"City Threat Level" Feature

Vae on 7/1/2010 at 23:59

A fail by killing in a no-kill mission has been a simple and effective way of making things more difficult in THIEF...but perhaps there is a better way to handle it.

A variable "City Threat Level" (CTL) feature could be used to dynamically increase difficulty by means of a security response to killing in THIEF. This increase in security/difficulty would act as a counter-balance to any killing in the City, and could replace the "brick wall fail" for killing that is used with the fixed difficulty levels, or work in conjunction with it. This "City Threat Level" (CTL) feature is a form of dynamic difficulty, and its' effect could also be scaled in relation to the fixed difficulty levels.

As the campaign progresses, the player would influence the CTL by his/her behavior. Killing in a mission would increase the CTL for the next one, and refraining from killing for a mission would decrease the CTL for the next...the more killing, the more difficult it will become for the player to achieve their objectives, as a result of an increase of security and alertness of the people of the City. These specific reactions are outlined below:

1) Less available unsecured loot

People are simply nervous about carrying or leaving valuables in unsecured areas, and are either locking them up and/or hiding them better, for fear of getting killed because they own them.

Mechanics:
This can work alone, or be used in conjunction with a randomized loot placement/amount system, and used as a modifier. The more killing, the greater this modifier would be, reducing available unsecured loot.

For those unfamiliar with the random loot concept, I recommend reading (http://forums.eidosgames.com/showthread.php?t=91959) "Randomized Loot" and (http://forums.eidosgames.com/showthread.php?t=89522) "Death to the Loot Stat".

2) More secure portals and containers

Again, people are getting scared of all the killings, and are keeping more of their doors and windows locked for safety...they are also buying better locks for portals (doors and windows) and for secure containers (safes, chests, jewelry boxes, etc.). Some are even starting to trap their locks, trapping more locks, or are creating better traps for those locks that are already protected.

Mechanics:
This can work alone, or in conjunction with randomized locks and traps, as a modifier. The more killing, the greater this modifier would be, resulting in more locked and/or trapped portals and containers. The difficulty for picking locks and disarming traps, on the whole, would also increase. If enough killing is perpetrated, even some locks may become unpickable...permanently denying any goodies or places of delight...;)

For those unfamiliar with the randomized locks and traps concept, I recommend reading (http://forums.eidosgames.com/showthread.php?t=97203) "Randomized Locks and Traps".

3) Heightened A.I. security

People of the City are becoming more alert and wary, now that they're aware that there's a killer on the loose. They have become more sensitive to strange noises and peculiar shadows...and are assuming the worst when in doubt, just to be safe.

Guards and the City Watch are increasing in their numbers and their frequency of patrols...and it's becoming less and less common to find one working alone. They also seem to carry torches more often.

Commoners and noblemen alike are walking the streets less at night, and more of them are deciding to arm themselves or walk with a friend.

Mechanics:
This can work alone, or in conjunction with a randomized A.I. system, as a modifier. The more killing, the greater this modifier would be, resulting in a steady increase of militant A.I. numbers, perception, intelligence, and physical capability (relative to what is implemented, and relative to the fixed difficulty level) near or in secure areas, with a decrease of unarmed civilian A.I. in unsecured areas.

Civilian A.I. could arm themselves with daggers or other small weapons, and some could have increased speed for quick attacks or flight. They would also be more likely to warn the nearest guard or the City Watch as the killings increase.

For those of you interested in how a more advanced A.I. might be implemented in T4, I recommend reading (http://forums.eidosgames.com/showthread.php?t=89009) "An A.I. Discussion".

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

Reactions could range from subtle to dramatic, depending on the CTL.

I believe the CTL feature would provide a more logical and believable reaction to the murders/disappearances in the City, and would discourage killing in a natural and immersive way...this is because the player would feel their impact upon the world as a consequence of their behavior...keeping killing and violence in perspective.

Difficulty would become a dynamic and continual reaction to the player within the world. As more violence would invoke more security as a counter-measure to the threat. Excessive blackjacking could also be used to increase the security, although the reaction and modifier to the CTL would surely be less.

This an example of how the CTL feature could be implemented...you need not agree with every detail in order to be for or against the basic concept.

Tension points

This is how tension could be quantified, without using the "instant fail" for killing on "expert" difficulty":

Killing:
Normal = 3, Hard = 5, Expert = 10

Knock outs:
Normal = 1, Hard = 2, Expert = 3

Alternatively, this is how tension could be quantified, when used in conjunction with the "instant fail" for killing on "expert" difficulty:

Killing:
Normal = 5, Hard = 10, Expert = FAIL

Knock outs:
Normal = 1, Hard = 2, Expert = 3

Keep in mind, any prescribed objective that would cause an "instant fail" (such as no blackjacking), would of course supersede this feature.

Threat Levels

The amount of tension points that are generated through violence will influence the behavior of the people of the City when certain thresholds are crossed, thus creating different threat levels, and a reaction to the perceived threat. These levels represent various stages of concern or fear.

Level 0 = 0-9 tension points
This is the normal level of tension in the City, with standard behavior and security measures.

Level 1 = 10-19 tension points
This is a slightly elevated level of concern. The effects are subtle. A little less loot lying around, some locks/traps are a little harder to pick/disarm. Both civilian and militant A.I. have not significantly changed their behavior, but there is a concern in the back of their minds.

Level 2 = 20-29 tension points
There is a considerable level of concern with a nervous tone. The effects are easily noticeable.

At least half of the valuables that were lying around have now been locked up or hidden away. Most of the locks/traps are harder to pick/disarm, and more of them are now in place.

Militant A.I. numbers have increased, and are noticeably more sensitive to strange noises and peculiar shadows.

There are less Civilian A.I. on the streets at night, and are less likely to carry any loot with them. They are also more sensitive, and are more likely to panic when facing an unknown.

Level 3 = 30+ tension points
People have now become scared or extremely nervous, and fear for their safety at night. The effects are obvious and pronounced.

Almost all of the valuables that have been lying around are now locked up or hidden away. All of the locks/traps have become difficult to pick/disarm, with some now unpickable. Practically every portal/container is now locked.

Militant A.I. have now doubled in number, and are staying together more often in pairs or groups. Many of them now also carry torches. Their perceptions have been heightened, and are now are on an alert status by default. Melee units are now quicker with their weapons and on foot. There are more ranged units and may even have gas or fire arrows at their disposal.

Civilian A.I. rarely roam the streets at night, and if they do, are most likely armed with a small weapon. Even indoors, many are now armed. Those that are armed may fight or flee, doing so with increased speed...and they will likely flee to the nearest Militant A.I.

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

The CTL system could be used to seamlessly trigger events in many different ways, including auxiliary story events and may even be used to dynamically alter or add fixed difficulty objectives, if desired.

Multiple CTL's could also be used, and they could be made to be sympathetic to each other, depending on preferred and logical design.

The CTL feature would enhance and dimensionalize the core design of THIEF, creating an even more immersive and dynamically responsive experience.

Chade on 8/1/2010 at 00:49

I didn't vote in the poll because the available options are really quite specific, and none were a good fit for me.

However: I like the idea of dynamic difficulty in thief. One thing I don't like about your suggestion is that risk and reward are inverely related: the game gets harder as the amount of loot goes down, and vis versa. And it makes the game harder for newbies then experts. This is the opposite of what we want!

I suggest we separate factors which make the game harder from factors which make the game more rewarding. Adding guards / increasing alertness makes the game harder. Removing loot makes the game less rewarding.

So punish killing by removing loot (i.e. your scared civilian mechanic). Newbies get less loot then experts. Punish success (in this case, stealing lots of loot) by increasing the amount of guards on the street. Experts have to handle harder levels then newbies.

Note the feedback mechanics here: ghosters who do not find a lot of loot will start tripping over unguarded loot sooner or later, leading to more heavily guarded levels. On the other hand, people who kill everyone in a level and then pick up all the loot will start to scare aware civilains, leading to leaner pickings untill they stop killing. This is all desirable IMO.

Bakerman on 8/1/2010 at 00:57

What Chade said. I like the idea of dynamically adjusting the difficulty, but the idea of the 'no killing' rules is that it's difficult to do. New players will be more likely to rely on killing guards, and if you then make the game more difficult for them it'll turn them off (or force them to kill more and more guards until they've maxed out the difficulty). It's a feedback loop that goes the wrong direction - it unbalances things instead of keeping the balance based on the player's skill level.

Vae on 8/1/2010 at 04:42

Quote Posted by Chade

One thing I don't like about your suggestion is that risk and reward are inverely related: the game gets harder as the amount of loot goes down, and vis versa. And it makes the game harder for newbies then experts. This is the opposite of what we want!

You are confused in your assumptions regarding the effects of CTL, thereby not understanding the balance.

Quote:

1) Less available unsecured loot

People are simply nervous about carrying or leaving valuables in unsecured areas, and are either locking them up and/or hiding them better, for fear of getting killed because they own them.

Mechanics:
This can work alone, or be used in conjunction with a randomized loot placement/amount system, and used as a modifier. The more killing, the greater this modifier would be, reducing available unsecured loot.

There need not be any less total loot. As I stated, there would be less available unsecured loot, not less available total loot. This means that more valuables would be locked in secure areas/containers or hidden better. This would be accomplished with loot placeholders that would accommodate this dynamic shift between secure/unsecured locations/containers and between open/hidden locations. The loot would not just disappear *poof*...:laff: ...However, a total loot reduction could be scaled with difficulty (e.g. Normal = 100%, Hard = 90%, Expert = 80% total available loot), if desired.

The effect of the CTL is scaled with the fixed difficulties in every regard, and would work in conjunction and in proportion with them, creating a balanced, dimensional, and synergistic difficulty system.

Are we clear on that now?...:)

Quote Posted by Bakerman

I like the idea of dynamically adjusting the difficulty, but the idea of the 'no killing' rules is that it's difficult to do. New players will be more likely to rely on killing guards, and if you then make the game more difficult for them it'll turn them off (or force them to kill more and more guards until they've maxed out the difficulty). It's a feedback loop that goes the wrong direction - it unbalances things instead of keeping the balance based on the player's skill level.

The CTL feature doesn't prevent killing, it just discourages excessive killing/violence for the expert and newb alike. This is scaled with the fixed difficulties to maintain balance.

Quote:

Tension points

Killing:
Normal = 3, Hard = 5, Expert = 10

Knock outs:
Normal = 1, Hard = 2, Expert = 3

This is just an example of the concept, and these values could be adjusted for more precise balance. Even if adjusted, the concept remains the same.

For example this could be:

Tension points

Killing:
Normal = 1, Hard = 3, Expert = 5 or FAIL

Knock outs:
Normal = 0, Hard = 1, Expert = 2

Even a newb doesn't need to go on a killing spree in order to complete a mission, but if he does, he will feel the natural resistance of City, just like everybody else...albeit to lesser degree, because of scaled balance...:)

In the future, I would appreciate a request for clarification before jumping to conclusions...thanks...:)

Chade on 8/1/2010 at 05:29

IMO you're not seeing the forest for the trees. The right feedback system won't require lots of little patches so that it works.

The difference between securing and removing loot is irrelevant detail. The fundamental concept here is reducing or increasing the amount of loot the player is expected to find: this can then be implemented in any appropriate way.

Delaying the punitive effects of CTL depending on level is a poor solution to the difficulty problem. It doesn't remove the undesirable feedback loop, it merely obscures it. Now you have the problem of balancing two contradictory game mechanics: one which punishes killing, and one which explicitly allows it! Far better to get the right feedback loops in the first place.

Vae on 8/1/2010 at 07:24

Quote Posted by Chade

IMO you're not seeing the forest for the trees. The right feedback system won't require lots of little patches so that it works.

Lots of little patches?

Quote:

The difference between securing and removing loot is irrelevant detail. The fundamental concept here is reducing or increasing the amount of loot the player is expected to find: this can then be implemented in any appropriate way.

If you are referring to the CTL feature, that is not the fundamental concept. If you are referring to another concept, then I am having a hard time understanding it because of how you are phrasing your responses.

Quote:

Delaying the punitive effects of CTL depending on level is a poor solution to the difficulty problem. It doesn't remove the undesirable feedback loop, it merely obscures it. Now you have the problem of balancing two contradictory game mechanics: one which punishes killing, and one which explicitly allows it! Far better to get the right feedback loops in the first place.

There would never be an undesirable feedback loop unless you wanted it that way. One would purposely have to jack up the tension point values...this is easily avoided. Remember, the CTL decreases as well as increases.

There isn't any contradictory game mechanics. Your mind only believes that because you mistakenly perceive the "allowance of killing" as the equal and opposite force of the CTL feature. The "allowance of killing" is a fixed negative absolute, and the CTL feature is a relative dynamic positive.

Because they are not contradictory absolutes (because CTL allows killing), and because the CTL feature is dynamic, it operates in harmony with it. There is no problem with balance, as this can easily be fine tuned for optimum results...just like the fixed difficulty system.

Briareos H on 8/1/2010 at 08:16

I don't think CTL should be related to no-kill objectives, as the latter are to me a strict rule which should be enforced throughout the game. You kill someone, you're dumb, you lose. If you want to kill people, then I'm all for a lot more custom and different difficulty options at the start of a mission, but make it static:

The first fundamental problem with CTL is straying away from the fact that Thief's core rule sets are strict and linear. At any moment, an experienced player knows what to expect from almost every variable of a mission, and this contributes to his enjoyment. Things can go a little awry, but because things going a little awry is a possibility made immutable, exhaustively understood from an AI behaviour perspective.
Dynamic changing of core simulation parameters is NOT a good idea, because a player's immersion in the game world relies too much on unharmed expectations.

The second fundamental problem with CTL is that it really only applies to free-roaming city sections and damn me if those weren't the most tedious and horrid thing in TDS. I'd be happy to see them gone.

jtr7 on 8/1/2010 at 08:47

It's actually far simpler, and elegant in the Thief tradition. It would be a couple of toggles and a small set of variables per mission, set by the level designers according to the fiction--not...patches (?). And since all the games have missions near each other as well as far, and there are missions close in time as well as a long time apart, the fiction would have the ultimate say. Just as even the missions themselves had consequences shaping the missions, so would this. It wouldn't work separate from the fiction but as an enhancement.

Also, those few variables would automatically handle changes in difficulty levels. The stats systems are already in place, and those numbers would become half of the variables the other numbers and settings would all modify the outcome.

It's not as radical or fundamentally deviating as it appears, but quite simple, but effective for immersion and realism, and it takes into account the obvious. A murder streak in a Hammer mission won't cause problems in a Pagan mission following right after it. In fact, there could be signs of celebration if the Pagans caught word of it--and no, I'm not even remotely alluding to a Faction Alliance Status. Screw that crap. :p

TDS already has half of the system in place behind-the-scenes, but not used in-game the way Vae proposes. Depending on the players' actions (the stats used to fill out the city district crime reports posted in-game), the TDS fences would comment on the player's job.

Heartless Perry might say:

Quote:

f11_02kill1: Say, I heard about the job you did at Rutherford Castle. Word on the street is you killed a whole lot of them guards up there. I always thought you preferred...you know...get in get out...no one even knows you're there--‘cept when they see their money's gone. But I guess you had to do what you had to do...to get the Opal I mean.
f11_02ko1: There's some talk on the street about the job you did at Rutherford Castle. Heard half the place woke up the next morning...and not near enough bandages to go around for all their headaches. You forget how to sneak past a guard, Garrett? (Laughs) Anyway, guess they should be grateful you didn't slaughter ‘em, eh?
f11_02seen1: You know, a couple of guards from Rutherford Castle are bragging they got a piece of you, Garrett. Well...it's nothing compared to what you got from them. It's who's got the goods in the end that counts, eh? Anyway, if guards were giving me trouble...I'd sneak up behind ‘em with my blackjack and...ah, why am I telling you that? You're the expert.
f11_02unseen1: Hey, I heard about the job you pulled at Rutherford Castle. Sounds like you had them running around in circles. And a bunch of ‘em laid off this morning--seems they didn't even know a thief was there and getting away with the family heirloom to boot. Warms my heart to hear about such a stealthy thievin' job.
f11_03kill1: Word on the street is...maybe you got something against the Pagans. Left so many of ‘em dead. Didn't sound like your usual M.O., you know...so stealthy you hardly leave a trace. You got guts, Garrett. I'll give you that. I mean, if it were me, I wouldn't want the Pagans mad at me, if you know what I mean.
f11_03ko1: Word on the street is...a whole mess of Pagans woke up this morning with bumps on their heads. Your handiwork, eh? Are Pagans too hard to sneak past or something? Anyway, I hear they got herby remedies for headaches and such, they'll be all right. Unless you stole their herbies too? (Laughs).
f11_03seen1: Word on the street is...you let a few Pagans get too close last night. Surprised they left you alive. Heard they's real vicious. Well, all that matters is you made it out...relatively unharmed, eh? Maybe next time you sneak up behind ‘em...womp ‘em over the head before they even know even you're there. Not that I need to tell you...Mr. Master Thief.
f11_03unseen1: Heard some talk on the street. Pagans got ripped off. Most of ‘em didn't even know anything was wrong 'til sun-up. And I said to myself...only one thief I know has the guts to take on the Pagans...get ‘em chasing after their own tails...and walk away with the goods. Had to be you, Garrett. Am I right?

And Black Market Bertha might say:

Quote:

f12_04kill1: I heard about your handiwork at that Hammer Complex, Saint Edgars...all those Hammers...murdered. I didn't figure you for the killing type. Na, figured you were the go in without anyone knowin' you're there type. Not my beewax either way. Maybe they deserved it...being in the possession of some trinket you wanted, eh? That'll teach 'em.
f12_04ko1: Guess you broke into St. Edgar's, huh? Heard a bunch of those Hammer fellas got knocked out...woke up next morning...none too happy. Thought it sounded like something you'd do. Why kill 'em, when you can knock 'em out just as easy? Or sneak past 'em, even better. Am I right? Not that it's any of my business....
f12_04seen1: You let a couple of Hammerites get a little too close at that Hammer Complex? That's what I hear anyway. Not that it concerns me any. But, they're claiming they actually got a piece of you. Well, you obviously got away in the end. And that's what counts, ain't it? You oughta try sneaking up behind those fellas next time...knock 'em out maybe.
f12_04unseen1: In the mood for some gossip? It's about you. Word is someone had those Hammers over at Saint Edgar's running around after their own shadows. Hardly knew someone had broken in. You're work, eh, G? Not that it's any of my business. I'll tell you this though, since I met you I'll never look at a shadow the same way.

So...for this CTL system, instead of hearing about it, or just hearing and reading about it as before, we would see it and experience it when it makes sense to, at a level according to the missions' settings, determined by the devs' knowledge of time, place, and Faction in relation to the previous mission. Adding in modifiers that work off the stats, or nullify them as needed, there is no requirement to record new voice-overs, or create new readables, and they can be tuned to serve the fiction, and the fiction can be made with the system in mind.

Beleg Cúthalion on 8/1/2010 at 12:25

The thing is that this system (which sounds quite pretty although I haven't read all of it) would probably only/merely work in some sort of City HUB or a larger City mission like TDS had. If they give us another one of these, I'd vote for implementing it, but not everyone was happy with these City sections and the usual missions probably won't benefit from the system.

jtr7 on 8/1/2010 at 20:12

No. It's not meant to ramp up and up and up, and having an open-world or hub is not acceptable for the idea behind this. It would only apply to missions close in time and proximity, which all the games had, but only half the time (TDS excluded). It's for those missions that are back-to-back or in adjacent districts, returns to the same districts later, and there would still be signs left over like if someone installed a security device for protection and never had it uninstalled. It eliminates going crazy in one mission and having no sign of it in the next, smoothing over continuity. Ramping up too often would create the nightmare of playtesting issues. The story provides natural caps on when and how much where the auto-mission-fail doesn't already cover it.

By the way, how does the Paranoia "system" work in TDS, that I've stumbled across in T3Ed, if it's even a system?