Consequences of violence -- have your say!

[Wilbefast]

Hi all. Today I have a question for the community

Overgrowth is to be a game about the consequences of violence, though whether those consequences will be interactive or not, and to what degree, are things as yet not entirely decided.

Since players are playing the game to be entertained by the fighting, I think we need new kinds of stories that can complement the gameplay without distracting from it. Here are some of the questions I'm thinking of exploring with the Overgrowth story to keep it focused on fighting:

• Is it better to win a fight dishonorably or to lose honorably?
• What are the consequences of sparing someone's life in a fight to the death, or killing in a fight just meant to establish status?
• Who would dare to attack a rabbit who has killed wolves with his bare hands?
• Does a reputation for great fighting skill make you safer or less safe?
• What does it mean to be a fighter in a culture that hates fighting?
• When is lethal force more effective than a show of strength?

http://blog.wolfire.com/2009/01/fightin ... in-itself/

The story will mainly be linear because we're just too small to be able to support an intricately branching decision tree. However we would like to include one or two decision nodes for key moments in the narrative if possible.

http://blog.wolfire.com/2009/07/wolfire ... questions/

With Overgrowth, the idea I am leaning towards for the story is to explore real consequences of violence -- focusing on different levels of force, when they are needed, and when they just make things worse.

http://blog.wolfire.com/2010/04/Saving- ... -at-a-time

If actions and their consequences were to be implemented, how would you like to see them materialize? Bearing in mind of course that Wolfire is a very small team: think too big and your idea won't see the light (albeit fun to ponder). Pro-tip: "branching" rhymes with "combinatorial explosion", so best avoid it. Think of a clever way of doing things though and the team might just be able to implement it.

I eagerly await your comments

[Hibbe]

Over the past few days I have been thinking about different situations that could occure in the overgrowth world.

Since there is no real story yet, Iv'e imagined cats hiring mercenary dogs to track and kill wolfs.
( Which, if someone reads this after the game is released might think is crazy )

This is a really interesting topic and very exciting to explore.

( I dodged a answer because I don't have one right now )

[etrapp]

Maybe they should implement some kind of reputation system, that changes the way other NPC's see you. Like if you brutally kill someone, others might be scared. Or if you win a fight honorably you are respected. Or something like that.

[Djemps]

I think something David might want to consider is how to better implement some of the current content that he has already placed in the game.

I'm thinking specifically of the 'choke out' and 'knife hostage' animations. For me, the ultimate question is 'what's the point?'.

Currently, the choke maneuver always ends in a death / knock out. So its only tactical purpose is to quickly suppress an enemy that might have caused a problem. You can reposition yourself a little bit, but the enemy usually chokes out before you can actually get to a new location.

In contrast, the knifepoint capture allows you the ability to decide when to finally kill your target. Not to mention the fact that you can move all over the place. We've been given a choice to kill or not. But what is the point of waiting? Why not immediately slit their throat?

For me, one solution is to implement a very simple ranking system into the game. For example:

Civilian < Guard < Captain < Boss / Wolf

numerically, it might look like:

0 < 1 < 2 < 3

These rank values might come into play when someone is taken hostage, or when Turner is fighting multiple enemies at once.

For example, a group of enemies might stay back and let you pass as long as the hostage you are holding is equal to or greater than their combined rank. If a guard has a value of 1, he won't attack if you are holding another 1_value guard at knife point. But, if you take a basic guard hostage in front of two of his buddies, they will disregard his safety and rush to attack Turner. How about a 2_value Captain? You would need to have to three guards in a group before they thought about attacking.

Similarly, ranking can be applied to an active battle. I'm no code expert, but I imagine there could be a way for a group of allied fighters to form a value based on the sum of their rank (three guards = group of force 3). In my mind I see two basic rules for groups:

1 - If the group's total force value is immediately cut in half, all remaining members temporarily pull back for a moment until a new group value is calculated.

2- If the group's total force value is immediately reduced to less than half, the remaining member(s) flee or surrender.

Let me try two examples in an attempt to make a little more sense...

A group of three 1_value guards are fighting Turner. Their total group value is 3. Turner kills the first guard. The group's value has been reduced to two. The remaining fighters don't even flinch and continue fighting with ferocity. The second guard dies. Now the group value has been cut in half from 2 to 1. The final soldier stops and backs up for a moment before re-engaging Turner. Of course, once he too is finally on death's door, he would probably turn and run according to the current game mechanics.

In a second example, Turner is fighting a 1_value Guard and a 2_value Captain. Their total group value is 3. Right off the bat, Turner kills the Captain. The lone Guard turns tail and runs since his group's value was immediately reduced to less than half. (3-2 = 1)

I've been thinking about this possible ranking system for a while now. I hope it makes a little bit of sense to the rest of you guys!

Also, concerning some of David's questions concerning force and culture:

I think that a pack based culture would place plenty of importance on the ability to show one's force without killing. This is how you rise to the top of the pack. But you still need a pack to fight alongside you!! If you kill everyone then you will be weak and alone in no time flat. Manipulating different animal cultures to your benefit sounds like a very simple, yet tactical game element. If Turner manages to slap down a Dog pack's Alpha male while still sparing his life, the rest of the pack might actually welcome him and fight at his side later in the level.

[Korban3]

^
That.

The group sum sounded along the right lines at first, but not quite correct. Then you explained it, with the two scenarios, and I was all "Perfect".
I'd love to see something like that. And maybe have the individual force modified by different attitudinals, of a sort. A pair of 1 rank guards see Turner. They chase him through some narrow, crete walled in areas. Turner kicks one, who's skull promptly impacts a wall and caves in, leaving a huge mess on the floor.
Still alive guard slips in blood.
Shits bricks.
No longer wants to fight.

So, it would look like:
My list of attitudinals is free, I am healthy, my group force is 2, I see a single enemy that I am not familiar with. We can kill him. Charge.
Enemy is running away, we get a confidence boost attitudinal, making us more brash. We run into a particularly tight area after enemy.
Enemy kicks buddy, buddy hits wall. Group force is halved. There's blood everywhere. Hesitate. Resume chase, but the confidence attitudinal is removed and we're given a shock one, deadening our reflexes just a bit.
Slips in blood and hits the ground, hard. We get chances of broken bones. We get the WTF attitudinal. If the character is particularly weak-stomached under-trained or just really has been fucked with, he may cut off the chase and vomit from nausea or flee from terror. Or assume the fetal position from mental overload.

Would be tough to implement, but it'd be cool as fuck and isn't too much a stretch of David's abilities, I don't feel. Then again, I'm not David, so I wouldn't know too much about that.

[Jacktheawesome]

I think something David might want to consider is how to better implement some of the current content that he has already placed in the game.

I'm thinking specifically of the 'choke out' and 'knife hostage' animations. For me, the ultimate question is 'what's the point?'.

Currently, the choke maneuver always ends in a death / knock out. So its only tactical purpose is to quickly suppress an enemy that might have caused a problem. You can reposition yourself a little bit, but the enemy usually chokes out before you can actually get to a new location.

In contrast, the knifepoint capture allows you the ability to decide when to finally kill your target. Not to mention the fact that you can move all over the place. We've been given a choice to kill or not. But what is the point of waiting? Why not immediately slit their throat?

For me, one solution is to implement a very simple ranking system into the game. For example:

Civilian < Guard < Captain < Boss / Wolf

numerically, it might look like:

0 < 1 < 2 < 3

These rank values might come into play when someone is taken hostage, or when Turner is fighting multiple enemies at once.

For example, a group of enemies might stay back and let you pass as long as the hostage you are holding is equal to or greater than their combined rank. If a guard has a value of 1, he won't attack if you are holding another 1_value guard at knife point. But, if you take a basic guard hostage in front of two of his buddies, they will disregard his safety and rush to attack Turner. How about a 2_value Captain? You would need to have to three guards in a group before they thought about attacking.

Similarly, ranking can be applied to an active battle. I'm no code expert, but I imagine there could be a way for a group of allied fighters to form a value based on the sum of their rank (three guards = group of force 3). In my mind I see two basic rules for groups:

1 - If the group's total force value is immediately cut in half, all remaining members temporarily pull back for a moment until a new group value is calculated.

2- If the group's total force value is immediately reduced to less than half, the remaining member(s) flee or surrender.

Let me try two examples in an attempt to make a little more sense...

A group of three 1_value guards are fighting Turner. Their total group value is 3. Turner kills the first guard. The group's value has been reduced to two. The remaining fighters don't even flinch and continue fighting with ferocity. The second guard dies. Now the group value has been cut in half from 2 to 1. The final soldier stops and backs up for a moment before re-engaging Turner. Of course, once he too is finally on death's door, he would probably turn and run according to the current game mechanics.

In a second example, Turner is fighting a 1_value Guard and a 2_value Captain. Their total group value is 3. Right off the bat, Turner kills the Captain. The lone Guard turns tail and runs since his group's value was immediately reduced to less than half. (3-2 = 1)

I've been thinking about this possible ranking system for a while now. I hope it makes a little bit of sense to the rest of you guys!

Also, concerning some of David's questions concerning force and culture:

I think that a pack based culture would place plenty of importance on the ability to show one's force without killing. This is how you rise to the top of the pack. But you still need a pack to fight alongside you!! If you kill everyone then you will be weak and alone in no time flat. Manipulating different animal cultures to your benefit sounds like a very simple, yet tactical game element. If Turner manages to slap down a Dog pack's Alpha male while still sparing his life, the rest of the pack might actually welcome him and fight at his side later in the level.

Well right off the bat this is completely awesome and should be included. As for what you were saying in the beginning about the choke and throat slit maneuvers, I think part of the problem is that they aren't finished yet. I'm almost certain we'll see more out of those. One thing I always thought was there should be an unarmed hostage maneuver that doesn't end in death. Currently, the choke starts immediately, and takes about 7 seconds to complete. There are ways to restrain someone without suffocating them, and that's what I would like to see first.

[BotheredMe]

I always find it interesting when I hear David talk about how he envisions the gameplay for overgrowth, especially when he added the fighting with scabbards bit, followed by a remark stating that it could be used to teach someone a quick lesson without anyone being hurt. As it seems, David has been adding so much depth to the AI, I'm not sure exactly how this will end up. My dream for this game, if not OG then the third installment, is for there to be a full spectrum of characteristics that determine the AI's actions.

I feel that David has hit it head on with the fear aspect, where they run away to get help when they are being killed. What really gets me is the new injury system where enemies and yourself will clutch at wounded portions of their bodies. When I saw this in one of the more recent alphas, I kept thinking that this wound end up serving as a choice module in which you may spare them or finish them off, one with the consequence of possibly being discovered and the other of avoiding notoriety, making guards in nearby cities less wary of your lethal intentions. Also, sparing them could hold a sort of relief of them aiding you in a future quest.

Also, your overall reputation should be accompanied with different reactions from enemy AI and friendly NPC's later on in the game. When you are well-known as a rabbit who has slain every type of beasts, large and small, enemy AI are less likely to confront you by themselves. Should a guard see you, they will quickly run for aid or slowly walk away to avoid triggering you to go chasing them. When in comes to big groups, I figure that after a certain amount of enemies have been killed, a fight or flight factor kicks in, maybe a future parameter, in which the AI chooses either to flee or to fight with all their might due to their own personal drive for pride/honor.

As for friendly NPC's, I imagined that being known as a sparing man (err... rabbit) who spared many lives would be praised by the communities whilst a ferocious man (err... rabbit... grr.) would be feared and NPC's would cower from your sight. You would get help from NPC's or be attacked due to your reputation and the choices you've made throughout the game. Perhaps someone trying to prove themselves as a worthy fighter would challenge you due to your high status, or maybe you partook in the takeover of a city that ended in many casualties and an NPC feels the need to bring justice to those casualties, kind of Rorouni Kenshin style. On that note, Sakabatou would be a great, non-lethal weapon. Just putting that out there

Those are just some ideas I've imagined from all the possibilities this game seems to be unfolding. I don't expect Hitman Absolution type AI, but I'll be damned impressed no matter how they end up creating this game. Go Wolfire!

[Wilbefast]

I'm thinking specifically of the 'choke out' and 'knife hostage' animations. For me, the ultimate question is 'what's the point?'.

You're right this feature has currently been implemented without a great deal of use in gameplay for the time being. Where the joke is concerned though, I think it was meant to give a definite advantage to using a knife.

Manipulating different animal cultures to your benefit sounds like a very simple, yet tactical game element. If Turner manages to slap down a Dog pack's Alpha male while still sparing his life, the rest of the pack might actually welcome him and fight at his side later in the level.

True, but the question I'm asking in this thread is how precisely this "very simple, yet tactical game element" might be implemented. The simplest way would be to script everything my hand, but that's not really a practicable solution for a sample team.

Still alive guard slips in blood. Shits bricks. No longer wants to fight.

One way this could be implemented is through a morale resource: each character updates an integer based on the perceived situation. Running out of morale means fleeing, which itself would influence the morale of allies, creating feedback loops (snowball effects) and other emergent results.

Morale represents the confidence that a unit has in victory. If a unit reaches "broken" morale, then it will rout. Morale ranges from eager, the highest level, to steady, to shaken and then finally to wavering, at which point the unit will most likely begin to rout.

Total War also added modifiers like "berserk" which rendered soldiers immune to moral penalties but impossible to control.

Anyway, such a system could either be event-based (the death of a friend is an event) or status-based (continuously query the status of all friends) or both. Notions of "leader" and "group" could be added, the latter being especially useful for caching neighbours for easy consultation.

a choice module in which you may spare them or finish them off, one with the consequence of possibly being discovered and the other of avoiding notoriety, making guards in nearby cities less wary of your lethal intentions. Also, sparing them could hold a sort of relief of them aiding you in a future quest.

a sparing man (err... rabbit) who spared many lives would be praised by the communities whilst a ferocious man (err... rabbit... grr.) would be feared and NPC's would cower from your sight. You would get help from NPC's or be attacked due to your reputation and the choices you've made throughout the game.

Specifics? For instance "reputation" could take the form of an integer updated between levels based on the number of enemies killed. On a micro scale this could interact with enemy morale: for instance the effect of a perceived morale-event could be multiplied by Turner's reputation variable.

Code: Select all

event(MoraleEvent e)
{
  switch(e.type)
  {
    // ...
    case(TURNER_KILLED_MY_FRIEND):
      morale -= FRIEND_KILLED_PENALTY * turner.reputation * affinity(self, e.whom);
    break;
    // ...
    case(TURNER_RUNNING_AWAY):
      morale += ENEMY_RUNNING_BONUS * turner.reputation;
    break;
  }
}

In this example reputation is a double-edged sword, as enemies are that much more inspired to be winning if you're a reputed fighter. Other things like "affinity" can be added too: rats for instance, would be very spooked but Turner killing a wolf. The same would not be true of wolves when a rat is killed.

[AAorris]

Ahh, I'm glad you're back, Wilbefast.

This kind of discussion and the dev logs are a huge reason why I preordered!

I have some ideas, but I'll take a moment to refine them.

[last]

I do like the ranking system idea, but what happens when someone messes with character parameters and make that character weaker or stronger? How to recalculate ranking then? Also there should be ranking for weapons to, because this is also effecting your fighting.

so the weapon ranking should go something like this: no weapons at all even not sheathed weapons, scabbards or something like them, knives, swords, and so on.
Oh and once the armor system is out, then this requires ranking to.

[Wilbefast]

I do like the ranking system idea, but what happens when someone messes with character parameters and make that character weaker or stronger? How to recalculate ranking then?

You could just include all the parameters in the ranking:

Code: Select all

rank := speed*SPEED_K + strength*STRENGTH_K + knife*KNIFE_K + sword*SWORD_K + ... 

Where knife and sword are equal to 0 (hasn't a knife) or 1 (has a knife). Linear functions is easy as hell for computers. This would actually be a better way of letting the system know the "value" of an arbitrarily-defined character to other arbitrarily-defined characters than setting out a bunch of pre-defined classes (civillian, soldier, etc).

[Jacktheawesome]

The problem with the character parameters is that they aren't immediately evident. Say a guard sees a captain: the guard will immediately know by the ranking insignia the captain carries that he faces a higher class fighter, and should run away or get help. You could then adjust that guard's parameters to have twice the strength of the captain. Well, the guard is now clearly more powerful than the captain, but the captain wouldn't realistically hesitate until he sees the strength of the guard. So with that factored in, the appearance of strength and actual strength are two completely different things that must be factored in.

[Wilbefast]

the appearance of strength and actual strength are two completely different things that must be factored in.

I was assuming that ranking would be used for allies, id est to estimate the value of a friend that has been killed. If we can assume that friends know each-others fighting abilities pretty well then we can give them this insider knowledge. As for enemies, fear of a strong enemy would be emerge from their killing strong allies, as this would greatly decrease morale. Meanwhile weak allies landing a blow on an enemy would raise morale more than strong allies landing a blow.

Can you modify character's degree of aggressive/defensive in Overgrowth? If so you could feed to morale attribute into this so that demoralised characters would be more likely to hang back and overconfident characters would run in without thinking too much about their safety.

[Jacktheawesome]

the appearance of strength and actual strength are two completely different things that must be factored in.

I was assuming that ranking would be used for allies, id est to estimate the value of a friend that has been killed. If we can assume that friends know each-others fighting abilities pretty well then we can give them this insider knowledge. As for enemies, fear of a strong enemy would be emerge from their killing strong allies, as this would greatly decrease morale. Meanwhile weak allies landing a blow on an enemy would raise morale more than strong allies landing a blow.

Can you modify character's degree of aggressive/defensive in Overgrowth? If so you could feed to morale attribute into this so that demoralised characters would be more likely to hang back and overconfident characters would run in without thinking too much about their safety.

Yes, of course the allies wouldn't be subject to this; we assume they would know their friends' strength. And as for enemies, yes. When the battle had begun, all would function normally, with morale dropping/rising based on the sum (adjusted for actual strength) of each side. The difference in appearance vs ability would manifest in the willingness to attack in the first place, not so much the battle itself. In essence, it would add an "underestimation" factor.

[AAorris]

Okay, I'm back, and I've collected my thoughts.

My ideas are focused on exploring the consequences of your actions
...in a narrative sense.
Well, I mean that the consequences of your actions are embedded into the story. The gameplay enforces a respect for the lore. Hmm, let's look at some examples to get on the same track:

Here would be my generic observations on each race's perspective on combat:

• rabbits may defend themselves, but would much rather run to live another day.
• Dogs would never run from a fight. neither would they hesitate to defend what's important to them.
• Wolves have utter confidence in their ability to survive. This confidence could often be their downfall.

The nuances of each of these concepts could reveal numerous things about gameplay. Finding these things would be intuitive, and really satisfying! Here are some cases I threw together. Each one sets Turner against the odds, where the player would need to use his knowledge of each race to survive the encounter.(Doesn't that sound exciting?)


Here. turner is surrounded by wolves in the forest. The player would look around and notice quickly that he is in a -bad- situation.

As far as the map goes, there is thick, thick bush all around where Turner would be caught and ripped to pieces if he tried to run, save one exit through the river.

The wolves would be circling as a pack, but initially, they wouldn't all fight together. Confident, wolves would not interrupt a 1 on 1 fight, but if the player tried to run, the wolves would all join in and make sure you didn't get away.

Therefore, the player would have to skillfully defeat the wolves, one at a time, until the gap between each wolf left an opening to the exit and turner might be able to manage an escape.



The second case is a stronghold full of raider rabbits. As far as raiders go I imagine two classes of them - the abusive leaders and the fearful pawns. For interests sake, I will assume that they can be classified like that for now.

The level here could be approached in many ways, but for the story we will imagine that there are friends turner is looking to free, and his goal would be to simply get the raiders the heck out of there, so he can get his friends away safely.

The simplest plan of action would be to break in, take a hostage, and use it to make the more fearful rabbits cautious. However, the ones who are brainless and/or in charge would approach. Cutting the throat of the hostage could earn some extra distance from the nervous rabbits, and turner could manage the attackers until everyone flees, or dies.

That said, in general, bodies could be used to wean out any unnecessary fights or create enough hesitation to have the advantage - and against other classes and races, could also be a sure way to get ganged up on.



It should be clear at this point the feeling I'm going for, and your opinions on if this makes sense or if Wolfire would enjoy such an approach ought to be set by now, but here's one more that I haven't really thought through.

It's a dog fort, perhaps it's an armory where they've taken and stored things that turner values. The dogs would be patrolling casually.

Turner would have to use stealth here, as if any dogs noticed a struggle, they would make enough noise to make sure that everyone on the map was after you. You should notice from the very beginning that you won't be scaring them off, and that they won't hesitate to dog-pile.

-------------------------

In the end, that kind of mechanic is something that would be a real asset to level design. It would create a feeling of increased immersion, choose-your-difficulty approaches, and punish players who try to do obviously stupid things. In lugaru, I was taught the hard way how NOT to do a lot of things.

Anyways, I just wanted to put a different perspective in. The mechanics you guys are all talking about are very interesting! Although I don't know if such complex ranking systems would be convenient for either the devs or modders. My views have similar, unexplained, or possibly worse consequences but I enjoy thinking about it! How do you guys feel about it... in a narrative sense? Perhaps a mix-and-move-forward is in order
...Now I need a nap.