Advertisement

Formalizing Experimentation In Game? (RPG-like)

Started by January 13, 2006 08:47 PM
22 comments, last by Jotaf 19 years ago
I think it's a neat idea but I don't agree with having stat gains for it as others have suggested; such a system should be a gain in itself and its use shouldn't be an affect on the world outside the simulation, lest it turn into a way to cop-out by grinding up to max stats. Also, with this in place, I would expect the game to be difficult enough that most would be inclined to at least check out the simbox before rushing into combat.

Ender's Game was a super incredible book; I almost never reread books but I've gone throught it multiple times.
It only takes one mistake to wake up dead the next morning.
The usual use I find for quickloading is for high reward with low risk. E.g. gambling until I win heaps of money; winning a combat with minimum casualties.

Risk and reward should be balanced: A player should infrequently be able to gain a large reward with little/no risk, just as they should seldom have to risk a large punishment for little/no reward.

Allowing time to be invested to reduce risk is a good idea, especially if this investment can not be easily recovered (e.g. by quickloading). The investment adds the time invested to the risk and thus should reduce the risk in other areas (difficulty) if the reward is to remain constant.


Simulating a situation based on the character's knowledge and imagination is a really cool idea. This means that they can invest time in multiple ways: imagining different scenarios; and gaining more information so as to come up with a more accurate picture of what could happen.

These simulations could offer higher bonuses if a simulated situation closely compares to the real thing.

Advertisement
Quote:
Original post by Frequency
I think it's a neat idea but I don't agree with having stat gains for it as others have suggested; such a system should be a gain in itself and its use shouldn't be an affect on the world outside the simulation, lest it turn into a way to cop-out by grinding up to max stats. Also, with this in place, I would expect the game to be difficult enough that most would be inclined to at least check out the simbox before rushing into combat.

Ender's Game was a super incredible book; I almost never reread books but I've gone throught it multiple times.
I think a small, temporary, stat. gain for the duration of the simulated situation would be quite fair as a compensation for the time invested.
Thanks for the feedback folks!

I agree that, if done, the simulation can't be as good as the original challenge or facing the challenge "for real" will just be grind. I was also thinking of making any stat / skill gains temporary so as to represent the bonus of "readiness."





In terms of actual implementation, what system would work best here? Should you:


  • Buy references to known threats and load them as programs (for example, "Dustrogue Combat Prep - Holodisc containing all known data on the fighting style the New California Dustrogue faction.")
  • Have to retreat back to a full-blown danger-room facility, or (in nanopunk style) be able to sit down anywhere and just jack into a portable deck
  • Be able to scan any area, such as one containing dangerous jumps, and test out dealing with known challenges before actually taking them on?


Also, do you think (in the spirit of items for an RPG-like game) that there should be prices, add-ons and levels of quality for a "danger room?"
--------------------Just waiting for the mothership...
Quote:
Original post by Wavinator

In terms of actual implementation, what system would work best here? Should you:


  • Buy references to known threats and load them as programs (for example, "Dustrogue Combat Prep - Holodisc containing all known data on the fighting style the New California Dustrogue faction.")
  • Have to retreat back to a full-blown danger-room facility, or (in nanopunk style) be able to sit down anywhere and just jack into a portable deck
  • Be able to scan any area, such as one containing dangerous jumps, and test out dealing with known challenges before actually taking them on?


Also, do you think (in the spirit of items for an RPG-like game) that there should be prices, add-ons and levels of quality for a "danger room?"


My thoughts here:

Your bullets are kinda outta order in the way I'm thinking of things, so I'll start at what seems to me to be the logical beginning. I like the idea of a "danger-room" moreso than the "plug-in anywhere" idea because of the fact that it is such a useful thing. If at any point we can withdraw ourselves and go play a practice round, we might as well just allow saving and loading at whim. :) However, danger-room facilities should be plentiful and easy to find, otherwise a useful feature will go lacking as it becomes "too much of a trip" to find a danger-room.

Moving to the next point, I like the idea that you could buy (let's call them holodiscs, I like that) holodiscs for certain "threats". Perhaps the danger-room facility will have a list of x,y,z holodiscs that you can play through (generic combats etc...) but you can also choose to have them load your own for you and play through it that way. This kinda, stimulates the desire to go out and explore the world as not only will you want to take on any comers that you might face, but you can also find holodiscs to play around with in the danger-room as well as study to learn how to better defeat your foes.

As far as purchasing upgrades go, I like this idea as well, perhaps the simulation becomes more realistic (closer to the actual combat style although never entirely of course). Other upgrades can tie in directly to point number three where we allow players to create their own holodiscs. Perhaps this will be through an area scanner, which allows them, through the use of some gadget or another, to store all terrain data for a particular location and save it as an environment that can either be explored freely within the danger-room or used as a backdrop for a combat scenario.

Also perhaps (creating a craftsman class?) the player could ultimately create their own holodiscs and peddle them off to the highest bidder. So if player X has the skill and feels he knows quite a bit about the different combat styles of the "New California Dustrogue Faction", he can create a custom combat scenario featuring them. This could be some sort of psuedo-scripting or something if we're speaking of hand-to-hand combat or just general ship selection/arrangment if we're talking a space battle.

When done up with these sort of nifty add-ons, the danger-room simulator becomes much more than just a simple process of trial and error, it can become a game in and of itself and I would speculate would be rather popular amongst the audience if done correctly.

I've put some thought into the idea of stat boosts, and have come to the conclusion that they would most definitely need to be temporary and should probably apply in greater or lesser degree based upon the fact of whether or not the player is fighting the same enemy they were fighting in the danger-room. Running a hand-to-hand combat trainer about the Dustrogue Faction will increase your speed/strength +3 for 100 cycles against a member of the Dustrogue Faction, but will only raise it +1 for 75 cycles against a different opponent. Stat boosts of course as well, would run concurrently, so a player doesn't spend an inordinate amount of time in the danger-room trying to amass stat bonuses of str +100/10000 cls.

And a final note about "purchasing upgrades". Perhaps at higher prices and later levels, players could buy their own portable danger-room simulator to carry around with them (in their ship/backpack/whathaveyou), especially if the world is large, this will allow higher-level players to feel free to wander the galaxy in search of evil-doers, rather than being tied closely to the nearest settlement as most new players are in open-ended games because of the necessity for healing/equipment.

That's my two thoughts, something to chew on,

Vopisk
As a game feature I think it is a great idea - but not as an alternative to saving and loading. I am in the "let the player save if they wish" camp. They bought the game, its up to them how they play it. If you want people to play all the way through then make the game reward doing that, rather than trying to force them not to.

If this idea was used as an alternative to being able to save-at-will then it would cease to be an "option" - the player would pretty much be forced to use this system or risk damage/death. I think this would create nagative feelings towards it is some/a lot of users.

As a feature of a game, which does allow saving at will, I think it is great. In a space game you would have the option to run a simulation multiple time before going into battle (and maybe adjust parameters). In an RPG you might be able to project your disembodied self into a location to explore or gather intel.
Dan Marchant - Business Development Consultant
www.obscure.co.uk
Advertisement
I've just been thinking along similar lines for the X series over at Egosoft. In X3 we had a fundamental problem: new players face an incredibly steep learning curve. Because of the tight production schedule, we didn't have the chance to really deliver any kind of tutorial or help type content to mitigate that. There's some stuff that makes it all a little less "what the hell is going on" for a totally new player, but it's still a pretty massive shock.

Personally, I hate the way most game tutorials are done. They remind me that I'm playing a game, and they're invariably pesky and interfere with the immersion. I've played quite a few games where the tutorial was integrated in the "New Game" process, and usually this is totally borked. However, ironically, the first X game I think is one of the few that really did it "right." The tutorial was presented entirely in the context of the gameplay (flight testing your new spaceship). The content was delivered in a way that was consistent with the game universe, not the real world (the voiceover doesn't say "press the Foo button" and remind me that I'm not, in fact, in a spaceship). There is a minor UI element that shows the actual hotkeys, but instead of being immersion-damaging, it simply acts as an interface between the piece of plastic in my hands and the spaceship that I'm supposed to feel like I'm flying. Since then, we've never really recaptured that feeling in the series, and I can't think of a single other game that made me actually want to sit through the tutorial, just because it was so cool.


[Pet peeve-slash-rabbit trail]
A classic example of ****ty tutorials is, IMHO, RTS games. Invariably it's the same retarded, idiotic thing. No RTS which I have ever played has succeeded in convincing me of my role as a "commander." Frankly, I don't think the damn designers really know what role the player is supposed to fill.

Consider a game like Warcraft III. I start out with some condescending, snivelling tutorial about building some farms, or some crap. I'm treated like a new, moronic, inexperienced sergeant. OK, that's fine - but the game doesn't stay consistent. It talks about the UI in the voiceover that's supposed to be "in game." It refers to things that the game characters shouldn't know about. It breaks immersion. Now, cracking the fourth wall can be extremely funny - if it is done right. Sadly, most designers seem to think that it's OK to just bulldoze the entire wall and pretend it doesn't exist, and that detracts from the experience.

So OK, maybe I'm a new commander and I need to be taught how to build farms. I can deal with that - in fact, it'd probably be fun if the game stayed self-consistent in that regard. But uh oh! I just finished the You're Retarded And Have To Build Farms Level, and now I'm suddenly this highly trustworthy, respected leader who is placed in charge of invading the enemy lands and conquering all the foes. Make up your mind... am I a brainless newbie, or a hero? Self-consistency gets violated so much in RTS tutorials that it's almost sickening. Any form of user education (tutorials) or experimentation (simulator mode) must remain self-consistent with the game's universe, or it will be jarring and uncomfortable to use.
[/Pet peeve-slash-rabbit trail]

So you're probably wondering by now what the hell tutorials have to do with "experiment mode." I think the problem is that most designers think of them as separate concepts (if they even think of "experiment mode" at all). In X2, we renamed the "tutorials" to "simulator missions." The game engine was the same, there was some voiceover to describe the point of each mission and offer helpful hints, etc. The simulator was totally disconnected from the gameplay experience, in every possible way, except that they both used the same space ships and graphics. I never did think that "Simulator" was an appropriate term. You couldn't even access the simulator from inside the game, which meant it was useless as a player education system. Half of our players probably never even realized it was there.

After the X3 release, we started talking about how to regain the tutorial aspect (helping newcomers get introduced gently to the world) without violating the game's universe integrity. IMHO this really comes down to teaching players the game, and having the game itself teach them, without having to step outside of the game and refer to real life (or at least, don't do this very often).

The concept I've been kicking around is to return to in-game education. I think this stems from the same basic issue as "Experiment Mode" - players need to learn how they can affect the world, but they don't want to sacrifice valuable playing time to do it. Especially in this era of increasingly casual gamers, we need a mechanism that lets people always move forward, or at least never go too far back/laterally. Quicksave/quickload is, in my mind, a cheap, stupid hack that gets slightly closer to the goal, but totally fails beyond that.

Really, both tutorials and experiment modes are the same problem - player education and exploration. They're also solvable with the same basic solution, I think. What we need is a system wherein players can poke and prod the world around them without risking damage to their "real" in-game persona. It's a sort of game-within-a-game, so to speak.


To use a concrete example, I'd take the X series (since I spend most of my design-mode time thinking about it anyways [wink]). Start the player out in their spaceship, and they have to get some kind of intergalactic driver's license. Fly around these space-cones, park between these holographic lines, get your little futuristic license card, have fun. Need to learn how to fight the Evil Space Aliens? Head over to the Star Arcade and play the "games" there. Need to learn how to make a profit trading Widgets with the Heathen Natives? Go off to some other kind of "simulator" and goof around.

There's two important aspects to this in my mind. First, the simulation has to have a rich subset of real gameplay, so that the player can genuinely learn by using the simulator. Second, the simulation should never affect players in a way that makes the "regular game" dull. Maybe you can have a huge, compelling, adrenaline-pumping battle in the Space Arcade, but you only ever get to fight the Blob UFOs instead of the Dreadfully Huge Cruisers of Explicit Death that the "real game" enemies fly. It's important not to sacrifice "real gameplay" to the simulator modes. I should never leave a simulator and think "aww, damn, I have to go earn that Massive Bonus Item again." I should leave the simulator and go "wow, if that's the kind of stuff I can do in the simulator, the real thing must be fricken amazing!" The benefits to the player should remain in the player's mind and experience only; they should learn things about the game, on a meta-game level, but the simulator is still just "fake" within the context of the game itself. If it's done any other way, your simulator simply becomes another way to grind and powerlevel. Boooooorrrriiing; sorry, but no, I don't need another Giant Random Wilderness of Wanton XP. I want to have fun.

Now, in a game like X, there's a huge amount of things you can do as a player. The bulk of those things are accessible from the very beginning of the game (unless you're a masochist and start on one of the harder difficulty levels). However, a player can have literally hundreds of hours of game experience, and still not know anything about Feature Foo. That means that, even to experienced players, the simulations have to remain compelling and educational. Most importantly, simulations cannot be condescending. If there is even an implicit assumption that the player is a "noob" when they enter a simulator, they're going to detect it, and probably be offended. Simulators have to remain consistent with the game atmosphere, and not assume anything about things outside the game (like whether or not the player is a newbie).



So there's my disjointed rambling on the matter [smile]

Wielder of the Sacred Wands
[Work - ArenaNet] [Epoch Language] [Scribblings]

I'm sorry I'm cutting trough the whole thread but I can't help myself :)
Sorry if this has already been said.

First, the possible flaws. There are a couple of risks with these simulations:
If they're too perfect, they'll be just like saving, and you can be certain that by sticking to the plan you'll be just fine, which ruins the experience - there are no more surprises.
If they're too far off from reality, the player will feel very cheated, and with every right of course!

Now I had an idea that could make it a lot more fun. Remember Ocean's 11, Entrapment and all those cheesy movies about big heists - thieves always rehearse before trying the big hit. But it's their responsability to make the simulation as close to reality as possible!
So delegate all responsability to the player. "Ok, at this point we get to the main gate... expect some resistance so I'm gonna throw in a couple of trolls" :P I'm assuming a kind of holo-room for this. You see, the player would build the simulation as he went along, according to his knowledge, and judge from that!
If he was wrong, too bad. Should've gathered more intel. A smart move is to always over-estimate your opponents.

This is immensely more fun, and you don't have to worry about balancing the simulation with reality.
Quote:
Original post by Obscure
As a game feature I think it is a great idea - but not as an alternative to saving and loading. I am in the "let the player save if they wish" camp. They bought the game, its up to them how they play it. If you want people to play all the way through then make the game reward doing that, rather than trying to force them not to.


Would your opinion about saving change any if:

  • The game had no "game over" (outside of the natural end of an optional plot)
  • Death was a setback rather than the end of the game
  • Part of the whole point of the game was avoiding setbacks and creating strategies for surviving and thriving


With saves, certain experiences are impossible: A choice in a dialog tree, for instance, means a lot less if the game tempts you to simply "reload and walk the tree." Loss of a critical character, betrayal and ambush, fateful decisions all have little or no meaning.

By convention, saving is not expected in some experiences. You can't incrementally save after each turn and roll in a flight sim, for example. Why not? Most racing games can't be saved just before a sharp hairpin or dangerous jump. Same with playing a basketball game.

Part of the reason we expect to be able to save is simply convention, but most of it is a rightful distrust that the game developer does not respect our time. Many seem to have the attitude that "if you can't make it then you need to just try harder" (and screw you if you can't-- you already bought the game and can't take it back once you learn level 18 is impossible, or whatever)

I'd like to try to change that paradigm, and get you into a confident playing mode where you don't miss any sort of out of game restore, but rather use ALL of the game's in-game means to balance risk, take chances, and get better and better at recovering from setbacks.
--------------------Just waiting for the mothership...
Quote:
Original post by Jotaf
So delegate all responsability to the player. "Ok, at this point we get to the main gate... expect some resistance so I'm gonna throw in a couple of trolls" :P I'm assuming a kind of holo-room for this. You see, the player would build the simulation as he went along, according to his knowledge, and judge from that!
If he was wrong, too bad. Should've gathered more intel. A smart move is to always over-estimate your opponents.

This is immensely more fun, and you don't have to worry about balancing the simulation with reality.


Sweet. That's very cool. So it's up to you to try to get layouts of locations, enemy types and some form of knowledge (not exactly sure how to convey this) of their behaviors?

Do you think you should be able to buy all of this as intel / information? What about hiring a character to scout and model the target, with some incorrect / incomplete information as a result of their lack of skill?

I'm assuming, btw, that even if simulation is safer, the player will eventually run out of resources and be motivated to actually take the risks that match the sim.
--------------------Just waiting for the mothership...

This topic is closed to new replies.

Advertisement