As a player, would you rather play as a super-human one-man army, saving humanity with one hand packing a bazooka-pistol while cradling a baby child with the other OR a faceless, nameless soldier, in a long line of 1000s of soldiers, with no will, and no real influence on anything that happens in the in-game universe? More importantly, after realizing that both of those options are gimmicky at best, where would you prefer to be placed on the continuum in between? Above all, why did the majority of game developers decide that players all want to experience the weight of the universe on their shoulders, every time they pick up a game controller? Halo, COD, I'm looking at you. What are your thoughts?

There's an old shooter called Operation Flashpoint where you got to experience both of these extremes (and the in-betweens).

You'd start off as a regular grunt, with your unscoped M16 that couldn't reliably hit anything past 100 yards, and you'd have to rely on the other dozen guys with you to fight effectively. In the later missions, you were a one man special forces bad-ass who could assasinate generals and take out whole bases filled with tanks.

While some of the solo missions were memorable, I really did love being a low-ranking soldier in part of a huge squad. Not being able to tell your squad-mates what to do, being told which direction to cover, etc, really made you feel like part of a team, not just like the only "real person" on the field. Having specialised responsibilities was cool too - as a machine gunner your suppressive fire could influence how well your squad could move up, as the RPG-guy you reaction times could determine how many of your squad-mates will be killed by a tank, etc... So being "just another brick" can still be meaningful when cooperation is essential to success.

There's also a lot of middle ground, like when you're the squad leader. As well as being in control of decision making, you also might get better equipment like a scope, but you still can't be a one-man-army - you've got to make use of your specialists to survive.

[Edited by - Hodgman on May 7, 2010 2:50:12 AM]

Quote:

As a player, would you rather play as a super-human one-man army, saving humanity with one hand packing a bazooka-pistol while cradling a baby child with the other OR a faceless, nameless soldier, in a long line of 1000s of soldiers, with no will, and no real influence on anything that happens in the in-game universe? More importantly, after realizing that both of those options are gimmicky at best, where would you prefer to be placed on the continuum in between?

Everyone wants to be a super hero. To keep it from feeling like a gimmic you can do two things. First, make sure the player knows from the start that you are superman. Look at prototype or infamous(didn't play the second), you are superman, and you become more and more super powered. The other thing is to do the Half Life or Doom 3 type progression. Both heros are just in the right place at the right time. They progress from crowbars and pistols, while running from zombies to godly weapons killing huge alien deamons. Either way they both end up being that "super-human one-man army", but the super-human aspect grows over the course of the game.

Quote:

Above all, why did the majority of game developers decide that players all want to experience the weight of the universe on their shoulders, every time they pick up a game controller? Halo, COD, I'm looking at you.

Partly it is to give purpose to the game, and partly due to limits on the gameplay.

The purpose comes from the fact that you can tell a "super-soldier" to do anything. A "generic soldier" on the other hand feels out of character doing many of the things a "super-soldier" would do. Part of why I disliked COD2 is the ally bots never felt like they helped enough, and so I was doing all the work. You can also put a "super-soldier" anywhere you want. A "generic soldier" on the other hand stops feeling "generic" in many situations. Consider the proportion of "attack" levels in Halo, Half-Life and Doom 3 compared to the proportion of "defend the objective" maps in the COD series. COD may be putting "the weight of the universe on your shoulders" but they do a really good job at seperating the elite from the generic (ie "soap" in COD:MW1/2 is elite, but "private johnson" is generic)

Considering that you are playing a game, many FPSs in particular will focus on you. There is only so much CPU power to spare for the AI bots, and the game won't be fun without people to shoot at. So many FPSs like Halo, Half-Life, and Resistance make you "superman" just so most the time you don't have AI buddies wasting all the CPU. Other games like Swat, Rainbow Six, or Army Of Two, limit you to a small squad. And games like COD have tonnes of AI, but they are all "turrets" who just stand and shoot, and their squad is almost always bigger than your squad. It is hard to be "just another soldier" without having lots of other equally good soldiers with you, so technology wise the superman is easier to manage.


Something like Swat 4 splits a nice balance. You are part of "SWAT" so that provides a backdrop for why you are in such strange situations. You are the team lead, so maybe better, but not "superman". You have 4 other members who you actually need for most the missions, so you look like a "generic SWAT member" since your 4 AI attempt to be your equal.

I don't need to be either.

I'm tired of always being the guy that gets ran around and does everything.
I'd like to be able to pick a role, like turret gunner or pilot, foot soldier or special forces.

Not to specialise, but just to pick an aspect that changes the way I play the game each time.

I'd like to be the main character of an interactive sff novel.

They make the player the focus of the game so that the pacing is entirely under his control. The alternative is to shift increasing amounts of it into the hands of the AI, and that's usually a bad idea.

The compromise between the two extremes, anyways, is to be important but unable to do everything yourself. If the player was the leader of a team, for example. It takes more finesse and effort, and players might not even notice/care, so developers may not see the incentive to care themselves.

Quote:

Original post by AethonicThe alternative is to shift increasing amounts of [control over pacing] into the hands of the AI, and that's usually a bad idea.

Why is it usually a bad idea? From my perspective, gaming struggles as a storytelling device because so much responsibility is put on the player. There's a reason why Tolkien, Stephen King, and Shakespeare didn't write choose-your-own-adventure novels.

Quote:

Original post by sunandshadow I'd like to be the main character of an interactive sff novel.

You mean in a game? Don't really understand what you're trying to say...

Quote:

Original post by KulSeran Partly it is to give purpose to the game, and partly due to limits on the gameplay. The purpose comes from the fact that you can tell a "super-soldier" to do anything. A "generic soldier" on the other hand feels out of character doing many of the things a "super-soldier" would do. Part of why I disliked COD2 is the ally bots never felt like they helped enough, and so I was doing all the work. You can also put a "super-soldier" anywhere you want. A "generic soldier" on the other hand stops feeling "generic" in many situations. Consider the proportion of "attack" levels in Halo, Half-Life and Doom 3 compared to the proportion of "defend the objective" maps in the COD series. COD may be putting "the weight of the universe on your shoulders" but they do a really good job at seperating the elite from the generic (ie "soap" in COD:MW1/2 is elite, but "private johnson" is generic)

You can tell a super-soldier to do anything but not a generic soldier? I understand that a generic soldier can't necessarily fight off an alien armada alone. But as one soldier in an army of generic soldiers, that could be done, and would create an interesting and unique gameplay experience.

I guess the problem might be that developers want to put the player in a situation where if he doesn't do well, the mission WILL fail. Perhaps that needs to changed. Gauge player success on a smooth scale, and have a narrow range of endings to each mission, depending on how well the player performed? In a real scenario, if 100 soldiers are sent into a battle, how much difference is one soldier's success or failure going to make?

Quote:

Original post by sunandshadowConsidering that you are playing a game, many FPSs in particular will focus on you. There is only so much CPU power to spare for the AI bots, and the game won't be fun without people to shoot at. So many FPSs like Halo, Half-Life, and Resistance make you "superman" just so most the time you don't have AI buddies wasting all the CPU. Other games like Swat, Rainbow Six, or Army Of Two, limit you to a small squad. And games like COD have tonnes of AI, but they are all "turrets" who just stand and shoot, and their squad is almost always bigger than your squad. It is hard to be "just another soldier" without having lots of other equally good soldiers with you, so technology wise the superman is easier to manage.

This brings up an interesting question. Is it better to spend CPU on AI or graphics? Which contributes more significantly to immersion? What if we regressed to Halo 1 graphics, but advanced AI so that the player doesn't feel alone?

Here's a question for everyone: how can we make players feel like part of a large team or army in a game?

Quote:

Original post by GiantGames Quote: Original post by sunandshadow I'd like to be the main character of an interactive sff novel. You mean in a game? Don't really understand you...

Yes, I mean a good game should be like an interactive novel and I should have the main character's role within it. A Zelda game is an example - the player in a Zelda game takes the role of Link, the main character of a fantasy story. But, the Zelda games are not very interactive. A dating sim is another example - in that kind of game the player takes the role of the male main character of a romantic or erotic story. It is interactive as far as the other characters go, but tends to not have combat or crafting or gathering, so the physical world is less interactive. Usually the world in a dating sim is not developed enough to have the quality of a good science fiction or fantasy novel, while RPGs may have more deeply developed worlds.

When I say an interactive sff novel it is similar to saying a choose your own adventure novel or I've also heard people use the phrase interactive movie. BTW the reason CYOA stories have not been more popular is that they take immersion-breaking extra effort on the player's part to read (not true of a videogame with an interactive story). Also CYOA games are inefficient to produce because the consumer does not experience 100% of what is produced. This part is the major reason interactive stories are not common in videogames.

Quote:

You can tell a super-soldier to do anything but not a generic soldier?

Quote:

Here's a question for everyone: how can we make players feel like part of a large team or army in a game?

I'm getting more at the fact that random soldier "Vasquez" isn't going to be tasked with specifics that "Bond" would.
"Bond" has everything. The explosives, the laser watch, the decoder, the communications, the silenced gun, can fly the plane, can drive the car, and gets the girl.
"Vasquez" has a gun. The coms man has the radio. Demo man has the explosives. Driver drives the transport. Pilot flys the chopper. Vasquez doesn't get the girl, cause they killed him when they set off the nuke (modern warfare 1).

When you want to make the player feel part of a team or army, the game can no longer be about "player do this!", it becomes "we have to do this!". You can't have the player be as central or otherwise more important that the other characters.

* They need teammates to cover. (escorts, team has to survive, clearing a path)
* Team mates to cover them. (player doesn't feel like everyone is shooting at them. Distractions).
* Other people who carry in equipment. (player doesn't have C4, someone hands it to him. Player has to go use the fallen comms man's radio)
* Other equipment the player isn't qualified to use, but relies on a team mate to use. (Natalia has to hack Goldeneye. Someone else flys the chopper or drives the transport)
* More equipment than one person can carry is needed to finish a mission. (SWAT4 equipment, L4D supplies)
* Actions that need team mates. (L4D "i'm down" mechanic. Army of Two mechanics)
* Lack of omniscient information. (No maps with enemy indicators. Maybe no maps or crude maps. Compass in place of MiniMap. Other characters provide information al-la COD4's "Over there! Behind the red car!" type hints, and L4D's "I hear a boomer".)

Quote:

This brings up an interesting question. Is it better to spend CPU on AI or graphics? Which contributes more significantly to immersion?

In game's I've worked on.
Upwards of 50% is graphics overhead.
Player code takes a significant chunk of time.
Animation has huge overhead (5-30%).
Physics has huge overhead (10-40%). This includes collision tests for everything related to the HUD, Player, and AI, not just the rigid body simulation.

AI tends to take a back seat in performance. If you run at 30FPS with 10 AI, then good. If you only run at 20FPS, then you cut it back to 5 guys and see if you hit 30.

At 30FPS, the eye is REALLY good at seeing things like 1 frame animation "pop" in a blend, and 1 frame animation "de-sync" between two interacting animations.
The eye is really good at detecting particles and geometry pop due to z-sort and transparency issues.
The eye isn't so good at determining if the scene is wrong though. As long as it is consistent. It's been shown that many "fake" lighting tricks for reflections and refractions on non-smooth surfaces still look "right".
You can break immersion with the smallest of graphical bugs due to how badly they pop out as broken.

Just like the graphics, your brain sees what is wrong with the AI, but ignores what is right. The L4D allies are very good AI. Everyone I know (me too!) yells at them when they do something that I internalize as stupid. The problem is that I can't internalize the AI's thought process, and so place my own assessment of the situation on them making them seem dumb.
You can quickly break immersion if these "dumb" moments happen too often, are too glaring, or screw the player over too badly.
On the other hand, little tricks like what FEAR did can help make AI seem very smart. FEAR's AI made decisions, then voiced their thought process. The AI would say things like "He's behind the desk, flush him out!", "Covering fire!", "We need backup", "grenade!". The AI might yell "grenade!" then sit there and get blown up. I still think their being stupid, but the way their conversations explain away their decision process fools me into thinking they are way smarter than they actually are. Their bad choices were due to my good planning.

Quote:

What if we regressed to Halo 1 graphics, but advanced AI so that the player doesn't feel alone?

May not have to. The issue isn't all the people and things around. But what part the player plays in relation to them. If I'm playing in a battle with 100 bots, but I'm still the one who has to move forward to hit some trigger to cause the next scripted event to happen (COD2, i hate you), then I'm not going to care about all the useless bots. If I'm playing L4D and a bot goes down, I'm still likely to run back save it, because everyone is part of the team. It only took 3 bots and I to make a team of 4 that I fell involved with due to the interactions.

[Edited by - KulSeran on May 9, 2010 4:57:02 PM]

Quote:

Original post by GiantGames Quote: Original post by AethonicThe alternative is to shift increasing amounts of [control over pacing] into the hands of the AI, and that's usually a bad idea. Why is it usually a bad idea? From my perspective, gaming struggles as a storytelling device because so much responsibility is put on the player. There's a reason why Tolkien, Stephen King, and Shakespeare didn't write choose-your-own-adventure novels.

It's a bad thing because if the AI is too good then the player becomes bored, and if the AI is bad then the player becomes frustrated. If you balance it to the middle ground, you've still gained nothing with the approach. Rather than leave it up to the AI, where random chance could swing it in either direction, you should leave it up to the presentation layer (not mechanical), e.g. scripted sequences. Virtually every storyline is going to be held hostage by the game mechanics, pacing (for the game and story) has to be regulated, not left up to chance. There are exceptions for games with highly fluid stories, but those are very rare.