets, or if maybe part of the fun comes from manipulating and navigating an environment that has lots of options (rappelling, going through ducts, causing problems as distractions).

I always think the secret of open ended gameplay is to expose, through story and gameplay, how the systems work, and then give you varying tools that let you experiment with tweaking that system.

Quote:

If you build a genuinely open-ended world, you'd have to populate each building and dockyard and stadium with genuinely scripted spy food in order for your spy players to get through the game without starving for fun.


Great observation. I have Project Eden on the brain right now, so I'm wondering how much of the threats you face are people and how much are automated. I don't know how much fun clipping cables and splicing in footage of a perfectly empty corridor would be in a game like Splinter Cell, but it would be interesting to consider better tools for scouting and stronger static defenses as one way of providing the meat.

After all, how is it that our heros in the movies do it so well? What "gameplay device" might they have that would raise the player's situational awareness?

Share this post


Link to post
Share on other sites
i think tj963 explained the idea which i named 'bottlenecking' a while back. Totally free form play would be quite daunting with anything other than simplistic goals in a relatively feature rich tool kit design. It would provide so much choice that the players would feel lossed, and wouldn't receive any feedback that what they are doing is correct, or, helping them acheive their goal.

Quote:

Lots of variety, but at the same time there was plenty of story between and in the missions, at points where you ended up regardless of how you got there.


This is basically what i meant. a 'level' (to use the term loosely) could be subdivided into smaller goals. They often already are, but most of the time their not essential other than to instill the feeling that you are doing something highly complicated, where in fact you are doing the same thing over and over again, in the same way, using the same tools/techniques.

For example, Objective #1 might be to disable the CCTV cameras. To do that, you make your way to the CCTV control room and disable them. From the control room, you have to do activity X and so on. Trying to fool the player into thinking this is a non-linear path. Even though it is just a linear path with diversions. What it boils down to is the same as "find the key, open the locked door", but with a different subject.

I would much prefer it if you had an option. Either disable to CCTV cameras, find another route, go the route with CCTV cameras but stay undetected. After all, how come some missions you 'have' to go and disable the cameras, but others there is no option to do that at all?

So, i think we have established that totally free form games would be more like applications or technical demos than games. Games must have some structured elements to them. Rules and victory/failure conditions etc. But, would it be a step in the right direction to introduce more sandbox design into games, or will certain games/genres forever be the structured movie like experiences?

EDIT: Just another example. Football (American or Soccer and indeed, many games derived from real life sports) games are basically sandbox designs through their very nature. Heres the rules, heres the actions you can perform, go and beat an opponent. To use an analogy, a structured soccer game would force you to pass to the winger, then run forward with him, cross the ball to the striker, the striker scores. Doing it any other way would not only lead to failure, but just isnt allowed.

Will some genres only ever be fun if they are structured?

Share this post


Link to post
Share on other sites

This topic is 4530 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this