Interesting question. What after a minute of thinking became quite obvious to me, is the strong link to "games in a classical sense"... you'd want a game which is simple enough for everyone to quickly get into, yet which is difficult to master. Who would have thought that greed can unintentionally result in something nice

Unless i'm missing something, i think one can summarize it like this:

For difficulty adaption to be non-annoying, it has to be at least one of the following:

1. Be part of the "reward" of playing the game - similiar to how levels get harder if you finish a level.
2. Be compensated with other rewards which make it worth it, so that the player actually wants the difficulty to increase. This may be the same as 1., just differently phrased, not sure.
3. be irrelevant to the "reward"-mechanics
4. Not actually being a difficulty in-/decrease, but just a modification in "how" the difficulty is created (i.e., enemies adapt to player tactics all the time).
5. Be player-requested (Example: Pirates! offers the player - after a very successful voyage - to change the difficulty level. Its up to the player to accept or not)

Really? Eqiupment is slightly randomised, although within such tight boundaries that you can pretty reliably tell (say) a Nord Footman from a Nord Warrior just by looking at them. Are stats random? Anyway, apart from that I think it does quite a bit of adjustment of damage done per hit etc, but the difficulty of a battle is really determined by the numbers and types of troops, and the situation.

Regarding troops, MnB really is similiar to the hybrid approach.

Troop stats work this way: In the troop definitions, only level 1 stats are defined. Then the troop gets leveled up to the defined level, by randomly assigning stats. But because of the starting stats, there is of course a tendency.

Troop equipment works this way: Every troop gets assigned about 3 times as much items as it will actually use. Then for each item type, the troop randomly picks between 2-4 items. It is possible to assign guarantees - i.e. "always mounted", "always has shield", "always has ranged weapons", etc.

Items use fixed templates, but then alow minor random variation (quite strict - i.e., one armor point more or less).

The battlefield as well (at least in open terrain) gets procedurally generated.

So, yes - of course there is lots of shaping going on, but that doesn't explain why with all the minor variation in so many places, it still remains kinda predictable regarding difficulty. One thing which may matter, is that the player has a lot of room to avoid potentially difficult situations - i.e., on a battlefield there is a lot of room to maneuvre. A topdown shooter for example doesn't have that - the margin between "too easy to avoid collision" and "unfair bullet hell" is quite small. So, i guess how "forgiving" a certain game-setup is towards difficulty spikes may be important.

Example: The lance in MnB is the only "instakill" weapon. Now, when the player faces a group of lancers riding frontally at him, thats a pretty hard spike which is difficult to survive - but: He can simply avoid it by turning and then approaching them non-frontally.

Interesting - in that case i seem to have played the "right" roguelikes (i prefer RLs which are a bit easier, because the difficulty of many roguelikes constantly keeps the player at the edge of the seat, so that every minor mistake can result... erm, perhaps thats the explanation?)

EDIT: Okay, i thought a bit further about this, and the problem of spikiness to seems to be related to all factors which aren't predicted and shaped. Often, those are things which depend on interaction and combination. Example: Combination of weapons. Combination of movements. Combination of formations, and so on.

However, there is an example which surprises me, because normally, it should strongly suffer from that: Mount and Blade battles (i mean only the battles themselves). Reason is this: All troop styles in MnB get significantly randomized. Character stats are different in every game, per troop (so, troop abilities arent even constant between games). Items get randomized on a per unit basis. And then you fight those on a 3d battlefield. How then does MnB manage to keep battle difficulty predictable?

I've never played a roguelike that wasn't too difficult for me, though.

This however is a design-choice by roguelikes, not a game-mechanic limitation - roguelikes do have a difficulty curve which follows very reliably what is intended by the designers. "Out-of-deep monsters" for example are no accident, but actually specifically requested in the game code, because otherwise, it wouldn't happen.

P.S.: Also, "target audience" really has nothing to do with the topic.

P.S. 2: There however is a problem with difficulty, when one starts to make stuff random, which's difficulty cannot reliably be rated. For example: Movement patterns.

@Paul:
Roguelikes beat your empiricism-obsession :-) However, roguelikes DID spend a lot of effort figuring this stuff out (that is, they didn't follow your advice : )

Also, yes - definatelly first go for rapid prototyping.

Not necessarily. For example, if you "rate" all abilities of NPCs with points, and then in lower levels give enemies less points than in later levels. It certainly cannot be as finely tuned as manual content, but if the data and code is there, you can shape the difficulty significantly.

What i consider a much more significant issue, is on how many layers one needs to take both, the generation, and the difficulty, into account. For example: who cares that your spaceships are pretty, as long as the formations and "rhythm" of the game is like a roulette?

What it boils down to is: When one does this approach, it isn't just a case of generating content - you also need to "teach" the game the gameplay mechanics.

My generic hint is: Don't do purely algorithmic content generation, except for things where it is efficient (i.e., generate a maze). For things which are not straight-forwardly defineable with an algorithm, at least the following options apply:

1. Do it algorithmically anyways. Your content will suck. You will get a very high amount of content. Effort will be low.

2. Do manual static content. Content quality will be very high. Variety will be low-medium. Effort will be high-very high.

3. Do a hybrid approach: Basically, manually create small unique content-parts, then write an algo which uses this pool of content-parts to build pseudo-randomized content. Your content will be medium-high quality. You will get high variety. Effort will be high-very high (however, the pay-off will get better the bigger your content-parts pool gets).

P.S.: How to get meaningful content with the hybrid-approach: This strongly depends on what you want to do. However, very often, it boils down to this: All your contentparts needs to be "tagged" with all kinds of classifications. Then, write content-templates, which use the content-pool in a selective way (basically, a bit like a "fill-in-the-blanks" puzzle). So, you basically manually specify "classes" of content, but allow selective randomization in the details.

Example: You do write a template for a "battlecruiser class" spaceship. You define that defenses should get higher weight and cargo capacity, engines and agility should get low weight. Furthermore you define that there should be no more then 4 times the same weapon. Then you leave the rest up to randomization.

I fully agree with this. However, i dont see how this would run contrary to allowing choices beyond "pro" and "contra".

Even if we stay with interpersonal relationships, there can be advantages and disadvantages to every option - even options which seem to always be the most desirable choice.

Example: The player can towards an NPC play an aggressive role, a submissive role, a mutualistic cooperative role, or he can seek to distance himself from the NPC. Theoretically, the most optimal choice for both would be to cooperate mutually (most advantageous to both). So does that automatically make this a perfect choice? Nope, because playing that role can be abused by the NPC (Betrayal) - or depending on the NPC personality, he may simply envy the Player for showing leadership-style strenth and therefore see him as a competitor instead. Or perhaps the NPC is shy and first needs some distance, before approaching again - there are many reasons why that apparent "optimal" choice, may not actually be optimal in every situation.

And please, dont tell me "we dont need roles". This is BS. There are ALWAYS roles, even if you dont label them as that. You always have things like "cooperate", "reject", "yes", "no", and so on... thats exactly what i'm talking about. Even if in an FPS you run away, attack, defend, cooperate... same thing. The only difference is that most games artificially restrict the choices to the popular "pro"/"anti" roles, even though logically speaking, there are ALWAYS at minimum those four choices available, unless they get artificially restricted. If the player were given free choice, he could always try to run away, always try to dominate, always be submissive and always try to mutually cooperate. Its just that most games often dont allow all those options, even though there is no explanation why the player cannot choose them (except of "the author doesn't want you to, because his game concept cannot deal with it.").

Those were just some examples and there are probably many more. You just need to think outside of the box of techniques, which you learned from static gameworlds.

Indeed, i just noticed another approach which is so simple that its ridiculous: The author doesn't even need to act as an invisible god most of the time. All he needs, is a sockpuppet which has the capabilities which the author requires. When a plot waypoint is triggered, all he needs to do is modify the AI of his sockpuppets to do what he wants, similiar to how an RP Gamemaster can use the abilities of NPCs for just that purpose.

My first thought was "strangely hypnotic".

I do like that the tone generation is not rasterized - just the clipping is distracting... oh and... when i accidentelly made it go up offscreen infinitely, i closed the page out of fear that it may fry my speakers

Personally, i can imagine this being a "strange" experience if it had changing backgrounds, some things to interact with, and changing sounds... but then, it may become illegal

Okay, apologies if this is a little bit rambling:

The reason I'm interested in strategy is because I wanted to look seriously into reasons WHY you might want to model complex social interactions in a game system. If it's simply to provide NPC agents that don't break immersion, then I have some serious reservations about that approach. And if you just want to model a system just to see what it does, and let the player poke at it, I guess that's fine although it seems more like a programming exercise than something that would necessarily be interesting to play with. And if you're NOT playing with it, it might as well NOT be a full-blown interactive element in your game.

I mean, it's all very well creating a complex conversation model, but what if the player finds an optimum strategy through that system that leads predictably to a win state?

I dont really see the problem. If you want a win-state, create a win-state, if you want no win-state, create no win-state. If you want multiple win-state, then create multiple win-state. If you want to leave it up to the player what HE/SHE considers a win-state, then give allow for multiple outcomes. Where's the problem again?

Furthermore, even though the majority of people may want to have "the one goal", that is not necessarily the case for everyone. To someone, the process of playing the game, may be the reward, and i doubt that those consider it just a "programming excercice", as long as the process of playing it is interesting.

To be frankly honest, you seem to be simply obsessed, that if there is no one predefined goal "given" to reach, and if there is no "competition", then its pointless. Nope - pointless is, when you create a game which's only potential appeal is competition, yet it offers no competition. This is pointless - not because of the lack of competetition, but because then there is nothing else left which your game has to offer!

EDIT: As for your concern about a "perfect strategy" - the only way to prevent that, is via randomness. Every mechanic which always works the same, is by definition predictable. If randomness is in, then it is just a matter of balancing the various options which the player and AI Opponent has. Still: Even this will not prevent that there will be a "perfect meta-strategy", unless the randomness strongly changes the game-balance every time. However, if you do not want the game to be trial and error, then you actually want the player - inside a single game - to come near an optimal strategy, because if you wouldn't allow that (i.e. constant randomization), then there would be no point to developing a strategy at all.

But a dynamic system that can be predictably stabilised isn't a very interesting dynamic system.

You're again asuming a single win-state and a "challenge" to reach just that.

You want your system to be poised at the point of complexity, halfway between stasis and chaos. So we need to consider the system in terms of strategic balance in order to build something capable of maintaining that unstable equilibrium in an interesting way. That's why I think in terms of opposing forces - creating the right mix of positive and negative feedback to create an unstable (and therefore interesting) game state.

If one only has experience with a certain mindset, then this is true. Outside of that - not necessarily. Not to create a misunderstanding here: I'm not saying that gameplay based on opposition is "bad" - i'm saying that your implication that only oppositional dynamics can be interesting, is a prejudice on your side.

Besides, every meaningful decision you make in life is a competition between opposing forces.

You're assuming that everyone thinks as you do. For example, there are people who like to "solve" things, rather than "beat them" :-)

Ideally, what you want is a system where the player is making choices that have the possibility to effect the outcome, both positively and negatively, in a way that requires proper consideration, rather than applying rote responses.

How do you know for so certain, that it is clear in every possible situation what the player considers "desirable", hmm? Are you assuming that choice is only there to give the player a puzzle to solve, not to let the player decide what he wants?

The rest of your "rambling" is more of such assumptions. If you want to create a game with a certain intention, and you want to do this as efficiently as possible, then you're free to discuss how to do that - but if you claim that it is the only possible intention why one would play a game, then you're definatelly trying to enforce your preferences on others.

If one is satisfied that player-taste in most games doesnt matter much - if one is satisfied that most games have just one "goal" to achieve - if one is satisfied that in most games, it is always the same battle between "good" and "bad" - if one is satisfied that most games have little to offer besides of the carefully precrafted path - then yes, of course there is no point to dynamics at all!

- Lyx

This is a bit like a "super-topic" targeting an issue, which covers both, the "Linear Stories vs Interactive Storytelling"-Thread (http://forums.tigsource.com/index.php?topic=6203.0) as well as the "A ludologically constructed conversation game"-Thread (http://forums.tigsource.com/index.php?topic=6242.0).

Designing a dynamic gameworld is a bit like creating a simulated world (with the scope depending on the game). So, you create all kinds of entities in an environment, define how things are connected, give those entities "intentions", and things start moving...

...and then? Okay, you maybe created an interesting explorable environment, but the game has no "point" besides of that. There is no macroscopic "progress" and "targeted movement" in the gameworld. Now, hardcore prodecural game proponents will probably argue that "an invisible god" (the author pulling threads) doesn't belong into such a game.

Why not? What certainly does not belong there, is an author enforcing directions the hard way with a totalitarian hands-on mentality, true. What also doesn't belong there is an author playing spellcaster, by making things appear out of thin air in a way, which breaks the carefully balanced environment and its mechanics, true. "Creation" and "Direct intervention" IMO is off the table for such a game.

But are those the only ways to introduce "guidance" in a gameworld? I dont think so. There are a few ways, how one can still let the author steer the plot in a harmonic way - and even in a linear (!!!) style.

A few examples:
- Biases. These environments work via mechanics - if the balance is shifted in one way, it will slowly tend into a certain direction.

- Plot Waypoints / Chapters. You can define certain states in the gameworld which trigger a transition. A transition to what? Well thats up to you! You can make the actors move to a different scene. You can make a significant modification to a part of the gameworld (attention: do not inject resources (no "casting out of thin air"), but rather transform already existing "substance"). There are many possibilities.

- You can combine the above two techniques. Imagine a flat sheet balanced on top of a ball. On the sheet are the actors in the gameworld. Now, you can make it so that movements on that sheet, tilt the balance in an escalating way - and for each direction in which the balance may escalate, you have a "plot waypoint" defined, which introduces an important transition of the gameworld. This basically would be close to the "branching" approach in traditional static worlds.

Those were just some examples and there are probably many more. You just need to think outside of the box of techniques, which you learned from static gameworlds.


EDIT:

When thinking about what kind of external (author) influence is acceptable, two major aspects seem to be significant:

1. If possible at all, resist the temptation to cast via magic. Do not enforce the changes which you wish "on top of the gameworld mechanics". Rather, use the gameworld mechanics to make the entities by themselves behave as you wish. In other words: Your modifications should be compatible with the gameworld mechanics, instead of being conflicting with them. Do not cast new things, shape existing things.

2. The player should not feel betrayed in his ability to decide and those decisions to have weight (consequences). I could now go into all kinds of detail, but the solution really is simple: Do not enforce you wishes on the player. Accept the decisions of the player and cooperate with him. Do not act like an author - act like a pen and paper roleplaying gamemaster! It really all comes down to that the player should not feel that you ignore his interests and strip him of his own significance.

Take into account that there are many one-way streets with which the player would be perfectly fine. For example, no player will complain that in a scene it starts to rain - and yet, the consequence is that this way you make sure that the player seeks shelter in nearby buildings - you just forced a plot transition without making it feel enforced - especially not if the player even gets options where to seek shelter

P.S.: Since you do seem to be interested in having some kind of "strategy" and stuff, it may be interesting to point out that two-player games in game-theory (NOT videogame theory) follow the same schema - i.e. "win-lose", "win-win", "lose-lose", etc. In other words, this really is a scheme which from a programming POV allows to make a lot of connections which the AI can analyze. It's just plain first-order logic - the "datastructure" is compatible with all kinds of things.

I'd like to add something regarding the "opposing forces" idea in the beginning of the first post.

I do not think that making the playfield simply "polar" is interesting and satisfying. It certainly is easy to understand, because humans of this culture (sadly) are so familiar with it. But it really "isn't as good as everyone says it is"

It's main problem is that it only understands opposition. Oh, sure - it allows an abusive relationship, and it allows situations in which both forces are "submissive" (which then strangely many consider as "friendly" or even "mutualistic") - it doesn't however change, that plain oppositional polarities only know war between forces, and thats it.

Think about it: You have all those games which try to give you "choices" - most annoyingly that "moral choice" crap, in which all you can do is choose between "pro" and "contra", between being "nice" and being "mean", being "egoistic" or "altruistic"... be honest, hasn't this always felt a bit primitive and naive?

I'll try to give you a different perspective, so that you can "see" the problems of the polar approach from a different angle. Some of you probably once heard about the 2D "Territorial map" which is sometimes used in communication psychology, right? If you dont remember or know it, its like this: We have two agents. Every agent gets assigned a value (so that you have two). That way, even if we stay at binary logic, you get four choices - the well known "I'm good,  you're bad", "I'm bad, you're good", "We're both bad" and "We're both good":

Now, the point here is: This isn't actually about psychology! Sadly, not many people understand this. It really is quite simple: You have a statement which consists of two values, which can each be positive and negative. Still doesn't ring a bell? How about this: A but NOT B, B but NOT A, A AND B, A NOR B.

This is something which applies to a lot of things. You can basically apply it whenever two aspects are relevant to a "rating". When are at least two aspects relevant? Well, basically in any kind of relationship. When i say relationship here, i do not just mean a personal relationship between humans, but any kind of relationship.

To give you a really primitive example in a game: The player may choose to reject/fight something, he may choose to accept/integrate into something, he may choose a free "alliance" in which both still retain their autonomy, or he may choose to simply avoid/noninteract with it. Doesn't this kind of choices sound a bit more interesting, than plain "friend/foe", "good/bad", etc. choices?

- Lyx

LOL

Dont say you were the one commenting on the emily short article on the frontpage? :-P

Okay, one more post, because i just noticed something significant. Basically, i'm going to go full-circle back to my first post in this thread.

In that post, i pointed out that most linear stories contain no mechanics, no "why". The story may explain the why, but if it happens in a program, the program actually understands nothing about whats going on. Thus, the program mechanics have no clue how to react, if stuff changes. The programmer basically has to constantly hold the hand of the program, and tell it how to deal with every single difference.

Later in that post, i mentioned that in my current project, i make it so that the decision-logic of the AI often (though, not always) does not depend on past events, but instead on the state of the involved characters. Now, the important difference between "event" and "state" in that regard is, that the app doesnt really understand an event - but it can understand and do maths with its state. So, in the case of events, you'd need to deal with every possible combination in the game as a unique exception - you again need to hold the app's hand - with states, you dont, because you can apply a generic decision logic.

Notice a pattern?

Now, with the talk about ecosystems, quests, consequences, plot and so on, i think we are dealing with a similiar pattern. Most current games have nearly no accountability at all. Things appear out of thin air and there is magic everywhere (why does that remind me about modern physics?). What's happening? Well, the programmer again is holding the app's hands, and tells it how to cast the right spells.

Hmmhmm. I wonder what would happen if you wouldn't start "spawning" things out of thin air, to begin with? I wonder what would happen, if instead you defined a limited set of resources for the entire gameworld, then define which products can be created from which resources. Interesting thing here is: You may skip steps via magic - as long as stuff doesnt appear out of thin air (no resources needed). I wonder what would happen if you wouldn't just "cast" stories of "loved ones" out of thin air, but instead programmed right into the game logic that NPCs may fall in love with each other. Furthermore, i wonder what would happen, if you then add need driven tasks to the world, so that things start moving.... hmmmm....

- Lyx

Okay, a bit more about consequences and ecosystems before i leave for today.

Lets first look at consequences a bit more. The link between consequences and story/guidance was mentioned before already. If the actions of the player are his story, then this is only possible if his actions have consequences. The first problem with this i guess is how to implement that. My prefered answer to that is an ecosystem. Not necessarily only in terms of resources. For example, if you pirate enough, the market and therefore its shopkeepers may become more poor. The issue is how to stop the balance of the system from going totally wacko.

This issue i think has at least two aspects. The first is which implementations there are at all to avoid problematic imbalances - so, the means to do it. The second aspect is how to do that without making the player again feel that his actions have no weight.  If every action immediatelly gets "repaired" someway, then it is quite obvious to the player that basically nothing changed.

Both aspects, as i will show, however share the same solutions. Regarding implementation, what IMO is needed is a need-driven ecosystem. If the balance is tilted one way, intentions to deal with it need to arise. If we stay with the quest-system, then the answer is quite simple: Create counter-quests and let not only the player do quests. Rather, create some kind of "job-market" in which the player as well as NPCs can take part, including the possibility of jobs interfering with each other (escort caravan / pirate the caravan).

However, this again brings up the "i didnt achieve anything" issue if the counterreaction is totalitarian. Depending on what is desired, there are multiple ways to fix that. One is to make counterreactions "soft" and "delayed" - so, allow inbalance, but after a while create counterreactions which not necessarily nullify the players actions, but which prevent a total collapse.

Okay, this way, we now do allow the player to affect the world temporarily. But no matter what he does, his actions still have no longterm consequences. Yet, it should also be obvious that we simply shouldn't allow a total collapse, because that basically would break the gameworld. By the way: Philosophically speaking, this is also the case with the world around you. You cannot make it collapse. You can create more or less longterm LOCALISED damage, but in the really really longterm, you cannot break it.

There is a reason while i capitalised the word "localised" above. What if we do allow localised collapse? What if we do allow the player to "break" parts of the gameworld, but not the entire gameworld? I guess the sceptical question to that will be "how can you prevent one from leading to the other"? If the player can break one location, why not the rest?

Simple: The same way as described before: Regeneration. Or even better TRANSFORMATION. Trivia: If the player destroyed a city, and then at a later time in the game, at that location something different would be built - would the player still consider his actions as having had permanent consequences? :-)

So my answer to this issue is: The player can shape/transform the landscape, but he cannot render it permanently unusable. He also in the longterm cannot stop the gameworld as a whole from making sure that it can sustain itself.

As i see it, there is one issue left with this. This can put the player into the role of a cheating bigass manipulator. The player may no longer consider himself living IN the gameworld... no longer being "immersed" in the gameworld - but instead simply look at the gameworld as a series of meaningless variables, ready to be manipulated however he wishes.

To understand the solution to this problem, it is necessary to be aware what the "role" of a manipulator is from a psychological POV. The manipulator sees himself as being disconnected/seperate from the things which he manipulates. In other words **he considers himself unaffected by the consequences of his actions**. What i'm hinting at of course is: The player may not escape the consequences of his actions - he must stay "in" the world which he affects, so that what he does, also affects himself (feedback). A simple way to do this is that the players actions depends on the availability of a resource-cycle. If he abuses the resource cycle, then this will also affect his ability to act himself. Example: Player raids convoys via weapon -> Weapon factories reduce production -> Player gets less ammo -> Players ability to raid is reduced.

There are of course many other ways to make sure that the player is affected by his own actions - like a boomerang. The above was just an example to visualize the concept.

I love the example of Mount and Blade. The first dozen hours I spent in that game were magical, I was in this massive world fighting great battles, laying siege to castles, pillaging the countryside with massive warbands, and winning jousting tournaments. I think I stopped playing once I realized that even though the map was huge, the possibility space and significance of the interactions were pretty limited. Beyond trading, which was the only thing I left out above, there isn't much variety to what you do. Mount and Blade turned out to be a few good ideas replicated across the board in a non flexible fashion. Every king is fighting a war that never truly ends. Every king has one usurper that needs your help. I'd say that's a game that failed to live up to my expectations specifically because they avoided having any sense of story or consequences. There are plenty of things to be interacting with but there is no weight to the interactions.

Well, i think here the aspects of "emergent mainstory" and "consequences" are important. Emergence is all about interactive feedback-loops. As i explained earlier with the quest example: Make the quest-result have consequences by modifying variables and aspects in the gameworld. Then make quest-spawning dependend on just those variables and aspects, and you got a feedback-loop. Add enough variables and aspects which influence each other interactively, and you get something which is much more than the sum of its parts. You get a dynamic "living" world.

The downside to this is: This significantly affects gameplay and overall game-balance. So, you need to have a system in place which can deal with this kind of dynamics. You need game-mechanics which can deal with temporary inbalances - just as any organism - material, biological, mental or cultural - have mechanisms to deal with that. Most games haven't. Their game-mechanics depend on stuff remaining statics.

Guess what? Mount and Blade also doesn't. Okay, i guess what i'm saying now will scare a few mount and blade fans. If you play mount and blade, then in many ways it will seem to you as if it has a dynamic ecosystem below it. It doesn't! It's all faked with hardcoded mechanics. The game-logic is cheating all the time, and AI-parties need no food, no money, no upkeep costs, no nothing, to do anything. It also has no limits whatsoever. All the stats which apply to the player? Irrelevant to AI-Players. Even things like AI-Partysizes are kept "believable" via magic - in the betas short before release, there were AI-parties with 1000+ troops, lords with -50000 gold, prisoner trains in the hundreds, and lots of other weirdness.

My point is: This game does not have a true dynamic ecosystem to deal with a truely dynamic ecosystem. It's all fake. And this is why all the things which you do in mount and blade have little consequences - first because the consequences again would require "magic", and second because the gamebalance-mechanics couldn't deal with it.

@Alex:

From:

Hi wanderer.  A terrible thing happened recently!  A <role> of the <race> has recently stolen our only magic <object>.  If you return it to us, we will be very grateful.

To:

"<greeting>. <this-is-way-bad>. <a/the> <dangerlevel> <role> of the <race> <has-stoeled> our <precious> <object>. <reason-why-we-need-it>. <help-us>. <please-return-it> and we will <reward-description>."

And thats just one template. I guess this also makes it more obvious, why previously i portrayed the availablity of a searchable database as so important.