Dragons and Castle-building

[PhoenixPhyre]

A common complaint is that castles should look different based on the existence of large flying predators such as dragons. I want to explore that topic a bit. Here, I'm going to stick to D&D-style dragons (since it's a complaint against D&D settings, most frequently, and this is the RPG forum).

What do we know about dragons?
1. They're large and solitary. Being large, they require significant food input (and thus territory). They're solitary (mostly not social with their own kind), so you don't have rookeries with multiple adults.
2. They're often territorial (with other dragons).
3. They're intelligent (even as whelps), growing more so as they age.
4. They're prideful. It's rare for one to serve another being unless forced.
5. They come in two divisions--metallic and chromatic.
5a. Metallic dragons are at worst stand-offish and arrogant, but are disinclined to raid castles except under extreme provocation.
5b. Chromatics are evil and greedy--the threat to mortals comes mostly from these.
6. The only significant threat to dragons comes from adventuring parties and other dragons.
7. While a dragon is devastating against living things, they're less so against stone structures (or at least no more so than siege artillery).

What does this imply?
Combining 1 and 2 implies that the expected number of dragons that claim any individual point as their territory is less than 1. Either there are no dragons around, or there's 1 (with the rare occasions that you build in the borderlands between two dragons' territory).

Combining 3, 4, and 6 implies that a dragon is rarely going to attack a fortification--raid farms, sure. Raid caravans for loot, sure. The risk of provoking adventurers (or running into a high level wizard), getting injured or worse, driven off in humiliation is not generally worth the reward of attacking a fortified structure. The exceptions are if you personally antagonized that one dragon. Then, and pretty much only then, will it likely come after you wherever you hide. Being intelligent, it's also likely that it will wait and ambush you (or bring minions, or both) when it does.

#5 implies that even if there's a dragon around, he's not always going to be hostile. Even the chromatics are known to be manipulators, rather than savage attackers. If they need to attack a fortification for whatever reason, they're likely to bring minions (kobolds are known for this).

#7 implies that the unique threat from a dragon is mostly to the people, not to the structures.

The effects on castle building
Building your defenses around a rare case (especially if it diminishes their value against the much more common case) is a bad idea. There's trade-offs--building to repel a dragon at a minimum takes resources away from defending against a regular army. In addition, dragons are raiders rather than occupiers. If, worst case, a dragon attacks and levels a keep, you run away and come back once he leaves. Dragons don't lair in above-ground things like that. You can often negotiate with them as well. Thus, building an anti-dragon keep is only rarely going to be a winning proposition.

Building a specialized anti-dragon fort will be useful if
a) there are known, hostile dragons in the area.
b) the fort controls strategic terrain (a pass through the mountains to feeding grounds, for example)
c) there is little threat from mortal armies in the area
d) dragons deviate strongly from the assumptions above.

Conclusions
So to me, the existence or non-existence of dragons makes little difference for castle building unless dragons are a known, very common, very hostile threat. You might have a few more covered courtyards and flexibly mounted heavy arbalests (that can serve to attack aerial creatures).

Side-note: I won't defend the Forgotten Realms over-all--I don't like it for many many reasons. This just isn't a good one to trash it on.

[neriractor]

Building a castle specialized to defend against dragon attackx also runs the risk of inciting a dragon to attack you once they figure out what you are trying to do.

[Max_Killjoy]

A common complaint is that castles should look different based on the existence of large flying predators such as dragons. I want to explore that topic a bit. Here, I'm going to stick to D&D-style dragons (since it's a complaint against D&D settings, most frequently, and this is the RPG forum).

What do we know about dragons?
1. They're large and solitary. Being large, they require significant food input (and thus territory). They're solitary (mostly not social with their own kind), so you don't have rookeries with multiple adults.
2. They're often territorial (with other dragons).
3. They're intelligent (even as whelps), growing more so as they age.
4. They're prideful. It's rare for one to serve another being unless forced.
5. They come in two divisions--metallic and chromatic.
5a. Metallic dragons are at worst stand-offish and arrogant, but are disinclined to raid castles except under extreme provocation.
5b. Chromatics are evil and greedy--the threat to mortals comes mostly from these.
6. The only significant threat to dragons comes from adventuring parties and other dragons.
7. While a dragon is devastating against living things, they're less so against stone structures (or at least no more so than siege artillery).

What does this imply?
Combining 1 and 2 implies that the expected number of dragons that claim any individual point as their territory is less than 1. Either there are no dragons around, or there's 1 (with the rare occasions that you build in the borderlands between two dragons' territory).

Combining 3, 4, and 6 implies that a dragon is rarely going to attack a fortification--raid farms, sure. Raid caravans for loot, sure. The risk of provoking adventurers (or running into a high level wizard), getting injured or worse, driven off in humiliation is not generally worth the reward of attacking a fortified structure. The exceptions are if you personally antagonized that one dragon. Then, and pretty much only then, will it likely come after you wherever you hide. Being intelligent, it's also likely that it will wait and ambush you (or bring minions, or both) when it does.

#5 implies that even if there's a dragon around, he's not always going to be hostile. Even the chromatics are known to be manipulators, rather than savage attackers. If they need to attack a fortification for whatever reason, they're likely to bring minions (kobolds are known for this).

#7 implies that the unique threat from a dragon is mostly to the people, not to the structures.

The effects on castle building
Building your defenses around a rare case (especially if it diminishes their value against the much more common case) is a bad idea. There's trade-offs--building to repel a dragon at a minimum takes resources away from defending against a regular army. In addition, dragons are raiders rather than occupiers. If, worst case, a dragon attacks and levels a keep, you run away and come back once he leaves. Dragons don't lair in above-ground things like that. You can often negotiate with them as well. Thus, building an anti-dragon keep is only rarely going to be a winning proposition.

Building a specialized anti-dragon fort will be useful if
a) there are known, hostile dragons in the area.
b) the fort controls strategic terrain (a pass through the mountains to feeding grounds, for example)
c) there is little threat from mortal armies in the area
d) dragons deviate strongly from the assumptions above.

Conclusions
So to me, the existence or non-existence of dragons makes little difference for castle building unless dragons are a known, very common, very hostile threat. You might have a few more covered courtyards and flexibly mounted heavy arbalests (that can serve to attack aerial creatures).

Side-note: I won't defend the Forgotten Realms over-all--I don't like it for many many reasons. This just isn't a good one to trash it on.


1) Presumes D&D-style dragons specifically.

2) Dragons are just a single example out of many many different threats that are overall common in high-fantasty RPG settings, that archetypical European castles are poor defense against. Fixating on dragons misses the point.

[Lord Torath]

1) Presumes D&D-style dragons specifically.Well, yeah:
I want to explore that topic a bit. Here, I'm going to stick to D&D-style dragons (since it's a complaint against D&D settings, most frequently, and this is the RPG forum).(emphasis mine)

[PhoenixPhyre]

1) Presumes D&D-style dragons specifically.

2) Dragons are just a single example out of many many different threats that are overall common in high-fantasty RPG settings, that archetypical European castles are poor defense against. Fixating on dragons misses the point.

I'm long-winded enough already--I tried to keep it to a single topic. But there are similar objections for all the other ones I've ever heard brought up. And you brought that one up explicitly, so I wanted to respond to it, in detail.

The fact is, that claims of "not making sense" can stem from a lack of knowledge on the one claiming as well as a lack of planning on the writer's part. I see this from students. "That doesn't make sense" they wail because they can only see the surface. Many things that "don't make sense" do actually make sense, once you see the underlying facts. Without those, any logical deduction is going to fail because it starts from faulty premises. And bias.

It's a lot better to simply accept things for what they are. Decry bad writing (DMPCs run amok, lazy plot devices). Have some epistemic humility. Realize that we don't know a lot of what we think we do, and most of what we do know is wrong. Accept good-faith writing with good faith, instead of trying to poke holes. Life's a lot more fun that way, I've found.

Same goes for rules. All rule systems break down when people try to find loopholes. That's because they're written by mortal people who are very imperfect, in imperfect language, to be interpreted by amateurs. Instead of looking to break things (argumentum ad absurdum is a fallacy, the way it's used most of the time), find ways to have fun within the system. If one system doesn't scratch a particular itch, use a different one. Learn to work within the implicit setting restrictions. If in a particular setting, the world hasn't been broken by <exploit X>, that may imply that <exploit X> doesn't work there. Setting overrides mechanics. The rules are for playing the game, not the game for the rules.

...and I'm rambling. Guess I must really be sick. Note: None of this is meant as an attack. More a cry for peace on forums, good faith to commentators everywhere.

[VoxRationis]

I could see a dragon attacking a keep, maybe even moving in afterward, for no other reason than to teach its builders a lesson.

The problem is that in D&D, once a dragon is large enough to look like a proper settlement-assaulting monster, by virtue of the way the game mechanics work, most options other than "hire high-level people to fight it" don't work very well. Well-designed fields of fire are great, but if your archers can't hit the darn thing more than 1/20th of the time, and that 1 in 20 fails to penetrate due to not being with a magic weapon or just not hitting hard enough, they don't really matter. Ballistae on universal joints are great, but they don't have bonuses to hit or anything that ignores armor, so that damage will likely just fail to materialize. Similarly, breath weapon damage climbs so dramatically that unless you built the whole thing out of asbestos, you're going to get the Harrenhal treatment. A young adult red dragon in 3.5 does an average of 55 damage, which means that it does 19 damage/pass to a masonry wall, which has 90 hp. Even if your entire castle is made of thick stone, with no wooden portions (rafters, support beams, rooftops, doors), it's going to melt its way through in 4d4+1 rounds, only 5 of which need to be spent actually in proximity to the wall. I'm not sure about how the math works in 5e, but breath weapon damage is strangely similarly high in AD&D (even though hit point scaling is much lower).

Edit: Well, this speech looks pretty silly coming after the previous post. I had the thing on edit mode for a while before finally hitting 'submit.'

[PhoenixPhyre]

I could see a dragon attacking a keep, maybe even moving in afterward, for no other reason than to teach its builders a lesson.

The problem is that in D&D, once a dragon is large enough to look like a proper settlement-assaulting monster, by virtue of the way the game mechanics work, most options other than "hire high-level people to fight it" don't work very well. Well-designed fields of fire are great, but if your archers can't hit the darn thing more than 1/20th of the time, and that 1 in 20 fails to penetrate due to not being with a magic weapon or just not hitting hard enough, they don't really matter. Ballistae on universal joints are great, but they don't have bonuses to hit or anything that ignores armor, so that damage will likely just fail to materialize. Similarly, breath weapon damage climbs so dramatically that unless you built the whole thing out of asbestos, you're going to get the Harrenhal treatment. A young adult red dragon in 3.5 does an average of 55 damage, which means that it does 19 damage/pass to a masonry wall, which has 90 hp. Even if your entire castle is made of thick stone, with no wooden portions (rafters, support beams, rooftops, doors), it's going to melt its way through in 4d4+1 rounds, only 5 of which need to be spent actually in proximity to the wall. I'm not sure about how the math works in 5e, but breath weapon damage is strangely similarly high in AD&D (even though hit point scaling is much lower).

That's covered by my #6 (the only predators of dragons are adventurers (and possibly high-level wizards)). Assaulting a keep is a good way of getting a bunch of those on your tail.

And in 5e, even a large enough group of guards (cr 1/8) are a threat due to bounded accuracy. A group of archers (CR 2-3) would be murder to a non-legendary dragon. A young red dragon would have serious issues with an army of 500 or so people--he'd murder many of them, but the arrows would bring him down or seriously injure him. Just another case where the issue is taking mechanics as physics crops up--3e had this odd insistence on running the actual world using the same mechanical pieces as the PCs, tying everything to level. It meant that you can't have a master farmer who's no better in combat than a 1st level commoner without some serious optimization. Much better to use the rules as merely a UI for adjudicating PC/world interactions in a game context.

[Max_Killjoy]

That's covered by my #6 (the only predators of dragons are adventurers (and possibly high-level wizards)). Assaulting a keep is a good way of getting a bunch of those on your tail.

And in 5e, even a large enough group of guards (cr 1/8) are a threat due to bounded accuracy. A group of archers (CR 2-3) would be murder to a non-legendary dragon. A young red dragon would have serious issues with an army of 500 or so people--he'd murder many of them, but the arrows would bring him down or seriously injure him. Just another case where the issue is taking mechanics as physics crops up--3e had this odd insistence on running the actual world using the same mechanical pieces as the PCs, tying everything to level. It meant that you can't have a master farmer who's no better in combat than a 1st level commoner without some serious optimization. Much better to use the rules as merely a UI for adjudicating PC/world interactions in a game context.

Better yet to have rules that don't cause those sorts of dissonances in the first place.

[PhoenixPhyre]

Better yet to have rules that don't cause those sorts of dissonances in the first place.

Can't happen. If people are looking for dissonance, they'll find it. No matter what you do.

I call it the Law of Conservation of Annoyance. I came up with it in the context of physical models, but it applies broadly.


All systems, models, or rule sets have a fixed amount of annoyance built in. The only difference is where it shows up. The trick is to find a system where the annoyance is contained somewhere you don't care about.

[Beleriphon]

Conclusions
So to me, the existence or non-existence of dragons makes little difference for castle building unless dragons are a known, very common, very hostile threat. You might have a few more covered courtyards and flexibly mounted heavy arbalests (that can serve to attack aerial creatures).

Side-note: I won't defend the Forgotten Realms over-all--I don't like it for many many reasons. This just isn't a good one to trash it on.

Honestly, I always thought the bigger threat to castles and other static fortifications was aerial cavalry like griffons or the like. A dragon is dangerous sure, but you just hide in the basement like the opening sequence of Skyrim.

[Max_Killjoy]

Can't happen. If people are looking for dissonance, they'll find it. No matter what you do.

I call it the Law of Conservation of Annoyance. I came up with it in the context of physical models, but it applies broadly.

There's a difference between dissonance you have to go looking for, and dissonance that comes looking for you, slaps you in the face, and then laughs about it.

Level-based systems are the latter, for the very reasons laid out here, such as "if the only way to advance at anything is to gain levels, how do you have experienced farmers and craftspeople who aren't also better combatants than 90% of your soldiers and guards?"

(Not that I agree with the notion that it can't happen -- at its core dissonance isn't about quirky edge-cases and deliberately-contrived problems -- but this doesn't even require that argument.)

[PhoenixPhyre]

There's a difference between dissonance you have to go looking for, and dissonance that comes looking for you, slaps you in the face, and then laughs about it.

Level-based systems are the latter, for the very reasons laid out here, such as "if the only way to advance at anything is to gain levels, how do you have experienced farmers and craftspeople who aren't also better combatants than 90% of your soldiers and guards?"

(Not that I agree with the notion that it can't happen, but this doesn't even require that argument.)

That's only a complaint against certain implementations of such systems, the ones that try to make the game rules cover the entire fiction's metaphysics. I agree that that attempt is bound to fail. Any attempt to make the rules cover things other than PC/world interactions will fail in similar ways.

That's because it's a category error. The game rules exist to create a playable, fun interface between the players and the fiction, often mediated through a DM. That's all they are. Of course there are better interfaces (for a given purpose) and worse ones (again, given a purpose). In my opinion, 3e D&D is one of the worse ones for any purpose I want to use it for--too much baggage and wonky assumptions. For others, it works well (whether they are into the character building metagame or the high power God play or whatever).

[Max_Killjoy]

That's only a complaint against certain implementations of such systems, the ones that try to make the game rules cover the entire fiction's metaphysics. I agree that that attempt is bound to fail. Any attempt to make the rules cover things other than PC/world interactions will fail in similar ways.

That's because it's a category error. The game rules exist to create a playable, fun interface between the players and the fiction, often mediated through a DM. That's all they are. Of course there are better interfaces (for a given purpose) and worse ones (again, given a purpose). In my opinion, 3e D&D is one of the worse ones for any purpose I want to use it for--too much baggage and wonky assumptions. For others, it works well (whether they are into the character building metagame or the high power God play or whatever).

It's not a one-way street.

The rules have to reflect how the (fictional) world actually works and what it feels like for the (fictional) person to live and breathe and move in that world -- AND the world has to work in a way that doesn't constantly contradict the rules we're presented with that tell us how our characters interact with that world and each other.

If we pit heavily experienced and successful warrior against a green soldier right out of training, and the rules tell us that the warrior is ten times more likely to win than the greenie, as opposed to twice as likely or one hundred times as likely, that is supposed to tell us something about the world they live in. It's not just a contrivance for gameplay.

If the people who live in that world are supposed to be capable of about the same sort of physical achievements as real-world people, but the rules result in characters routinely being incapable of clearly 2-foot obstacles or capable of nonchalantly leaping over 10-foot walls, then there's something inherently wrong, and either the rules or the world need to change.

[Yora]

Honestly, I always thought the bigger threat to castles and other static fortifications was aerial cavalry like griffons or the like. A dragon is dangerous sure, but you just hide in the basement like the opening sequence of Skyrim.

If you want air protection, underground bunkers are the way to go. Except for light and the possibility of being smoked out, I don't see any significant drawbacks compared to a regular castle.

(I think I once saw a backstory somewhere of a dwarf stronghold being simply flooded by the besiegers.)

[VoxRationis]

Well, underground bunkers don't really command territory in the way that a castle does. It's way easier to just ignore a bunker entrance than a well-placed castle.

Also, yes, flooding might be a concern. Thus fell Castamere (I swear, I'm not trying to make so many ASOIAF references; they just keep coming up).

[PhoenixPhyre]

Well, underground bunkers don't really command territory in the way that a castle does. It's way easier to just ignore a bunker entrance than a well-placed castle.

Also, yes, flooding might be a concern. Thus fell Castamere (I swear, I'm not trying to make so many ASOIAF references; they just keep coming up).

Right. That's one of the trade-offs I mentioned. Castles are pretty-well optimized for the combination of medieval war-making technology and land-army-based threats. If you instead try to optimize for the (rare) cases of dragon attack, you'll compromise the (more common and generally worse) case of defending against an invading army.


It's not a one-way street.

The rules have to reflect how the (fictional) world actually works and what it feels like for the (fictional) person to live and breathe and move in that world -- AND the world has to work in a way that doesn't constantly contradict the rules we're presented with that tell us how our characters interact with that world and each other.

If we pit heavily experienced and successful warrior against a green soldier right out of training, and the rules tell us that the warrior is ten times more likely to win than the greenie, as opposed to twice as likely or one hundred times as likely, that is supposed to tell us something about the world they live in. It's not just a contrivance for gameplay.

If the people who live in that world are supposed to be capable of about the same sort of physical achievements as real-world people, but the rules result in characters routinely being incapable of clearly 2-foot obstacles or capable of nonchalantly leaping over 10-foot walls, then there's something inherently wrong, and either the rules or the world need to change.

Sure. But those are implementation details, free parameters that can be twiddled to suit the need. At least in well-designed (robust, generalizable or modular) systems. Those have nothing to do with NPCs though--NPCs facing NPCs win based on story concerns, not on mechanics. Part of that story is verisimilitude and immersion. But the NPCs don't have to go through the tedious interface called the game rules. Whatever should happen (either to suit the story or to maintain verisimilitude or whatever) just happens. It's only when PCs get involved that the rules must be invoked, and those are mutable. One must also beware of importing real-life assumptions where they're not explicitly imported by the system. If you impose additional assumptions, you're the one bearing the risk of breakage. Not the system or the setting.

In other terms, each game system has a volume (in abstract space) of optimal uses. Using them for settings or games outside this range causes issues. Some systems (GURPS) have a wide range where their performance is acceptable. But it still falls short in some cases--running a high fantasy game with heroic people doing heroic things is a place I've been told GURPS (with its very deadly combat) is less well suited for. Others have very narrow focus--Exalted or the other WW products are tied way too close to a single setting and game style to make them usable outside that.

5e D&D does what I want it to, for the games I want to run. Not perfectly, but close enough and with enough flexibility that I can override the rough spots as needed. That's because I like the heroic high fantasy genre. Both gritty "death is easy" games and gonzo God-Games are not to my taste. I can see myself playing 4e to scratch the "small-squad tactical combat" itch I get occasionally. 3e? Not a chance. It's a combination of bad (self-contradictory) assumptions, source bloat, and needless crunch. Those are all personal judgments, however.

Part of the problem with FR (and all the other non-Eberron D&D settings) in 3e is that they're not native to the system and carry a large baggage of legacy design assumptions. As a result, the verisimilitude suffers greatly. Those are pretty sacred cows by this time, though, and last time they tried to slaughter them (4e) it cause a serious uproar and (in part) doomed the success of that edition.

[Darth Ultron]

Conclusions
So to me, the existence or non-existence of dragons makes little difference for castle building unless dragons are a known, very common, very hostile threat. You might have a few more covered courtyards and flexibly mounted heavy arbalests (that can serve to attack aerial creatures).


I agree.

You can also add to your list about things known about dragons:

8.Good and even neutral dragons can be counted on to defend and oppose evil dragons. Or more basicaly the metallics oppose the chromatics. Often places in a kingdom or area will be under the, often unknowingly, protection of a good dragon.

9.The general, vague worldview of a dragon sees humans much the same way humans see animals or even bugs. So if some bees make a hive way out in your back yard...well away from your house...most people just tend to let them.

10.As dragons are powerful fantasy magical creatures...the way to oppose them is with fantasy and magic...not mundane things. For some versions of D&D this breaks the idea that ''the whole world must be sad, pathetic, weak commoners'', but once you get past that it works. This would give the average castle some sort of anti dragon weapon or protection.

[PhoenixPhyre]

I agree.

You can also add to your list about things known about dragons:

8.Good and even neutral dragons can be counted on to defend and oppose evil dragons. Or more basicaly the metallics oppose the chromatics. Often places in a kingdom or area will be under the, often unknowingly, protection of a good dragon.

9.The general, vague worldview of a dragon sees humans much the same way humans see animals or even bugs. So if some bees make a hive way out in your back yard...well away from your house...most people just tend to let them.

10.As dragons are powerful fantasy magical creatures...the way to oppose them is with fantasy and magic...not mundane things. For some versions of D&D this breaks the idea that ''the whole world must be sad, pathetic, weak commoners'', but once you get past that it works. This would give the average castle some sort of anti dragon weapon or protection.

Finally a point I can agree with from DU. To expand a little:

8a) and even the evil ones tend to be less "destructive brute from the sky" and more manipulator. At least the older, more powerful ones. The only ones stupid enough to go raiding are the ones that are least likely to be successful at it.

9a) The secret to dealing with dragons is not meddling in their affairs. For you are crunchy and taste good with ketchup. The secret behind how dragons get old is that they realize that that forbearance has to go both ways. Kind of "you ignore my occasional pantry raids and I'll ignore you digging around my back yard."

10a) Or the fortress might know someone who can come make the dragon's life a pain if it gets obnoxious. After all, forts are expensive to build and maintain--organizations who can do that can also have powerful people on retainer. Or other sources of help, like divine aid or other dragons. And dragons, in part due to their long lives, are not known for taking unnecessary risks.

[Max_Killjoy]

Sure. But those are implementation details, free parameters that can be twiddled to suit the need. At least in well-designed (robust, generalizable or modular) systems. Those have nothing to do with NPCs though--NPCs facing NPCs win based on story concerns, not on mechanics. Part of that story is verisimilitude and immersion. But the NPCs don't have to go through the tedious interface called the game rules. Whatever should happen (either to suit the story or to maintain verisimilitude or whatever) just happens. It's only when PCs get involved that the rules must be invoked, and those are mutable. One must also beware of importing real-life assumptions where they're not explicitly imported by the system. If you impose additional assumptions, you're the one bearing the risk of breakage. Not the system or the setting.


That's where we're going to have to disagree -- I should be able to put two NPCs into the system, and use the system to have them interact if I want to, just as if they were two PCs or a PC and an NPC, and (especially if I run the interaction many times to get an "average"), the interactions should feel like I'd expect them to feel based on who those two characters are supposed to be.

[PhoenixPhyre]

That's where we're going to have to disagree -- I should be able to put two NPCs into the system, and use the system to have them interact if I want to, just as if they were two PCs or a PC and an NPC, and (especially if I run the interaction many times to get an "average"), the interactions should feel like I'd expect them to feel based on who those two characters are supposed to be.

Why would you want to? That's not what the system was designed to do. In any system. Ever. The human bandwidth required for such an attempt, if nothing else, is prohibitive. The world exists entirely as a foil for the players. A stage. This is intrinsic in fiction--the fictional world has no independent existence. You can't simulate it at anything approaching fidelity--any attempt to do so will only make things worse. Not to mention--any system that could adequately simulate enough such interactions (including simulating an economy) would be extremely unplayable.

This goes against what you always claim, namely that a system should do one thing and do it well. Trying to make a system that can simulate a living world without DM involvement is trying to do a billion, separate, very difficult things. You'll have to choose what matters, and focus on that or have a mess. That's the biggest lesson to me of 3e D&D--it tried to stuff everything into the same system and failed miserably. Not because it didn't try hard enough, but because the task is sysiphean.

[CharonsHelper]

Well-designed fields of fire are great, but if your archers can't hit the darn thing more than 1/20th of the time, and that 1 in 20 fails to penetrate due to not being with a magic weapon or just not hitting hard enough, they don't really matter.

Even in 3e - just get a bunch of 'lesser orb' or acid arrow wands. A dragon's touch AC is crap, and a few dozen wand users hidden with Improved Cover (giving improved evasion) would shred any dragon. I actually had that be the fluff reason why cities put up with egotistical/annoying mage guilds.

[Max_Killjoy]

Why would you want to? That's not what the system was designed to do. In any system. Ever. The human bandwidth required for such an attempt, if nothing else, is prohibitive. The world exists entirely as a foil for the players. A stage. This is intrinsic in fiction--the fictional world has no independent existence. You can't simulate it at anything approaching fidelity--any attempt to do so will only make things worse. Not to mention--any system that could adequately simulate enough such interactions (including simulating an economy) would be extremely unplayable.


I didn't say ALL characters ALL the time, I said any two (or whatever) characters -- NPCs should be buildable / mapable using the same rules as PCs, and should be able to interact within the system the same way PCs do -- have to no, but able to yes. Not that "short form" layouts for minor NPCs is a bad idea, but there should be nothing inherently different that breaks if the system used to build PCs is used to build an NPC, or if the system itself is used to resolve a conflict between any two fully-built characters.

I cannot approach a "secondary world" as existing as a "stage" or "foil" for the players (RPG) or for the protagonists (RPG or fiction)... I cannot treat it as set dressing or a means to an ends. If I'm going to do it justice, I have to treat it as if it were a real place that would exist without the players or the protagonists.

And when I'm playing, the last thing I want is the sense that the entire world and all the events therein revolve around my PC. I don't want to be The Chosen One, the Savior of All Good People, I don't want Narrative Causality and a cast of NPCs who are sadly inept and utterly doomed without my character to make it all better.

This has nothing to do with mathematically simulating an entire world in the sense of an astrophysicist simulating the formulation of a planetary system or an ecologist simulating the ecosystem of a jungle island.




This goes against what you always claim, namely that a system should do one thing and do it well. Trying to make a system that can simulate a living world without DM involvement is trying to do a billion, separate, very difficult things. You'll have to choose what matters, and focus on that or have a mess. That's the biggest lesson to me of 3e D&D--it tried to stuff everything into the same system and failed miserably. Not because it didn't try hard enough, but because the task is sysiphean.


Did I say that it should run without the GM?

[PhoenixPhyre]

I didn't say ALL characters ALL the time, I said any two (or whatever) characters -- NPCs should be buildable / mapable using the same rules as PCs, and should be able to interact within the system the same way PCs do -- have to no, but able to yes. Not that "short form" layouts for minor NPCs is a bad idea, but there should be nothing inherently different that breaks if the system used to build PCs is used to build an NPC, or if the system itself is used to resolve a conflict between any two fully-built characters.


The rules exist to play the game. If you're playing NPC vs NPC, you're not playing the game as designed. Why should the rules work for this?



I cannot approach a "secondary world" as existing as a "stage" or "foil" for the players (RPG) or for the protagonists (RPG or fiction)... I cannot treat it as set dressing or a means to an ends. If I'm going to do it justice, I have to treat it as if it were a real place that would exist without the players or the protagonists.


But you can do that without a game rules system (or with a rule system designed for that task, instead of one designed for engaging game-play). That's the whole point. The parts that make it a living world are independent, and aside from the parts that make it a game. You don't use one to do the other. That's asking for disappointment. The reason PCs are different from NPCs is that a separate human being (with agency and with desires for fun separate from your own) plays them. A DM who kills off NPCs callously is very different from one who does the same to PCs. The inclusion of other people changes the nature of the game and requires separate, special handling to preserve fun.



And when I'm playing, the last thing I want is the sense that the entire world and all the events therein revolve around my PC. I don't want to be The Chosen One, the Savior of All Good People, I don't want Narrative Causality and a cast of NPCs who are sadly inept and utterly doomed without my character to make it all better.


No one said it had to. The NPC / NPC interactions just don't need to use the same mechanical infrastructure as PC/NPC interactions, because there's only one real human doing the interacting. Things just happen in whatever way you choose. Whether that's with an eye to "what makes sense to you" (which is very different than what "makes sense to me," or narrative causality, or any such thing, that's up to you. It's always up to you (the creator)--there's no one else there. Mechanics are suggestions, baselines, defaults. They're designed to be overridden where necessary.

The idea that this dooms stories to being about special snowflakes/promised messiahs is a bias that you're trying to promote to fact. It's unjustified. You've got this strange (to me at least) dichotomy that either the entire world uses a unified mechanical infrastructure or it's "narrative causality". I for one, don't see it that way. PC/NPC and NPC/NPC are completely separate levels of abstraction that demand separate handling. If you try to handle a bulk material using the same techniques you do for individual atoms, you're in for a world of hurt. Same goes here. Use the technique that fits what you're trying to do. The point of the game is to have fun, not create a maximally "believable" (for whatever fixed meaning that might have) world. Those two are often in competition with each other--the most believable world out there is the real one. And it's not all that fun for games, especially if you like fantastic things.

And the real world isn't that believable. "Absurd" (low-probability) things happen constantly. Don't make the mistake of trying to answer all the questions. You can't do it. No one can. And trying to do so creates a rigid product that isn't useful for it's prime purpose, having fun.



This has nothing to do with mathematically simulating an entire world in the sense of an astrophysicist simulating the formulation of a planetary system or an ecologist simulating the ecosystem of a jungle island.


Did I say it was? No. If you need to be able to cover any arbitrary interaction between any two NPCs and you need to stuff all of that, including economics, politics, and all the other moving parts that make a real world, into a player-accessible framework, you're going to end up with an utter mess. The two things (PC/NPC interactions within a specific genre of game and NPC/NPC interactions within a living world) are on completely different scales, for completely different purposes with contrasting goals and so deserve separate mechanical elements. It's like insisting you have to be able to dig all holes with a fork or it's a bad fork. That's not what a fork is for, stop using it as a backhoe or blaming it for being a bad backhoe.



Did I say that it should run without the GM?

That's an extrapolation on my part--otherwise you're just playing with yourself.

[Aliquid]

If an NPC decides to help out the PCs during an altercation (that is taking place with a different NPC) then you have some NPC vs NPC action going on. This isn’t uncommon for games I have been in.

As a player, if I would be pretty annoyed if the DM just hand waved it so that NPC vs NPC interactions like that resolved in whatever way the DM wanted.

“Of course the blacksmith successfully helped us chase off the goblins. It was predetermined by the DM.”

“Of course the blacksmith died trying to help us chase off the goblins. It was predetermined by the DM”

[PhoenixPhyre]

If an NPC decides to help out the PCs during an altercation that is taking place with a different NPC, then you have some NPC vs NPC action going on. This isn’t uncommon for games I have been in.

As a player, if I would be pretty annoyed if the DM just hand waved it so that NPC vs NPC interactions like that resolved in whatever way the DM wanted.

“Of course the blacksmith successfully helped us chase off the goblins. It was predetermined by the DM.”

“Of course the blacksmith died trying to help us chase off the goblins. It was predetermined by the DM”

If they're part of a combat encounter involving the PCs, then they're part of a PC/NPC interaction (by definition). However, if NPCs are taking to NPCs, you don't have to use the same persuasion rules (for example). Those exist to see how well the PC deals with an NPC (between actual human players), not for the DM to play with themselves.

[Max_Killjoy]

If they're part of a combat encounter involving the PCs, then they're part of a PC/NPC interaction (by definition). However, if NPCs are taking to NPCs, you don't have to use the same persuasion rules (for example). Those exist to see how well the PC deals with an NPC (between actual human players), not for the DM to play with themselves.

The system exists to map and arbitrate.

An NPC who is as strong as PC Bob should have the same STR score as PC Bob.
An NPC who is as smart as PC Eliza should have the same INT score as PC Eliza.
An NPC who is as good at fixing cars as PC Reggie should have the same "fixing cars" ability (however that works) as PC Reggie.
Etc.
Etc.
Etc.

And if the GM wants to determine what happens with an NPC "offstage" (to use the set-dressing metaphor) in a non-arbitrary way, then the GM should be able to use the same rules that the PCs would use.

[Grod_The_Giant]

Leaving aside Max_Killjoy's obligatory tangent railing against D&D/non-simultationist rulesets... there's a definite difference between defending against 3.5 dragons and against 5e ones.

5e isn't that bad, since bounded accuracy means you can hit a dragon (a bog-standard Guard has a 1 in 4 chance of hitting an Adult Red Dragon) and they don't have DR or spellcasting. You need a lot of archers, and you'd need cover for them. How much cover is kind of hard to determine, given how vague the rules about damaging objects are, but you've got plenty of stone towers and such lying around. You'd take heavy casualties, especially in the initial minute of the attack when everyone is running around in a panic, but a castle is big enough that there would still be more guys out of range. A few hundred archers scattered around a big castle would be enough to make even a big dragon wary, I'd think. Maybe have some stone shelters scattered around the battlements/on top of towers.

3.5... you're pretty much screwed, since big dragons are going to be affected only by crits, and will have DR on top of that, and are spellcasters. I'd think the best you can do then is to have a basement bunker to hide the nobles in, and hope the dragon gets bored/adventurers show up before it digs its way to you.

[LordCdrMilitant]

I just like to mount antiaircraft weapons on castle walls. Siege machines mounted on fast-tracking mounts that permit high-angle fire, with ammunition fused to explode on contact or at a prescribed altitude.

Apart from dragons, who present a negligible threat to well fortified locations, armies have airships and can hypothetically tame flying beasts, so it does make logical sense to be prepared to defend against aerial attack.

[Darth Ultron]

I thought of another thing...

11.Timescale. Dragons live for over a thousand years and even more so they sleep for massive amounts of time. Just normal sleep can last years for a dragon, plus hibernation and this does not count magical sleep. A dragon can easily sleep through the whole existence of a human kingdom.

[PhoenixPhyre]

Leaving aside Max_Killjoy's obligatory tangent railing against D&D/non-simultationist rulesets... there's a definite difference between defending against 3.5 dragons and against 5e ones.

5e isn't that bad, since bounded accuracy means you can hit a dragon (a bog-standard Guard has a 1 in 4 chance of hitting an Adult Red Dragon) and they don't have DR or spellcasting. You need a lot of archers, and you'd need cover for them. How much cover is kind of hard to determine, given how vague the rules about damaging objects are, but you've got plenty of stone towers and such lying around. You'd take heavy casualties, especially in the initial minute of the attack when everyone is running around in a panic, but a castle is big enough that there would still be more guys out of range. A few hundred archers scattered around a big castle would be enough to make even a big dragon wary, I'd think. Maybe have some stone shelters scattered around the battlements/on top of towers.

3.5... you're pretty much screwed, since big dragons are going to be affected only by crits, and will have DR on top of that, and are spellcasters. I'd think the best you can do then is to have a basement bunker to hide the nobles in, and hope the dragon gets bored/adventurers show up before it digs its way to you.

Two things about 3.5 dragons--

1) the castle under 3.5 assumptions would also likely have access to higher levels/magic weapons. Because Magic Mart. As much as I don't like that way of things, it's the established thing. That is, you're causing yourself difficulties if you assume big dragon vs level 1 guards. The system doesn't assume that, so why should you?

2) I find it unlikely that an established dragon (adult+, the big ones) would be likely to attack a fortified area, for the reasons given above. The only ones likely to are the young adults trying to make a lair/name for themselves. And those are small enough to be handled normally (if with significant losses).

But either way, designing your fortifications around a rare, cataclysmic event like a dragon attack is foolishness. Yes, sure, have AA weapons available. Flying creatures of lesser size are (relatively) more common. But designing a specialized anti-dragon castle? That's like trying to harden a convenience store security system against James Bond (or the Mission Impossible guys). Even if you could theoretically do it, it's a total waste of time and effort. Just evacuate, minimize your losses, and rebuild when the thing moves on. Rampaging adult+ dragons are natural disasters, not military threats the vast majority of the time. Treat them as such.

And the whole point was that saying "it doesn't make sense to have a regular castle due to dragons" was a result of bias and refusal to look at the actual facts rather than an honest assessment of a setting or a system. The same analysis can be done for high-level wizards, flying cavalry (although there you'll actually want to take some countermeasures, as the expectation value of the loss is greater), etc. As long as mortal, regular armies are the dominant threat (both in number and in consequences--a castle can be rebuilt. Occupation and invasion are more difficult to handle), castles will continue to look like earth ones, for the same reasons.

Dragons, wizards, etc. are like air power in WWII. Capable of causing massive damage, but not sufficient to win a war. You still need infantry divisions to actually hold ground.

[Darth Ultron]

And the whole point was that saying "it doesn't make sense to have a regular castle due to dragons" was a result of bias and refusal to look at the actual facts rather than an honest assessment of a setting or a system.

Also most people are stuck in the idea that ''the game world must be 1st level'' or otherwise ''just like mundane Earth''. But this is not true at all. Everyone is locked in the thought that while dragons can have epic power and magic...the rest of the world must only have rocks and sharp sticks.

A typical castle might have such things as a Dragonward (that keeps dragons out) or a powerful offensive weapon like a Disintegration Cannon.

[Bogwoppit]

Also most people are stuck in the idea that ''the game world must be 1st level'' or otherwise ''just like mundane Earth''. But this is not true at all. Everyone is locked in the thought that while dragons can have epic power and magic...the rest of the world must only have rocks and sharp sticks.

A typical castle might have such things as a Dragonward (that keeps dragons out) or a powerful offensive weapon like a Disintegration Cannon.
I tend to agree - the immense expense of building a castle would mean that to NOT spend several thousand GP on making a few magic items for attack and defence would be silly.
These items don't have to be exclusively aimed at dragon attack either: Darth's disintegration cannon would work against enemy siege engines too.

In practice, what does this mean though? Looks to me like we end up with a standard-looking medieval castle, with some magic items boosting its defence / offence.

[Max_Killjoy]

I tend to agree - the immense expense of building a castle would mean that to NOT spend several thousand GP on making a few magic items for attack and defence would be silly.
These items don't have to be exclusively aimed at dragon attack either: Darth's disintegration cannon would work against enemy siege engines too.

In practice, what does this mean though? Looks to me like we end up with a standard-looking medieval castle, with some magic items boosting its defence / offence.

That would at least count as addressing the elephant dragon in the room.

In most of these settings, instead, it's always a total lack of preparation and total panic and shock at this unknown thing and OMG all our thatch or wooden rooves are on fire and men dying as they try to attack with swords... EVERY SINGLE TIME there's a dragon attack (or an attack by any one of a near-endless list of other fantastic elements).

[Knaight]

A typical castle might have such things as a Dragonward (that keeps dragons out) or a powerful offensive weapon like a Disintegration Cannon.

This would be more than enough to make a typical castle no longer a regular castle, but one shaped by the fantasy world. The complaint regarding FR is exactly that this doesn't happen, with a dragon as an example.

It's far from the best example too - basically any reasonably ubiquitous trainable creature with a high burrowing speed (and even 5ft/round is more than a foot per second given a double move, and thus completely ridiculous by real world standards) could cause all sorts of undermining problems and poses a fairly large threat.

[PhoenixPhyre]

This would be more than enough to make a typical castle no longer a regular castle, but one shaped by the fantasy world. The complaint regarding FR is exactly that this doesn't happen, with a dragon as an example.

It's far from the best example too - basically any reasonably ubiquitous trainable creature with a high burrowing speed (and even 5ft/round is more than a foot per second given a double move, and thus completely ridiculous by real world standards) could cause all sorts of undermining problems and poses a fairly large threat.

Is it possible that they do have such things and they're just not mentioned (because 99+% of the time they're irrelevant?) I mean most setting books don't do much with castle layouts and defenses. And most of the novels date to 2e, where the threats (and system!) were very different. And the novels don't even use the game system at all, really.

[DaveOTN]

One of the most memorable moments of the 3.5 campaign I played in college came when my druid wild shaped into a dire badger and, over the course of about 20 minutes, collapsed an evil wizard's tower.

Burrowing is in many ways a bigger threat. On the other hand, it gives you the perfect opportunity for squad-level combat during a siege - teams of sappers and counter-sappers, tunneling into each other and fighting quick, brutal actions underground as one side tries to collapse the tunnel and the other tries to hold it long enough for a larger force to get through.

Look at Pietro Micca for inspiration, who lit a couple of barrels of gunpowder on fire around him to keep the French out of Turin in 1706.

[PhoenixPhyre]

One of the most memorable moments of the 3.5 campaign I played in college came when my druid wild shaped into a dire badger and, over the course of about 20 minutes, collapsed an evil wizard's tower.

Burrowing is in many ways a bigger threat. On the other hand, it gives you the perfect opportunity for squad-level combat during a siege - teams of sappers and counter-sappers, tunneling into each other and fighting quick, brutal actions underground as one side tries to collapse the tunnel and the other tries to hold it long enough for a larger force to get through.

Look at Pietro Micca for inspiration, who lit a couple of barrels of gunpowder on fire around him to keep the French out of Turin in 1706.

This I agree with (mostly). I'd expect the anti-sapping defenses to be a unstated in most RPGs unless that was a strategy specifically envisioned by the scenario designer. IIRC, burrowing speeds don't work through stone unless specifically stated as such, and most don't. Earth glide (for earth elementals) doesn't leave a hole. This makes even having a stone foundation (which is needed for structural stability) does a lot of work against simple enhanced burrowing techniques. Basically, things like that makes traditional sapping easier, but not revolutionarily so.

[Darth Ultron]

This would be more than enough to make a typical castle no longer a regular castle, but one shaped by the fantasy world. The complaint regarding FR is exactly that this doesn't happen, with a dragon as an example.


What regarding FR? It is the perfect example of a world made in a logical way. FR is full of things that stop dragons and keep them in check. For example: Waterdeep has a Dragonward. And the setting is full of magic ''not stated out in boring rules''.

And that does not even cover the powerful NPCs.

[Max_Killjoy]

The rules exist to play the game. If you're playing NPC vs NPC, you're not playing the game as designed. Why should the rules work for this?


Why shouldn't the rules for a character be the rules for a character be the rules for a character?

Why shouldn't any two characters be able to fully interact?

Why shouldn't any character be able to interact with the setting and the rules on equal terms with other characters (proportionally to their relevant abilities)?




But you can do that without a game rules system (or with a rule system designed for that task, instead of one designed for engaging game-play). That's the whole point. The parts that make it a living world are independent, and aside from the parts that make it a game. You don't use one to do the other. That's asking for disappointment. The reason PCs are different from NPCs is that a separate human being (with agency and with desires for fun separate from your own) plays them. A DM who kills off NPCs callously is very different from one who does the same to PCs. The inclusion of other people changes the nature of the game and requires separate, special handling to preserve fun.


In order to preserve the fun for me, and not cause "wait, what the hell?" moments, the parts that make it a living world, and the parts that let the players interact with it, and the parts that let characters and the environment and so on interact, all need to be compatible and in sync.

In order to preserve the fun for me, the fictional world and the other characters in it besides the PCs need to feel like they have an existence independent of the PCs, that they aren't just waiting offstage drinking coffee, and hitting the catering truck for a wrap or a doughnut, and doing costume changes, while they wait for their next scene with the PCS.

Plus separate rules means more rules to keep straight -- and also potentially more complexity and more edge cases and more unintended interactions.




No one said it had to. The NPC / NPC interactions just don't need to use the same mechanical infrastructure as PC/NPC interactions, because there's only one real human doing the interacting. Things just happen in whatever way you choose. Whether that's with an eye to "what makes sense to you" (which is very different than what "makes sense to me," or narrative causality, or any such thing, that's up to you. It's always up to you (the creator)--there's no one else there. Mechanics are suggestions, baselines, defaults. They're designed to be overridden where necessary.

The idea that this dooms stories to being about special snowflakes/promised messiahs is a bias that you're trying to promote to fact. It's unjustified. You've got this strange (to me at least) dichotomy that either the entire world uses a unified mechanical infrastructure or it's "narrative causality". I for one, don't see it that way. PC/NPC and NPC/NPC are completely separate levels of abstraction that demand separate handling. If you try to handle a bulk material using the same techniques you do for individual atoms, you're in for a world of hurt. Same goes here. Use the technique that fits what you're trying to do. The point of the game is to have fun, not create a maximally "believable" (for whatever fixed meaning that might have) world. Those two are often in competition with each other--the most believable world out there is the real one. And it's not all that fun for games, especially if you like fantastic things.


Sorry, when someone says that the PCs are different and get different rules because they're the PCs, and that the fictional world is there to serve as a foil and stage for the PCs and nothing more... that's immediately what that implies to me, that the entire fictional world revolves around the PCs and that they're special because they're the PCs (as opposed to being the PCs because they're exceptional, or even just because they're the character's that the players have chosen to somehow focus on).




And the real world isn't that believable. "Absurd" (low-probability) things happen constantly. Don't make the mistake of trying to answer all the questions. You can't do it. No one can. And trying to do so creates a rigid product that isn't useful for it's prime purpose, having fun.


The history of fiction is full of stories ruined by low-probability pileup, where event after event is "excused" by being technically possible.




Did I say it was? No.


And yet your comments keep coming across as if that's what you think I'm talking about, as evidenced by the rest of that paragraph as follows...




If you need to be able to cover any arbitrary interaction between any two NPCs and you need to stuff all of that, including economics, politics, and all the other moving parts that make a real world, into a player-accessible framework, you're going to end up with an utter mess. The two things (PC/NPC interactions within a specific genre of game and NPC/NPC interactions within a living world) are on completely different scales, for completely different purposes with contrasting goals and so deserve separate mechanical elements. It's like insisting you have to be able to dig all holes with a fork or it's a bad fork. That's not what a fork is for, stop using it as a backhoe or blaming it for being a bad backhoe.


Again, this isn't about a mathematical system for simulating an entire world.

It's about any character having the same tools (proportional to their abilities) to interact with each other and with the setting, if for no other reason than to represent what they're capable of relative to each other and the setting, on the same scale, and inform one's decisions.




That's an extrapolation on my part--otherwise you're just playing with yourself.


Or, you know, there's an NPC interaction or task not directly involving the PCs, that maybe the GM wants to neutrally resolve because it wouldn't be fair in that specific circumstance to do it by outright fiat...

[Batou1976]

That's where we're going to have to disagree -- I should be able to put two NPCs into the system, and use the system to have them interact if I want to, just as if they were two PCs or a PC and an NPC, and (especially if I run the interaction many times to get an "average"), the interactions should feel like I'd expect them to feel based on who those two characters are supposed to be.

I find I must concur.

Sure, as DM I can predetermine which (group of) NPC(s) prevails in an encounter, in service to the narrative I'm presenting... but I don't/ shouldn't have to if I don't want to.

Maybe I don't want the outcome to be predetermined (and not dependent on the PC's actions), maybe I want the victor to be random... or maybe I don't particularly care who wins, but want a way to determine a victor that's a bit more involved than "heads Team A wins/ tails Team B wins". Maybe one of those groups of NPCs is acting in support of the PCs, but isn't under the player's direct control.

Whatever the case, I prefer NPC:NPC interactions to operate under the same rules as PC:NPC ones. The way the NPCs "stat block" is derived doesn't necessarily have to match the way PCs are generated, but the mechanics governing how they interact with the world should.

[PhoenixPhyre]

I find I must concur.

Sure, as DM I can predetermine which (group of) NPC(s) prevails in an encounter, in service to the narrative I'm presenting... but I don't/ shouldn't have to if I don't want to.

Maybe I don't want the outcome to be predetermined (and not dependent on the PC's actions), maybe I want the victor to be random... or maybe I don't particularly care who wins, but want a way to determine a victor that's a bit more involved than "heads Team A wins/ tails Team B wins". Maybe one of those groups of NPCs is acting in support of the PCs, but isn't under the player's direct control.

Whatever the case, I prefer NPC:NPC interactions to operate under the same rules as PC:NPC ones. The way the NPCs "stat block" is derived doesn't necessarily have to match the way PCs are generated, but the mechanics governing how they interact with the world should.

Do you really go through the whole combat/social routine (especially with players at the table) when the PCs aren't involved? That sounds like an atrocious waste of time and effort. If the players aren't at the table, then it's manual monsterbation (as I heard it put on these forums). Either way, it's really pointless, because all it does is create the illusion of neutrality. THERE IS NO NEUTRAL PARTY. Everything except the PCs exists and acts if and only if the DM puts it there. Pushing the responsibility for outcomes onto the system is a way of shirking responsibilities.

One of two things is true--either the system dictates every choice made (and thus has to cover every single possibility to exhaustion) or it doesn't. If it doesn't, the outcome is entirely up to DM fiat. Because the DM decides when to invoke the system and this taints everything that follows. You can't escape that responsibility by blaming the system.


But all of this is off topic. The entire point of this thread is not to rehash old, immovable arguments. It's to show that the accusation of "it doesn't make sense" as a reason to dislike a setting or system is frequently a lazy accusation born of ignorance or bias. Given any non-trivial system, and any non-trivial setting, a plausible argument can be made for just about anything with enough knowledge. Setting/system dissonance is, in the vast majority of cases, either from misunderstanding the setting or taking the system out of context. The rest of the time its because you're trying to smash a genre-bound system into a setting that doesn't do that genre. Or misunderstanding what the system is trying to portray. Or all of the above.

[Batou1976]

Do you really go through the whole combat/social routine (especially with players at the table) when the PCs aren't involved? That sounds like an atrocious waste of time and effort. If the players aren't at the table, then it's manual monsterbation (as I heard it put on these forums). Either way, it's really pointless, because all it does is create the illusion of neutrality. THERE IS NO NEUTRAL PARTY. Everything except the PCs exists and acts if and only if the DM puts it there. Pushing the responsibility for outcomes onto the system is a way of shirking responsibilities.

If by that you mean something along the lines of "do I have Left-hand-sock-puppet talk to Right-hand-sock-puppet then roll a die to determine which sock puppet won the argument, all while the players are watching and wondering how they ended up with such a bug-nuts insane DM"... then no I don't do that. What I do is think to myself "Chamberlain Bob wants to persuade Guard-Captain Flashheart to deny the fencing master's petition to open a sword school in the Otyugh district" then roll Chamberlain Bob's persuade skill (or whatever else may be relevant) against Flashheart's relevant opposed ability/skill to see if the Chamberlain succeeds. Then, if/when it becomes relevant to the adventure, I declare the result to the players. I often prefer to do this rather than just decree one way or the other, or flip a coin.




But all of this is off topic. The entire point of this thread is not to rehash old, immovable arguments. It's to show that the accusation of "it doesn't make sense" as a reason to dislike a setting or system is frequently a lazy accusation born of ignorance or bias. Given any non-trivial system, and any non-trivial setting, a plausible argument can be made for just about anything with enough knowledge. Setting/system dissonance is, in the vast majority of cases, either from misunderstanding the setting or taking the system out of context. The rest of the time its because you're trying to smash a genre-bound system into a setting that doesn't do that genre. Or misunderstanding what the system is trying to portray. Or all of the above.

All true enough... unless you're talking about FATAL or World of Synnabar(sp?). Either of those games seem to be complicated simply for the sake of being complicated, to the point where many things just don't make sense (but this isn't the only reason I dislike either of those). But the less said about them (FATAL in particular), the better... :smalleek:

I had something else... but alas it has escaped me now. Ah, the joys of approaching middle age... :smallsigh::smallfrown:

[PhoenixPhyre]

If by that you mean something along the lines of "do I have Left-hand-sock-puppet talk to Right-hand-sock-puppet then roll a die to determine which sock puppet won the argument, all while the players are watching and wondering how they ended up with such a bug-nuts insane DM"... then no I don't do that. What I do is think to myself "Chamberlain Bob wants to persuade Guard-Captain Flashheart to deny the fencing master's petition to open a sword school in the Otyugh district" then roll Chamberlain Bob's persuade skill (or whatever else may be relevant) against Flashheart's relevant opposed ability/skill to see if the Chamberlain succeeds. Then, if/when it becomes relevant to the adventure, I declare the result to the players. I often prefer to do this rather than just decree one way or the other, or flip a coin.


That's valid, but having to create all the background stuff for the characters, to me, either makes it a foregone conclusion (especially in a system like 3.5 where the range of possible skill values is really high, so someone who focuses on a social skill will make that check 99% of the time against someone who doesn't so focus), or it's basically random and I'd be better served assigning probabilities and rolling d100. And in no case can I claim that the system made the decision--after all, I assigned those skill points myself (or chose to take the NPC stat block) and so to me it's a case of "oh, I want to randomize this" convenience rather than a requirement for making it a living world. YMMV.



All true enough... unless you're talking about FATAL or World of Synnabar(sp?). Either of those games seem to be complicated simply for the sake of being complicated, to the point where many things just don't make sense (but this isn't the only reason I dislike either of those). But the less said about them (FATAL in particular), the better... :smalleek:

I had something else... but alas it has escaped me now. Ah, the joys of approaching middle age... :smallsigh::smallfrown:

I can commiserate on the bold part. I (jokingly) swear I have juvenile-onset Alzheimer syndrome--I do the old-person "walk into a room, wander about wondering why I came, leave, and then realize why I was there in the first place" thing frequently.

[Max_Killjoy]

Do you really go through the whole combat/social routine (especially with players at the table) when the PCs aren't involved? That sounds like an atrocious waste of time and effort. If the players aren't at the table, then it's manual monsterbation (as I heard it put on these forums). Either way, it's really pointless, because all it does is create the illusion of neutrality. THERE IS NO NEUTRAL PARTY. Everything except the PCs exists and acts if and only if the DM puts it there. Pushing the responsibility for outcomes onto the system is a way of shirking responsibilities.

One of two things is true--either the system dictates every choice made (and thus has to cover every single possibility to exhaustion) or it doesn't. If it doesn't, the outcome is entirely up to DM fiat. Because the DM decides when to invoke the system and this taints everything that follows. You can't escape that responsibility by blaming the system.



I thought better of you than to resort to juvenile sexual metaphors to belittle a position different from your own.

And you're really telling me what you've never had a situation arise in which is was more advisable to determine something not directly involving the PCs via mechanical rather than purely fiat means, for fairness or because you didn't have a preferred outcome or to introduce risk or for any other reason?




But all of this is off topic. The entire point of this thread is not to rehash old, immovable arguments. It's to show that the accusation of "it doesn't make sense" as a reason to dislike a setting or system is frequently a lazy accusation born of ignorance or bias. Given any non-trivial system, and any non-trivial setting, a plausible argument can be made for just about anything with enough knowledge. Setting/system dissonance is, in the vast majority of cases, either from misunderstanding the setting or taking the system out of context. The rest of the time its because you're trying to smash a genre-bound system into a setting that doesn't do that genre. Or misunderstanding what the system is trying to portray. Or all of the above.



Yes, of course, if you go far enough down the rabbit-hole of nested excuses and "explanations", of course any objection can be belittled as the supposed "ignorance" of those who make the objection.


Consider for a moment a hypothetical show about some bounty hunters. Every episode features one of their hunts or other jobs going sideways, so that they end up not getting paid. Comments are made as to how tight their budget is. And yet, despite this, they always have at least enough fuel and ammo to get through the episode, they always keep their vehicles function. Despite never once being shown pulling a successful job and getting paid, despite no explanation for how they're paying for anything.

Some viewers will "explain" this with something like the following -- "The episodes don't show the boring successful jobs, just the entertaining complicated messed-up jobs." We're never shown, never told, never informed by the creators of the show, never given any hint that those other jobs exist, but some fans will insist that those jobs just must exist... they'll doggedly cling to that excuse.

Other viewers will say it doesn't matter, because the point of the show is to entertain, not to make sense.

And then there are those of us who will say "Based on what we've actually been shown and told within the show, this doesn't add up." The first sort of viewer will say "You're ignorant" because we don't accept what they've made up to paper over the problem. The second sort of viewer will say "You're ignorant" because we're supposedly missing the "point" of being "entertained", even though the holes specifically make the show less entertaining for us.


Show me a fantasy setting that is stuffed full of an endless array of aerial and tunneling magical and other fantastical threats... and yet is dotted with the sort of castles found in real medieval Europe showing no change whatsoever and no consideration given to the multitude of different threats that real castles did not face, and I'm not going to pretend that it doesn't matter, and I'm not going to make excuses for it along the lines of "oh well the defenses must be there even if we never see any hint of them".




That's valid, but having to create all the background stuff for the characters, to me, either makes it a foregone conclusion (especially in a system like 3.5 where the range of possible skill values is really high, so someone who focuses on a social skill will make that check 99% of the time against someone who doesn't so focus), or it's basically random and I'd be better served assigning probabilities and rolling d100. And in no case can I claim that the system made the decision--after all, I assigned those skill points myself (or chose to take the NPC stat block) and so to me it's a case of "oh, I want to randomize this" convenience rather than a requirement for making it a living world. YMMV.


I don't think anyone is claiming that it's a purely "mindless" neutral decision/outcome made purely by the system. (Or claiming that this is a necessity for a "living world".)

You seem to be arguing that because it can't be purely decided by the system, one might as well just decide for one's self or flip a coin?

[Bogwoppit]

I can totally see how this discussion about game systems for NPCs helps resolve the discussion about castles and dragons :smallbiggrin:

Having a castle mage will let you deal with most issues where a standard RL castle design falls down (ha!) against fantasy threats.
The idea that a wizard of so.e sort was a standard member of staff for a castle is built right into the rules of BECM DnD, for exactly those reasons.

[Thinker]

Why shouldn't the rules for a character be the rules for a character be the rules for a character?

Why shouldn't any two characters be able to fully interact?

Why shouldn't any character be able to interact with the setting and the rules on equal terms with other characters (proportionally to their relevant abilities)?


I'd like to address these points if you don't mind.

Player characters and nonplayer characters have different roles within the game. As a GM, I don't need very much information about most nonplayer characters for them to fulfill their function. The only stats/abilities I need to worry about when creating an NPC are those relevant to how they are going to interact with the player characters. If they're going to be witnesses to a crime, I really just need to know their personalities, professions, and what they saw, heard, and felt. If they're healers at the temple, I need to know how much of a "donation" they'll charge and how much they heal. Rarely do I need to know how much HP they have or what sort of attacks they'll need. In the cases where I do, I probably won't need to know what the NPC's profession skill is like. The less prep work I need to do for an NPC encounter, the better so I leave as much out as I can. If the players interact in an unexpected way - for example, they might try to talk to a hostile NPC, I can come up with those numbers on the fly based on the personality and disposition of the NPC. I know the system I am running well enough to not need to worry about the NPC's level or the exact distribution of character points as if it were a regular player character to create a challenge for the players.

As for the nonplayer characters interacting with one another offscreen, that rarely relies very much on stats so much as personality, disposition, and capabilities. The players aren't likely to witness an interaction between two NPCs (that would basically just be me having a conversation with myself), but they are likely to encounter the results of an interaction as dictated from those three things - personality, disposition, and capabilities. The capabilities aren't numbers so much as abilities - command of a gang, authority with the law, ability to cast healing spells, etc. The NPCs can fully interact, but I'm not interested in rolling dice for offscreen effects. The interactions must make sense - a mousy librarian probably shouldn't be the one who broke the punk's arm - but she might call the cops on him if he's smashing up her library.

The player characters are the focus of the game. The nonplayer characters really don't matter that much. It is rare for any of the groups I have GMed for to meet nonplayer characters more than a few times. So long as those interactions made sense, everyone is happy. I don't need a game system to give me hard and fast rules for those interactions. I'd like things to be logically extrapolated - if there is healing magic, nonplayer characters should certainly have access to it, even if it is not expressed in exactly the same way as for player characters. I also don't need the players to have their abilities expressed in the same way as the nonplayer characters.

Let's say that Cure Light Wounds heals 1d8 health on a target touched. The cleric can cast that spell 3 times in a day, but the cleric also has a host of other abilities and training to worry about. I might say that the temple healer can heal on a target touched for 1d8 health 8 times per day. She didn't receive exactly the same training as the cleric and she specializes in healing. I haven't come up with a class for her, but I know that she only really needs to be good at healing. I don't find this to be unreasonable, especially in a class-based system since each class is a full-package of abilities. I could imagine a class if I needed to that had the ability to cast Cure Light Wounds extra times. In a point-based system, things become even easier - she's hyper-specialized in healing and would probably be squashed by a bug the first time she entered a dungeon.

I guess you could say that I believe that the rules of the world still apply to NPCs, but I don't think you need to know the specifics or as much information about the NPCs.

[Max_Killjoy]

I'd like to address these points if you don't mind.

Player characters and nonplayer characters have different roles within the game. As a GM, I don't need very much information about most nonplayer characters for them to fulfill their function. The only stats/abilities I need to worry about when creating an NPC are those relevant to how they are going to interact with the player characters. If they're going to be witnesses to a crime, I really just need to know their personalities, professions, and what they saw, heard, and felt. If they're healers at the temple, I need to know how much of a "donation" they'll charge and how much they heal. Rarely do I need to know how much HP they have or what sort of attacks they'll need. In the cases where I do, I probably won't need to know what the NPC's profession skill is like. The less prep work I need to do for an NPC encounter, the better so I leave as much out as I can. If the players interact in an unexpected way - for example, they might try to talk to a hostile NPC, I can come up with those numbers on the fly based on the personality and disposition of the NPC. I know the system I am running well enough to not need to worry about the NPC's level or the exact distribution of character points as if it were a regular player character to create a challenge for the players.

As for the nonplayer characters interacting with one another offscreen, that rarely relies very much on stats so much as personality, disposition, and capabilities. The players aren't likely to witness an interaction between two NPCs (that would basically just be me having a conversation with myself), but they are likely to encounter the results of an interaction as dictated from those three things - personality, disposition, and capabilities. The capabilities aren't numbers so much as abilities - command of a gang, authority with the law, ability to cast healing spells, etc. The NPCs can fully interact, but I'm not interested in rolling dice for offscreen effects. The interactions must make sense - a mousy librarian probably shouldn't be the one who broke the punk's arm - but she might call the cops on him if he's smashing up her library.

The player characters are the focus of the game. The nonplayer characters really don't matter that much. It is rare for any of the groups I have GMed for to meet nonplayer characters more than a few times. So long as those interactions made sense, everyone is happy. I don't need a game system to give me hard and fast rules for those interactions. I'd like things to be logically extrapolated - if there is healing magic, nonplayer characters should certainly have access to it, even if it is not expressed in exactly the same way as for player characters. I also don't need the players to have their abilities expressed in the same way as the nonplayer characters.

Let's say that Cure Light Wounds heals 1d8 health on a target touched. The cleric can cast that spell 3 times in a day, but the cleric also has a host of other abilities and training to worry about. I might say that the temple healer can heal on a target touched for 1d8 health 8 times per day. She didn't receive exactly the same training as the cleric and she specializes in healing. I haven't come up with a class for her, but I know that she only really needs to be good at healing. I don't find this to be unreasonable, especially in a class-based system since each class is a full-package of abilities. I could imagine a class if I needed to that had the ability to cast Cure Light Wounds extra times. In a point-based system, things become even easier - she's hyper-specialized in healing and would probably be squashed by a bug the first time she entered a dungeon.

I guess you could say that I believe that the rules of the world still apply to NPCs, but I don't think you need to know the specifics or as much information about the NPCs.


Regarding the parts I bolded:

In most campaigns I've run and campaigns I've played in, there were a lot of recurring NPCs, and fewer of the "temple Healing Priest #319" sort. PCs didn't just wander around through nameless towns full of nameless NPCs spending a night and then heading off for the next town. The towns and villages and crossroads, the inns and taverns and businesses, the people who run them and work in them, all get names and little bits of color and personality. Most of my players have expressed interest in repeat visits to the places and people I describe, and they've tended to try to pull NPCs into their adventures rather than shut them out.

I don't run class-based systems, so "make this fit into a class", or "make up a class to fit what this NPC does" is never a concern for me.

[Loxagn]

Dragons in D&D are something of a problem. A few major issues with defending against a dragon are as follows:


Dragons can fly.
They can fly under their own power, and quite fast compared to other creatures. This makes attacking them actually quite difficult, as most attackers would be restricted to attacking with bows and crossbows. Assuming an adult Red dragon, and a 3rd-level Warrior with 15 Dexterity with Weapon Focus (Longbow), the Warrior can't hit the dragon on anything other than a 20, at which point they will deal 1d10 damage with a 5% chance of dealing 2d10 instead. What's worse, Adult red dragons have DR 5/magic, meaning that at that point a single soldier has a 2.5% chance of dealing between 1 and 5 points of damage, with a .1125% chance of dealing between 1 and 15 points of damage. Even assuming an army with 500 archers, during a given round on average 12 of them will deal an average of about 32 damage, factoring in critical hits (rare as they are). This against an enemy who will have, on average, 253 HP. Assuming none of the archers die, this means it will take, on average, eight rounds, or less than a minute, for an army to kill an adult red dragon. These are seemingly good odds, and with savvy troop placement it will be difficult to deal much damage to your forces.
Dragons have breath weapons.
As written in the book, this is not much of an issue. An adult red dragon can deal, on average, 66 fire damage in a 50-foot cone, every 2.5 rounds. With strafing tactics, this can become a deadly weapon on the unfortunate commoners, experts, and aristocrats that populate your city and decimate buildings. Assuming optimal archer placement, a dragon will be able to use this on average three times before being brought down; however, factoring burning buildings and cover, the dragon can do a great deal of damage this way.
If the dragon has, say, metabreath feats, however, especially Enlarge Breath, this rapidly becomes a nearly-instantaneous gamewinner for the dragon. We won't consider that, though, for now.
Dragons are terrifying.
Frightful presence is nothing to sneeze at. If a dragon passes by even slightly close on one of its strafing runs, then 95% of the populace in its immediate vicinity are going to flee in terror, soldiers included. It has no reason not to swoop overhead, roaring in rage and raining fire down on the landscape. This has a drastic effect on reducing the amount of its opposition, as even those optimally-placed archers will flee in terror before the enemy.
Dragons are spellcasters.
This is, admittedly, an extreme hurdle to overcome. Every red dragon above Young age has access to the spell 'Protection from Arrows'. They're even likely to cast it on themselves before starting an engagement with a town, since parent red dragons will teach their young how amusing it is to watch the poor humans try and fail to do anything to you. DR 10/magic is an insurmountable hurdle for our archers, dropping the chances of dealing even a single point of damage to the dragon in a round to .1375%. This is the real gamechanger, as even if one archer confirms a critical for maximum damage every single round against the dragon, it will still take 24 rounds to actually down the dragon. And that's only a first-level spell. Adult red dragons can cast up to 4th-level. A red dragon can easily cast Confusion or Greater Invisibility and, suddenly, cause exponentially more havoc than ever before.
And, most dangerously of all, dragons are intelligent.
All of these would be fine if dragons were big, dumb lizards who flew and used magic. But the problem with that is... they're not. An adult Red has an intelligence of 16, which puts it on par with any human wizard (assuming 15 starting int) below 8th level. This means that, while they might be prideful, while they might be arrogant and overconfident, a dragon is not stupid. A red dragon who feels slighted can, as easily as breathing (literally), swoop in in the dead of night above the clouds, dip below the cloud layer, and breathe a cone of fire that will take several hours to recharge, but will set the town and the landscape in a ten-mile radius ablaze, instantly killing everyone inside with less than about 30 hp (which is most of them), and about 95% of everybody with less than 50. Even restricted to 'Core', the dragon can spend a couple of rounds setting things on fire with relative impunity, cast Confusion or even Contagion in the town square to cause a little bloodshed, and then make itself invisible before soaring off to its lair, before returning every couple of days to repeat the process all over again. If it feels like it, it can polymorph itself into a palace guard and then use its Suggestion SLA to cause the king to declare martial law and order his soldiers to butcher entire districts of the town which have been 'infected'.


An adult red dragon can, in essence, utterly destroy a settlement if it feels like it ought to. The only places which would be immune to this are large metropolises likely to have a few high-level casters on hand, or underground settlements.
Really, the only way to deal with a belligerent dragon is to either have an artifact arrow of Dragon-slaying in your arsenal somewhere, or else have adventurers on tap. You could ask your court wizard to try and poke the dragon with Shivering Touch, as well. A fortified underground bunker would also be a good idea, probably.
Diplomacy is also a universally-good option. It might seem horrific to offer up a virgin sacrifice once a week to appease your new draconic tyrant, but it's probably going to be enough to buy time for adventurers who know what they're doing to get here and help, which is a much better scenario than an enraged dragon deciding your town should just burn now.

Why an adult red dragon for these numbers? Well, frankly, because you're not likely to run into a younger dragon alone. If you find a young dragon, then Mom or Dad are probably nearby. And if your city has, for some reason, managed to piss off an older dragon, then... good luck with that.

[PhoenixPhyre]

Dragons in D&D are something of a problem. A few major issues with defending against a dragon are as follows:

<snip>

An adult red dragon can, in essence, utterly destroy a settlement if it feels like it ought to. The only places which would be immune to this are large metropolises likely to have a few high-level casters on hand, or underground settlements.
Really, the only way to deal with a belligerent dragon is to either have an artifact arrow of Dragon-slaying in your arsenal somewhere, or else have adventurers on tap. You could ask your court wizard to try and poke the dragon with Shivering Touch, as well. A fortified underground bunker would also be a good idea, probably.

Yes, dragons are a problem. But none of that speaks to the points I brought up in the OP, which boil down to

a) rarity. The vast majority of settlements (especially fortified ones as was being discussed) are unlikely to be in the territory of even a single adult red dragon.
b) intelligence => less likely to raid fortifications. Remember that the point I was addressing was that the threat of dragons would cause castles (specifically) to be built differently or it "wouldn't make sense." A thorp or village? Sure. That's toast. But dragons don't get to be adults by being foolhardy. Attacking a fortification means pissing off people who have access to things that can hurt them.
c) trade-offs. Building an anti-dragon fortification (on the rare chance that a dragon's going to attack) makes it much less useful against a more mundane threat (which are much more common and worse, since dragons aren't occupying forces).
d) counters. What kind of castle has level 3 fighters as it's worst threat? Any serious castle would have multiple spell-casters (especially clerics, who can deal with most of the fear relatively trivially), magic items (3.5 assumes that such things are pretty darn common), and high-level guards.

It makes much more sense to have a refuge (probably underground and fortified) that you pull everyone into and hide until the thing goes away. Or not piss off the dragon in the first place. Even adult reds aren't usually marauders--young ones might be, but adults have more sense. These are not Skyrim dragons that attack towers at random--they're thinking beings with agendas and a low tolerance for personal risk.

Dragons aren't social beings, especially chromatic ones. They kick their young out just past the wyrmling stage--you're more likely to find a solo young/young adult raiding cities than a full adult (by which time they've got a lair, probably).

[Lord Torath]

Dragons in D&D are something of a problem. A few major issues with defending against a dragon are as follows:
*snip*This is very Edition dependent. 1E and BECMI dragons can breathe 3x per day, for damage equal to their current hp. 2E dragons can breathe once every 3 rounds for a set range of damage based on color and age, with no options to "conserve" breathe for a larger area of effect. 5E dragons are subject to Bounded Accuracy, meaning the 3rd-level soldiers will hit roughly 25% of the time (if I understand what I've read about 5E). 2E dragons gain spells randomly, so there is no guarantee they will get Protection from Normal Arrows. Not all 1E or BECMI dragons get spells.

If you focus on the general traits all dragons have in common (flight, good AC, breath weapon, etc.), you will get more generally useful insights.

[Loxagn]

<snip>

My apologies. I was assuming 'default' D&D setting, in a situation in which one was likely to see angry dragons (such as the lake town from The Hobbit whose name escapes me at present). With that in mind, having spellcasters on tap is a good counter, as I said, as is a fortified underground defensive position (which conveniently works against more mundane threats as well). Also, I agree with you wholeheartedly that 'Not pissing off the dragon in the first place' is your best possible option. Settings where there are large numbers of high-level casters hanging around are not going to have a problem with this, since it's in everybody's best interests to live and let live. But in a 'default' D&D setting, where casters above level 5 or so are rare for all but the largest cities and your biggest problem is usually bandits on the roads? If there's an angry dragon nearby, best to stay out of its way.

EDIT: Also, it would seem I thought I was in the 3.5 forum. Whoops! I'll have to edit the former post to consider alternative 'power sets' for dragons.

[PhoenixPhyre]

This is very Edition dependent. 1E and BECMI dragons can breathe 3x per day, for damage equal to their current hp. 2E dragons can breathe once every 3 rounds for a set range of damage based on color and age, with no options to "conserve" breathe for a larger area of effect. 5E dragons are subject to Bounded Accuracy, meaning the 3rd-level soldiers will hit roughly 25% of the time (if I understand what I've read about 5E). 2E dragons gain spells randomly, so there is no guarantee they will get Protection from Normal Arrows. Not all 1E or BECMI dragons get spells.

If you focus on the general traits all dragons have in common (flight, good AC, breath weapon, etc.), you will get more generally useful insights.

I'm willing to consider even the worst-case scenario (3.5e rules). But if we do so, then we also have to grant the rest of the 3.5e assumptions (such as common spell-casters/magic items, etc). I agree that moving away from 3.5e the threat diminishes fast (but unpredictably so--different editions are quite different, after all). A 5e castle (with a decent amount of moderate CR inhabitants, such as veterans, archers, mages/priests, etc) is unlikely to be threatened by an adult red dragon unless it catches them unaware--the breath is only 60 ft, the fear is only 120 (and archers have much larger range), and it has no spells/resistances. The other editions I don't know enough to make any claims about.


My apologies. I was assuming 'default' D&D setting, in a situation in which one was likely to see angry dragons (such as the lake town from The Hobbit whose name escapes me at present). With that in mind, having spellcasters on tap is a good counter, as I said, as is a fortified underground defensive position (which conveniently works against more mundane threats as well). Also, I agree with you wholeheartedly that 'Not pissing off the dragon in the first place' is your best possible option. Settings where there are large numbers of high-level casters hanging around are not going to have a problem with this, since it's in everybody's best interests to live and let live. But in a 'default' D&D setting, where casters above level 5 or so are rare for all but the largest cities and your biggest problem is usually bandits on the roads? If there's an angry dragon nearby, best to stay out of its way.

But that's not the 'default' setting (at least for 3.5e, where your rules come from). The 'default' setting (either greyhawk or FR) is loaded with mid/high-level magic and magic items (your average castle would have a mid-level cleric and his assistants at bare minimum, and the guards very likely might have magic weapons). In the default settings as well (FR was the source of this post, as it was the one decried), dragons don't go around raiding castles much (unless it's plot significant). And one of my whole points was that, from a world-building perspective, you're very unlikely to see angry dragons. So building special fortifications to deal with those threats is pointless (and expensive)--it's like worrying about "what if the Mission Impossible team decides it wants to infiltrate our data-center."

There's the concept of the expected loss: It's

(1) <loss> = (risk x probable loss).

The corresponding expected ROI of a change (a better anti-dragon fortification) is

(2) <ROI> = (risk x (adjusted probable loss - status quo probable loss))/(cost).

In the case of a dragon attack, the probable loss is cataclysmic, but the risk is so small that the expected loss is small. Since the maximum benefit (the numerator of (2)) is small, and the cost is large, the ROI is bad. This means that it's rational to have "regular" castles even if there are dragons around.