How would you design gameplay for a situation where a player has to fool a bunch of enemies into doing something?

Here's an example scenario: You're trying to get some resource, like water, for your little town but it's being guarded by a group of bandits. The town is remote, there's no real law and you can't fight them because they're too strong.

So you want to use the following ruse: You create or acquire some item that's attractive to the bandits, like food, then give it to them. Once the bandits use the item something negative happens-- they're cursed, spooked, poisoned, mentally enslaved, whatever ultimately renders them harmless.

You want the player to be concerned with

• The time it takes to prepare the ruse (ex: invent the poison) vs. how long the town can do without the resource
  
• How effective the ruse will be, including
  
  
  
  • How suspicious will the bandits be, determining if they even fall for it?
    
  • How fast will they accept and use what you give them (this MUST be somewhat random so the situation can be repeated)
    
  • What they will have to do in the meantime while waiting for the ruse to take effect
    
  
  
• What the larger consequences will be for the ruse (i.e., will someone come looking for the bandits?)
  

What information does the player need to see to evaluate this situation? What gameplay verbs would you use to execute the plan?

Maybe It Could Work Like This?
A simple way to do this would be something like a "Ruse Skill" versus a Gullability stat. The gameplay would then revolve around a special give action ("give as a ruse"). Enemies could then have a Suspicion stat that lowers as you succeed in raising your social standing with them through skill use.

If enemies were part of known factions then faction strength and character could determine if someone comes looking. If a new group of bandits show up, they'd have a higher Suspicion stat, so you'd have to craft better ruse items to get them until they were gone.

Kind of bland, so I'd welcome improvements.

How often does this scenario occur (compared to all the activities that the player can do in the game)? Can the player always try to trick an NPC, or can the player only trick specified NPC?

Quote:

How would you design gameplay for a situation where a player has to fool a bunch of enemies into doing something?

What happens if the player fails the first time? What do you want the player to do to make the ruse successful? Do you really mean 'has to' or just 'could'?

If tricking people is one of the core activities in your set of gameplay, then I think it should have the same level of complexity and universality as your other activities.


An idea:

Make ruse items rare items that the player would save up for special occassions. The effects of the items are described, it tells the player whom it may be good against and how to use it. The difference between this and your option 1 is that the player cannot intentionally create or acquire ruse items (or items that turns an ordinary item into a ruse item). It is like a trump card that the player has been carrying. The point behind this is to break the use of ruse a routine, so everytime the player uses it, it feels special.

Concept: make something interesting by disallowing it to become a routine.

You occasionally have them in strategy games. For instance in a lot of games the AI is super aggressive to transports(empty), even chasing with units that could never catch them. You can also trick some AIs into fortifying areas that "are worse" for defense strategy wise by always attacking with a few units on X front, instead of the front you actually plan on.

It’s an interesting challenge and something I’ve been working on myself as I want to create game play heavily revolved around fast talking, deception, and infiltration.

The game play approach I’ve taken is key and threshold. To perform a task, or over come an obstacle requires a “key” and possibly a threshold value. A key could be a skill, piece of knowledge, type of object, or a specific object. While a threshold is the value that must exceeded to succeed. So to unlock a door requires either a key or an item that contains the lock pick property and overcoming a security threshold value. There is also a whole host of entity chaining and onioning but I won’t derail this thread by going into the details.

I haven’t extended it to NPC/Factions yet but I would take the approach that Factions have traits, wants and needs.

So to trick the bandits you’d need something they want or need, then convince them to accept it. Throw in some situation and history modifiers, and take into account the bandits traits and you’d get a result of some sort.

In your example the lets say I’ve got a leg of roast lamb that I’ve laced with some poisoned herbs. The bandits want food and since I’m great cook it’s got a score of 50 points, the threshold to succeed might by 50. Now you throw in a bunch of other factors to see if I succeed in getting them to take the poison.

Example:
If I show up to offer the lamb as a gift they are instantly suspicious -10
If I ask to trade it for a barrel of water +5
If they are running out of food +20
Current reputation with the Bandits -10
Bandits have the greedy trait +5

And so on until I get the final score and they decide to take it or not. The main game play then comes down to 4 phases The Prep, The Plan, The Execution, and The Aftermath

Prep work would involve preparing gathering the poison and preparing the poisoned leg of lamb.

Planning would be about setting the situation in the most favourable way. Sending hunters out to cause bears to come down the mountain early, setting up scent packs to scare away animals to make it harder for the Bandits to find food.

The Execution comes when I try get them to take the lamb, whether I try and offer is a trade for water making them feel they have upper hand or try give as gift or tribute. Do I give it to a sentry, or the leader, are others around?

The Resolution comes from the execution and the reactions of those involved. Maybe the sentry didn’t share the food with others. The bandits might react accordingly with one of there members poisoned if they realize he was poisoned. They might all enjoy the food and then die. Do the bandits have friends that will come looking? Are they just a small branch of a major gang?
But then that’s just one path, how would it change if instead of food I offered them a large ruby leave the village in peace? Depending on that bandits and how it was given they might leave, they might kill each other over the ruby, or they might raid the village to steal anything else that of value before moving on.

But that might over the top for what you want. I’m trying to build game play around all of that.

If you're giving NPCs a Suspicion stat, it should also affect how they react when your intentions are sincere.

Thanks for the feedback, everyone (esp. since I wasn't really able to get back to this post).

Quote:

Original post by Wai How often does this scenario occur (compared to all the activities that the player can do in the game)? Can the player always try to trick an NPC, or can the player only trick specified NPC?

I was thinking about how this might work for all NPCs in a sandbox environment.

Quote:

What happens if the player fails the first time?

At the moment it seems like the situation needs to be repeatable and there needs to be a range of responses (insults, theft, violence) as possible responses. I'm again seeing a meter/stat/value governing this which increases with failure. At max maybe the player is told "they'll kill you if you try something like that again."

At that point, the player needs to avail themselves of other non-ruse options (flee, go to down to hire mercenaries, etc.)

Quote:

What do you want the player to do to make the ruse successful?

Best I can think of at the moment: Combine objects to make a ruse object then use maybe use character skills (flatter, fast talking, lying) to get it accepted.

Quote:

Do you really mean 'has to' or just 'could'?

Figure of speech, but I used has to to mean most desirable course of action out of all potentials.

Quote:

Make ruse items rare items that the player would save up for special occassions. ... The point behind this is to break the use of ruse a routine, so everytime the player uses it, it feels special.

I agree that making it special would be best else why wouldn't the player continually spam the option to perform ruses. Alternately, I could use some sort of governing resource that naturally imposes a cooldown or forces a variation in strategy.

Another, more complex approach would be to somehow give the items traits that map to specific NPC traits. Starving folk like food, greedy folk like gold, that sort of thing. There might then be the additional dimension of how the object will be used-- food is eaten, gold is not. The strategy might then be getting the information that makes up the opponent's weakness and then creating/finding the ingredients.

Quote:

Original post by TechnoGoth The game play approach I’ve taken is key and threshold. ... There is also a whole host of entity chaining and onioning but I won’t derail this thread by going into the details.

This sounds interesting. Actually if you have some time a few details might be helpful. Do you see any advantage to multiple thresholds with specific modifier categories, or do you think it would be easier to implement and understand if done on a simple linear scale?

Quote:

Throw in some situation and history modifiers, and take into account the bandits traits and you’d get a result of some sort.

History modifiers would be a good way to help deal with past player failure.

Quote:

And so on until I get the final score and they decide to take it or not. The main game play then comes down to 4 phases The Prep, The Plan, The Execution, and The Aftermath

Item creation / inventory use really make this easy to understand. Aftermath could easily be handled by logic that evaluates status effects and attacks.

Quote:

Do I give it to a sentry, or the leader, are others around?

It would be good to try to think in terms of behaviors the player can observe and exploit in this case. Maybe an NPC has a hording behavior, for instance, and you have to somehow be given enough information to know he won't share the lamb.

Quote:

The bandits might react accordingly with one of there members poisoned if they realize he was poisoned. They might all enjoy the food and then die. Do the bandits have friends that will come looking? Are they just a small branch of a major gang?

I can't seem to see a way of getting this across to the player except by just boldly (and unimaginatively) telling them. "Bandit group X is an isolated bunch of outlaws with no friends or family..."

Quote:

But that might over the top for what you want. I’m trying to build game play around all of that.

Not at all, that's good grist for emergent gameplay. Love to hear more.

Sure not a problem,

To start off the core of my game engine revolves around an entity framework, in this case an entity can represent anything from a concept to an in game faction.

Entities have 4 key pieces of functionality
Associations
States
Blocks
Hooks

Associations are other entities contained or connected to the entity

States govern the current state of the entity its available actions if any, and state transitions.

Blocks are other entities that prevent the entity from transitioning from one state to another; typically the blocking entity prevents the change by being or not being in a particular state.

Hooks are connected to state transitions and trigger an event on the watching entity.

That’s all a bit abstract but to here’s a working example:

Entity: Door
- States: Opened, Closed

Entity: Spin Lock
- States: Unlocked, Locked, Disabled
-> Associations: Level 1 Access Code

Entity: Key Card
-> Associations: Level 1 Access Code

Door has a block on Closed -> Opened when the Lock is the Locked state.

In this case you’ve got a door with a lock, and a key card. The door can’t be opened if the lock is locked. The lock requires level 1 access to unlock. The key card can the unlock the lock as it has

The difficult of the challenge can be increased through chaining by adding an alarm to the lock that has a hook on Locked->Disabled which causes the alarm to sound if the lock is disabled.

Onioning can also be used increase the difficulty by adding an additional separate challenge like a security camera facing the door. Which provides an added layer of difficulty not connected to any of the other challenges.

The Key and threshold based action all comes into play when trying to over come those challenges. If we look back at the basic task of unlocking the door. It can be accomplished by acquiring a keycard with level 1 access, which could be bought, stolen, found, or downloaded.
Or tool with lockpick entity could be used to pick the lock. In which case lockpick and level 1 access would both be “keys” as they allow actions various actions to be performed on the lock. When using the lockpick tool however there is threshold check which a simple score check where if the security level of the lock is exceed by the lockpicks intrusion level+ the players security skill + modifiers the lock is disabled. You mention multiple threshold levels and to be honest I didn’t initially think of them but they could fit in nicely in the system. A second threshold on the lock test would mean the lock is unlocked rather then disabled which mean you wouldn’t have to first disable the alarm if you pick the door well enough.

I want to play around with modifiers a lot as part of the system as well. Have couple of drinks with an ex security guard at SafeCo and find out about a flaw in their ElChepo brand locks giving you +5 modifier against them until flaw is patched. Or working in the dark and suffering -20 on the attempt.