Eberron - Scarred Elf Campaign

You Always Need A Map

One of the impediments I encounter to playing more frequently is the amount of time it takes to draw decent maps.

I had thought that the issue arose from the highly tactical combat engine associated with 4th edition D&D. Pretty much every scene in every published adventure has a – usually small – battle map, in full colour, with most of the relevant details placed right on the map. These maps are usually about 8 × 10 squares, or about double that size if large creatures are involved, and involve a lot of heavily contrived terrain features.

For a number of reasons, those maps don’t sit well with me, and don’t suit the way my group plays the game. To start with, they are not nearly large enough. My players are not content to be narratively railroaded into the “start” position on the map: they want to move cautiously through any area, and control their approach.

Moreover, battles often spill off the edge of the map, into corridors in a dungeon, or into the wild world in a town or the wilderness. NPC’s try to run for reinforcements, or PCs need to make a strategic withdrawal (and many monsters won’t stop at the door in pursuing them). When I play in my friend Brent’s campaign, maps that are 100 or 200 squares on a side are not uncommon, and we often use the whole map.

In addition, the maps need a fair number of terrain features to make them interesting, and for my own verisimilitude those features have to be naturalistic. Many 4e published maps are filled with contrived obstacles like acid pits, lava pools, and the like. Moreover, many of the mundane features are inappropriately scaled, like roads and bridges that are too narrow for carts, or rooms that are too large for their purpose and difficult to fill with appropriate furniture.

I run my games in MapTools, which theoretically allows for infinitely large maps. That part suits my purposes just fine, but drawing those maps takes a lot of time. Dungeons are not bad, since there are inherent constraints. I can put an entire dungeon level on a MapTools map. Once the tokens are placed it is obvious what rooms will naturally support each other for reinforcements, and if a battle spills out into a corridor, I know exactly what that corridor looks like.

But wilderness encounters or even urban encounters are not so easy. Wilderness encounters require the placement of interesting terrain features organically over a large area, and urban encounters require mapping out a significant portion of a neighborhood (to say nothing of if the PCs decide to randomly look for a building to duck into). That is a lot of work. I’ve begun making large wilderness and city geomorphs to mitigate the problem of running off the edge of the map in play, but city geomorphs in particular take time to produce and most of them are largely devoid of detail because I just don’t have time for it.

(Wilderness geomprphs are easier because Wolph42’s Bag of Tricks has a nice randomizer for MapTool that lets me populate an area with trees and terrain features).

So I have been looking for shortcuts for a while. Most of my players are still suffering from post-killer-GM-stress-disorder, and entirely distrust that the world that is described by the GM will have any relationship to the world that is actually going on in the GM’s head. They assume that at any moment combat can break out, and at least one of them feels a need to have a good idea of the physical layout for even the most innocuous social interactions.

This was not an issue thirty years ago, when we were playing as teenagers. By and large we played in dungeons of one sort or another, which are much easier to manage. We also had a lot of time on our hands, more than enough to draw lots and lots of detailed maps.

But the other reason we didn’t run into the problem was that we did not have the sort of unconstrained play that I am currently going for. For example, right now, Koln is in a small city, and he knows his sister has been kidnapped and is being held somewhere in that city. He has several clues about her whereabouts, but how he chooses to interpret those clues and where he goes for information is entirely up to him.

And because Koln can go anywhere in the city, and talk to anybody, and fight them or chase them or run away from them, it makes it very hard to plan encounters and have a battle map ready for each one. I really needed to find a way to abstract things. But now I am back to detailing my geomorphs. And, sadly, this is a unique city and I won’t be able to use what I create here anywhere else.

Since it appears that I am not going to be able to abstract encounters, and am going to need to detail every street, plaza, market and throneroom the PCs could possibly enter, I am moving to Plan “B”: how to simplify my maps and make them more quickly. But perhaps that is a topic for another post.

A Realization

I’ve come to realize that in a very real way I am no longer playing 4th edition D&D.

I don’t mean I don’t follow the rules. I very much enjoy the balance and (for an encounter- and monster-designer) predictability of 4e’s rule set. And when I make houserules, they are nearly always a novel application of an already existing rule, with careful attention to game balance. No, what I mean is, the way I run my games is philosophically very different from the way I think a typical 4e game is run.

One concrete example of that difference relates to a difference in the relative merits of “crunch” and “fluff” in the game. In 4e, the flavour text of powers or other rules elements is expressly irrelevant to how they function. This can lead to a disconnect between the narrative depiction of the functioning of a power and the effects of that power in particular circumstances.

For example, a character has trained in a particular move, a sweeping blow that knocks the legs out from under a creature. Mechanically, the creature is now subject to the “prone” condition. So, how does this work on a gelatinous cube? The rules tell us that in these circumstances, the cube is not physically prone, but is “writhing or unsteady, rather literally lying down.”

I have two issues with this. The first has to do with the functioning of the condition. Among other things, the “prone” condition grants the cube “a +2 bonus to all defences against ranged attacks from attackers that aren’t adjacent to it.” That makes sense for a humanoid, who is lying flat, and makes a smaller target from range. But it makes very little sense for a 10′ × 10′ × 10′ creature. Knock it over, and it is still the same sized target.

So, maybe you squished it flat instead? But then there is a disconnect with the narrative – remember, the warrior trained in this leg-sweeping move, not a flattening move. So instead, we narrate that the character used an entirely different move, doing so proficiently despite the fact the PC has never encountered this situation before.

That may be fine for some people, but it isn’t for me. The narrative continuity of the world has established that the PC has trained in a particular move. He can use that power, as opposed to other powers that he does not know, because of that training. The physics of that power are now baked into the game. I dislike changing the physics, or narrative continuity, out of slavish adherence to game mechanics.

Now, it is possible that with sufficient application of imagination, I can come up with a solution that resolves all of these problems at once. But I haven’t yet, and certainly am unlikely to do so on the spur of the moment at the table, for this and a multitude of other similar situations that can arise.

A similar issue arises from the fact that the mechanics of a power do what they say and nothing more. Unlike the mechanical-narrative-game physics disconnect, this is more of a playstyle issue. As an old-time gamer, I prefer a game where the characters have the ability to resolve problems with the creative application of their existing abilities.

If there is a fire, and a PC has a power that is described as creating a whirlpool of water, I want them to be able to use it to put our a fire even though the mechanics do not specify that “any flames in the area are extinguished” (with a caveat that it doesn’t apply to zones that do fire damage).

My solution? The descriptive text tells us what the power actually does; the mechanics, on the other hand, are merely the recommended method of adjudicating the effect of using the power, which the DM can vary in appropriate circumstances.

For this to work, it is essential that both player and DM understand, and are in agreement respecting, the way a power works in the physical game world. Nothing is more frustrating for a player than when a DM vetoes an action that clearly works in the player’s mind. And the DM should not lightly interfere with the mechanic, but only in the clearest of cases, especially if the player has a reasonable explanation as to how the PC adapts to the circumstances.

On the other hand, you are not limited to the existing flavour text, much of which appears to have been written without a great deal of thought. Any physical description of the effects of the power that matches the mechanics should be acceptable, as long as both player and DM agree on it.

So what happens if the PC hits a gelatinous cube with a proning power? Well, with the power described, I can understand how damage to the base of such a creature would unbalance it until it rights itself with a move action. And I understand how that could make it harder for the cube to move, or attack, or defend itself. So you could retain all of the mechanical aspects of the “prone” condition, except the defence against ranged attacks.

Note, if the attack was against an ochre jelly, which is low to the ground anyway, slicing it up a bit might make it a harder target to hit from range. Maybe I’m reaching, but I would like to rule in favour of the mechanic working as designed whenever possible. And if it was some sort of shield bash attack against fortitude, instead of a leg sweep, I might rule that the cube had been flattened somewhat. Like I said, only interfere in with the mechanics in the clearest of cases.

All of this is contrary to the philosophical approach to 4e, as expressed in the rules. If my interactions with other 4e players on the internet is any indication, my approach is anathema to the average person who still chooses to play that edition of the game. But it is absolutely necessary in order to play the sort of game I want to play.

Dealing with Character Death

I swear I drafted this before the Angry GM wrote his article ”Death Sucks”, which includes the system he calls the “Scars of Peter Molyneux”, I just hadn’t got around to publishing it.

The reason I still play 4th edition, instead of trying to coax my players into trying 5th, is because I like the combat engine so much that it makes more sense for me to tweak the non-combat portions of 4th edition to make them act like other editions, than it does to tweak the combat rules of another edition to suit my tastes.

Part of the reason this is easy to do is because the non-combat portion of 4e are nearly non-existent, which makes it easy to create rules without messing much up. Another reason is that, if you know a game element really well, you can modify that element with a reasonable degree of certainty that it will not mess up game balance.

There are sacrifices, however. Character generation is a pain in the ass, and difficult to do quickly or on the fly. This is only exacerbated by my use of programmed tokens in MapTool, which greatly speed up play but can take quite some time to prepare. Creating a low-level PC token and testing it can take about an hour, for example.

As a result of this (as well as the fact that my plots are character driven), character death is a pain in the ass. Oh, I will happily kill off NPC allies, but I am very reluctant to kill PCs. And the problem is exacerbated because the availability of Raise Dead is very limited in Eberron. Most priests are not clerics, those that are only cast spells for the faithful, House Jorasco is expensive, and even if you can find someone to cast the spell, on Eberron it often doesn’t turn out quite right.

However, I still want there to be consequences for screwing up. So, there came a time when a character I didn’t want to die would have died by the rules, and I decided to put that character in a coma instead. I figure, if the rules would have killed a character, I can do nearly anything I want to the character short of killing it, and the players can’t really complain.

After that, I started thinking of all sorts of interesting things I could do to characters short of killing them. Taken down by a power that blinds, and fail your third death save? You are permanently blind. Same thing happens, but with a power that slows or immobilizes? Your leg is broken: you are prone and cannot stand without assistance.

Some of these injuries will heal on their own eventually, and some are permanent. Since the “dead” state could be cured by an 8th level “Raise Dead” ritual, in order to make this change no more onerous than the existing rules, I needed fixes for these problems that were no more expensive than “Raise Dead”.

And “Raise Dead” as an 8th level ritual? That always seemed a bit low level for such powerful magic. I mean, in other editions it is available around 9th level, so in 4e I figure it shouldn’t be available until low paragon tier at least. I ended up making it a 14th level ritual.

So suddenly I found myself with a reason to create rituals like “Cure Blindness/Deafness”, which existed in previous editions. And magic prosthetics to replace severed limbs.

I haven’t had a chance to use it since (subsequent character deaths could not plausibly be turned into anything else), but I think this solution will add a lot of nice potential story elements. D&D is short on stories that can be told around injuries or amputations, unless you use one of the rather lethal critical hits tables that were in vogue for a while.

I was content to just go with that for a while, but my tinkering with diseases, poisons and long term status effects inspired me to add another layer. More on that later.

Koln Campaign, 992 YK, Shadow Marches
or, Why I Don't Use Skill Challenges

With the prologue over, it was time for Caius to have an opportunity to play Koln. I figured that after being injured by vampires, Anika would return to her family in the Shadow Marches. Since I like a character-driven game, I wanted to really root Koln in the Shadow Marches and Tharashk culture, create a sense of attachment to his family.

This session failed to accomplish that goal.

Oh, it started out well enough. I was rather proud of the session introduction, which I think linked the prologue to the main story in the player’s mind but told nothing whatsoever to the character. But it quickly went downhill from there.

You see, the session was supposed to be Koln’s test of Siberys, where he obtained, or failed to obtain, his dragonmark. It is supposed to be a test that brings out extreme stress, which causes the dragonmark to manifest if it was going to do so. And I thought that the best way to do that was to ditch Koln in the woods and force him to find his way home.

But it was a mistake, both in conception and execution. The first problem was that Caius did not have enough reference points to engage in the culture in a meaningful way. So there was a druidic ritual, which he really just sat through and should have been narrated as exposition rather than played; followed by the test of Siberys which, if Caius had already been immersed in Eberron, would have felt like a matter of great significance; but for a player who was not yet familiar with the setting, it was a poor introduction.

Then there was the resolution mechanic. The test of Siberyis is a non-combat encounter, or series of encounters. According to the 4e rules as written, it cries out to be a Skill Challenge. But it suffered, as many Skill Challenges suffer, from trying to cram more checks into the challenge than really ought to be necessary to achieve the goal.

Koln needed to use his skills to get home, but he also needed to use them in a way that would trigger one of the powers of the Mark of Finding to manifest. And there were really only a few skills that I could think of that would trigger any given power (and Caius didn’t come up with anything different). Which meant if he wanted to succeed in the challenge, he basically had to spam checks. And as a GM, I had to narrate spammed skill checks into something interesting.

I think the player enjoyed it well enough (particularly because of the combat encounter, which I will get to later), but it left the GM wanting to blow his brains out. And it underscored the basic problem that I always run into when creating skill challenges, which is that most interesting non-combat obstacles, or even series of obstacles, do not predictably require 4-23 checks to resolve.

I underscore “predictably” because Skill Challenges contemplate that the GM know how many successes are required to accomplish a task, and of what type, which I feel imposes an unneccesary restriction on gameplay. Players need to be free to find their own solutions to problems.

Checks ought to be made when the GM is adjudicating the results of a specific action attempted by the PC, and the number of checks should be determined by how far the GM thinks that specific action would take the player toward accomplishing his goal. Moreover, the type of check should be determined by the nature of the solution that the PC attempts.

I underscore “interesting” because often there are only a few skills that may realistically be applied to a situation, and the GM therefore needs to either allow spamming of the same skill, or to shoehorn in primary skills that really are of questionable relevance.

Like convincing a national leader to listen to your argument by performing a feat of athletic prowess in front of him, and thereby convincing him that you are a formidable fighter and somehow an expert on tactics and strategy. Because kings respond to nothing so well as a diplomat who breaks off negotiations to start performing one-handed push-ups.

This is what I do now. I create an obstacle with a solution that is known to me. I figure out how many successful skill checks would be necessary to overcome the obstacle. For every four checks using my hypothetical solution, I assess experience for an at-level monster, and that is the amount of experience the encounter is worth. The players then chooses whatever actions they wish, and I adjudicate those actions as necessary. I don’t even track successes, although sometimes I track failures. If the party accomplishes the goal, they get the experience regardless of what skills they chose or how many successes they achieved.

Of course, in practice, that really happens, because I rarely create a situation where more than three checks are required.

Anyway, getting back to the session, from Caius’ point of view it was not a total bust, because at one point Koln failed a check in the Skill Challenge which triggered an encounter (don’t tell me 4e doesn’t have random encounters). Koln fought a spiretop drake when he wandered to close to her nest. Koln then decided to go looking for her eggs and found them (by a remarkable coincidence, they were worth exactly the amount of a treasure parcel of Koln’s level).

Caius decided that Koln would keep one of the drake eggs, and train the hatchling. I updated the spiretop drake’s stats to MM3 levels and made it into a Companion Character, replacing the CA mechanic with Hunter’s Quarry, and thus was born one of Koln’s most enduring allies.

That damn drake is bloody effective in combat, other than the fact that he doesn’t stick around to draw his fair share of the attacks; there has been more than one battle where it was all down to the drake, and he pulled through. Generally, about halfway through a given battle, the monsters realize how dangerous he is and try to focus fire on him, but because he is flying only the ranged monsters can even attempt it.

In fact, I have actively tried to kill him off on more than one occasion, to start a subplot I have cooking, and I haven’t even come close.

Oh, and Koln also got his dragonmark. So now he always knows which way is north.

You can find the session summary here (the italicised portion is a modified version of my notes for the session introduction).

Koln Campaign, Prologue

First, a couple of comments about rebuilding Koln in 4th Edition. When I started the Koln campaign, Caius had only played a couple of sessions with the 4e rules, and I was not much more experienced with them. As a result, mistakes were made.

In his 1e incarnation, Koln was a fighter. Period. Even though Caius had developed some pretty decent houserules regarding multiclassing, I don’t think Koln ever picked up a second class. Moreover, it was a big part of his character conception that Koln should wear plate armor. And in 4e, most fighters don’t get plate without a feat. But Koln was also supposed to have a dragonmark, which took up his level 1 feat slot. So if he wanted to be a fighter with plate at first level, his only option was to play a knight.

If I was to do it now, I would give him a dragonmark within the story without making him take the Mark of Finding feat. I was already giving him the Know Direction cantrip using the Dark Sun rules, so I could have made it work within established cannon, and built a regular fighter with plate proficiency. But I didn’t know the system well enough to know what things I could mess with, and what would have unintended consequences later on, so I stuck to the rules.

When I was designing the first session for Koln, I wanted to get into the back story. There was a part of the back story that I wanted Caius to know, but didn’t want Koln to know. And the backstory was interesting enough to play, except that Koln was not yet born when it was happening. So I decided the first session would be a prologue.

I had Caius play Lubash Tharashk, a dragonshard merchant in Karrnath. I built Lubash as an 8th level berserker, because: (a) I wanted to give Caius a taste of what characters could do at a higher level; (b) Lubash was an half-orc, and would traditionally be a type of barbarian; and © the berserker had a defender’s aura similar to the knight, which would give Caius an opportunity to get used to the mechanic.

The story I was going for was, Lubash is travelling in a coach with his pregnant daughter, who is crying. The coach is attacked a vampire, the coachman and horses are killed, and then the vampire attacks Lubash and his daughter. Lubash fights it as best he can, but dies, and the last thing he sees is his daughter being attacked by the vampire. So I start with a good fight, and a mystery, and then jump ahead to the main story with Koln as a young man.

The problem is, Lubash, as run by Caius, proved to be an able fighter despite being quite under-optimized. The vampire was an 8th level elite monster, and I expected it to do very well against Lubash and his pregnant, equally suboptimal, second level ranger daughter, particularly because I knew Caius would use the defender mechanic rather than the striker mechanic in order to protect the daughter. However Caius has a knack for strategy, and using the daughter as a flanking-buddy, Lubash beat up the vampire pretty badly. The vampire was forced to flee.

So, now what? Well, the vampire has a healing mechanic using his blood drain power, and they are in a town, so the vampire goes hunting a nice family of human minions. Screams ensue, Caius figures out what is going on, so Lubash and his daughter try to find a place to make a stand. Fortunately I had an interior map for an inn handy, so I dropped that onto the VTT map in time for the vampire to track them down. I didn’t want to repeat the vampire-flees-and-recovers-by feeding-on-human-family thing all night, so I added another vampire buddy.

From that point on, the timing of the combat worked out nicely. Lubash died on cue, just in time to see the second vampire slashing up his daughter’s face.

Railroady? So what? Lubash was built to die. Caius had no attachment to him, and it didn’t really remove his agency, because he got to control how Lubash died. Caius got some backstory, and enjoyed himself much more than he would have if I had given him a sheaf of notes to read instead. He became really invested in the character of the daughter, which would prove to be important later on. And he had a free trial run with the defender’s aura mechanic. Win-win.

For those who want the back story to the back story, here is the spoiler. Lubash is Koln’s grandfather, and the daughter is Anika, Koln’s mother, who is pregnant with Koln. The whole family are half-orcs, most of whom can pass for human. Anika left House Tharashk in order to marry into the ir’Koln family, as required by the Korth edicts. The ir’Kolns are a Karrnathi noble family with substantial holdings. King Kaius disapproves of the marriage because he does not want a dragonmarked house gaining influence in Karrnath through strategic marriages.

Kaius pressures ir’Koln to call off the marriage, but ir’Koln is in love and he refuses. So Kaius forces the issue by turning ir’Koln into his vampire spawn, and compelling ir’Koln to obey. Ir’Koln has no option but to obey, but he is still enamoured of Anika. Having lost his conscience in the transformation, he gives in to his base instincts. If he can’t have her, he will fix it so no man will want her. He creates his own spawn, and orders them to find her and disfigure her face.

So now we have the elements of the Koln back story. Koln has lost not one, but two birthrights: the first when his mother leaves House Tharashk; and the second when his father fails to marry his mother. What is more, according to the Korth edicts, he cannot claim both. He can either ask to rejoin House Tharashk, or he can try to claim his noble title.

You can find the session summary here (the italicised portion is a modified version of my notes for the session introduction).

Star Wars: The Force Repeats

My eleven year old daughter tells me I shouldn’t be so hard on the new Star Wars movie, because at least it teaches people to recycle.

Wandering Monsters

Earlier today I was reading the Angry GM’s most recent article about traps. It was an excellent article, as usual, and I agree with pretty much everything in it. However, also as usual, it got me thinking of a lot about issues related to encounter design.

There is a point in the article that, in order to discourage pixel-bitching, there needs to be a consequence to taking the step of looking for traps. And before he even said it, I was already thinking about wandering monsters. Because how and when to use wandering monsters has been very much on my mind lately.

In the old days, pretty much every dungeon included a table for wandering monsters, and the DM would check periodically (generally every turn, or 10 minutes) for the appearance of a random creature which may or may not have some relationship with the dungeon ecology (or lack thereof).

It was time-based mechanic, and because it was a time-based mechanic, it does not work well in the 4e system without adaptation, because 4e treats time very differently than 1e did.

Everything in AD&D took a lot longer. In general, exploration (travel) took about 10 times longer (in-game), and exploration (searching) took 30-40 times longer (also in-game). Combat is about the same, but only if you include 4e’s short (5 minute) rest as part of the combat duration. (Read the AD&D DMG and the 4e Rules Compendium if you want details; summarizing in this way allowed me to cut six long paragraphs of analysis that I doubt anyone wanted to read.)

The long and the short of it is, both because of changes in the system and changes in the way I want to structure my games, time is no longer an appropriate mechanic for organizing random encounters. What I want to do instead is link random encounters to the choices that players make. That is, I want the chance of a resource wasting encounter to be the consequence of players choosing to take an action rather than forego it.

Take searching for traps. In game, this slows down character movement through the dungeon and increases the likelihood of some creature stumbling upon the party before they get to the next room. Out of game, it breaks the narrative. Both as a DM and as a player, you only want the party to be looking for traps when there is a reasonable chance that traps might be present; you don’t want them checking for pits every 5 feet, for example.

So, if the party checks for traps, perhaps you should check for a wandering monster. This makes the decision to look for traps a wise one if the party has reason to believe that there might be a trap present, and a foolish one if there is no indication that a trap might be present. A corollary of this is that the presence of traps should be telegraphed in some way; look at Angry’s article if you want clever ideas as to how to do this.

The same logic applies to searching for secret doors, so let’s turn instead to opening locks. Opening locks is a bit different because it can be perceived as not being a real choice; however, this is only true if (a) there are no other options available for exploration in the dungeon (you have explored every open or unlocked room on this side of the door); and (b) the door is too strong to force.

If the door is not too strong to force, then you have the option of attempting to force it. If you fail on the first attempt, you have made a bunch of noise and alerted any creatures on the other side of the door. If you succeed, you surprise any monsters on the other side, but you still make a bunch of noise and may alert other monsters in the area.

If you attempt to pick the lock instead, you have a much reduced risk of alerting whatever is on the other side in the event of a failure. This means that if there are no other options for exploration on this side of the door, choosing to pick the lock rather than force the door becomes the default choice unless there are also consequences associated with the choice to pick the lock.

This is where wandering monster checks come into play, because when the PC chooses to pick a lock, there is a risk of a random encounter. Moreover, if they fail to pick the lock, and have to try again, they run that risk a second time.

Random encounter checks are therefore linked to the decision to take an action rather than the time it takes to complete that action. This is not as much of a departure as you might think, since time in D&D was always an abstraction, and since most of the significant out-of-combat actions in AD&D took ten minutes anyway (for example, a cursory search of a 20’ x 20’ room, or searching a 10’ x 10’ area for the trigger to open a secret door), which was generally enough to trigger a random encounter check.

Now let’s look at short (5 minute) rests. A short rest is the default action to take after a combat, in order to recharge encounter powers and regain lost hit points. As they are generally used, there are no meaningful choices associated with the decision to take a short rest. Even if one introduced a random encounter mechanic, the risk of proceeding without regaining hit points or encounter powers would far outweigh the relatively small chance of a random encounter.

But what about the decision to take a second short rest? Suppose the party healer is a sixth level bard. If she uses her song of rest, then for each healing surge a party member expends, he gains additional hit points equal to the bard’s charisma. But if the bard instead takes a short rest and uses Majestic Word, for each healing surge spent the party member gains additional hit points equal to the bard’s charisma plus 1d6 hit points. Without some sort of time pressure, the optimal choice to make is to take several short rests in a row, and spam Majestic Words.

So perhaps the first short rest is free, but additional ones trigger a random encounter check.

I would like to play with that a little. But first, let’s talk about one of the things I do with encounter design. When I have a number of rooms, not too far apart, with allied creatures, and the PCs encounter the creatures in one of them, I assume that team monster will send for reinforcements as soon as they realize they are in trouble (generally partway through round 2). If the PCs fail to prevent monsters from escaping and summoning reinforcements, they could find themselves fighting a series of encounters with no short rest in between.

As a result, I tend to design connected areas as though they were a single encounter, fought in waves, and I set the difficulty for the entire area at the extreme edge of what the PCs can handle – usually double the XP budget for an at-level encounter. Now, they only face this if they fail to shut down the call for reinforcements, so if they play their cards right, they could instead be facing a series of easy encounters, and take very little damage from each of them.

As it stands, despite easily overpowering the easy encounters and taking little damage, there is little incentive to skip the short rest and press on to the next encounter. And on the one hand, good for them if they have managed the encounters well.

But on the other hand, from a narrative perspective it makes sense if allied creatures check in on each other from time to time. And maybe the party should make decisions about setting up defences while it is resting. And maybe there should be a real choice between pressing on with surprise, or risking discovery during a rest. So maybe there should be a chance of random encounters – or rather, a random chance of an encounter, since the creatures involved in the encounter are likely to be the allies of the creatures that were just killed. So maybe the first short rest shouldn’t be a freebie.

I still am not sure what I am going to do with this. For now, I think I will consider employing it on a case by case basis.

The Koln Campaign

I thought I would talk a bit about the Koln campaign.

Back in the 80s, my friend Caius ran a campaign on his homebrew world that included a significant NPC by the name of William von Koln. (This was, by the way, the campaign in which I ran my favourite character, the name of which eventually became my internet alias).

Koln arrived on the scene on the very first session, and stayed with the party more or less throughout the run of the campaign. There was a period when Caius and one of the other people in our group were switching off as DM from time to time, so from time to time Koln was run as a PC, but he never really lost that NPC feel, and most of the significant events for Koln occurred off-screen.

Koln was a half-orc fighter who could pass for human. He had been dispossessed of his birthright, and his major motivation was a desire to reclaim it. I can’t remember exactly what his birthright was, if I ever knew, but I think he overshot the mark to an degree, because he ended up conquering the entire of alternate-universe Germany.

The campaign was a long time ago, and I have forgotten most of it, but we must have played White Plume Mountain at some point, because Koln ended up with the soul-sucking sword Blackrazor. This was a big point of contention between Koln and my character, Beoric, because Beoric objected to the use of the weapon for that purpose, and where and when Koln used the sword drove a lot of the interaction between the characters. Koln and Beoric were friends and allies, and Beoric often struggled with Koln’s amorality.

Koln always wore green plate and a horned helm, and had an aspect of the wild hunt about him. When he was a PC, the other DM gave him the ability to shapeshift into a wolf, for what reason I have never been clear.

I think Beoric started out as a houseruled ranger/thief. The thief component was added because in those days (the campaign started around 1984) it was the only way your woodsy ranger could stalk prey or climb a tree. He was a half-elf because only non-humans could multiclass. What I really wanted was to make him a gish, and Caius eventually made a houserule for adding classes which worked reasonably well, so I added magic-user (that’s “wizard” to you young ’uns) to the class mix.

We travelled with Donovan, a duelist who eventually multi-classed to duelist/assassin; Richard Kaylen, a fairly combat focussed cleric; and Raymond, an irascible old magic-user. There may also have been a paladin or paramandir called Edmund Radon, and a crazy cleric named Baltor, in the party; I can’t remember if they were from that campaign or another.

Anyway, after Caius had played a few sessions as Gaius Cannith, and started to get used to the fourth edition system, he asked if we could reboot Koln, in Eberron, as a full PC. Oh, and could he please have a dragonmark. I agreed, because I thought it would be fun to run a campaign where I already knew the PC very well and could easily predict his choices (and therefore avoid a lot of wasted prep).

So this is what I had to work with. I knew Koln was a half-orc, and that he had been dispossessed of his birthright. He needed a dragonmark, which was by default the mark of finding, which meant he was connected to House Tharashk. Oh, before I go on…

CAUTION: Spoiler warning! Use your mouse and click and drag over the block of gray to reveal the spoiler.

As a member of House Tharashk, the easiest way for Koln to be dispossessed is for him, or his parent, to be excoriated from the House. But that did not fit with a big part of Koln’s story, which is that he reclaimed his birthright by force of arms. So I conceived of a backstory to allow for that.

Koln’s mother fell in love with a member of the Karrnathi aristocracy. Under the Korth edicts, she had to give up her membership in the House in order to marry him. Koln was conceived during the engagement, but his father broke off the engagement before Koln was born. As a result, Koln was dispossessed of two birthrights, his noble title and as an heir of House Tharashk. Since he can’t have both, he will eventually have to choose between them.

This fits with my general approach to running a campaign for Koln. Koln’s alignment was neutral, but he tended to move back and forth between good acts (he was loyal and trustworthy) and evil acts (like using Blackrazor). Since as an NPC he didn’t have to face a lot of tough choices, as a PC I want to really stress the tension between the differing aspects of Koln’s personality, and his conflicting desires.

For example, there is a conflict between Koln’s desire to be a leader of men (which he can pursue if he chooses his aristocratic heritage), and his affinity with nature and the hunt (which is strongly associated with House Tharashk). Likewise, I have given him two weapons, one of which is a soul-sucking equivalent to Blackrazor, the other of which is the opposite.

I also wanted Koln to have to face the sorts of choices that he did not previously face, either because (a) he was an NPC, or (b) we were playing this in our teens. More than 30 years later, I want the reboot to have a level of nuance and sophistication that we were not capable of back then.

So I gave Koln a family. Moreover, I spent considerable effort in characterizing his family members, and gave them a big role in his life. I mean, the bulk of his motivation is related to his birthright, so shouldn’t family be important?

Nor are they just McGuffins; most of them have a significant role to play in many of the encounters. I have really enjoyed the interaction between Koln and his little sister, as well as Koln and his mother. The scene where Koln’s mother’s secret was finally revealed to him was particularly poignant.

Assuming Koln ever makes it to Karrnath, I am really looking forward to him meeting his father. Whom he will eventually learn is now a vampire, with no intention of ever giving up his title to Koln.

I have had to make significant changes to Beoric, now that he is not a protagonist. Previously, Beoric’s defining characteristic was leadership, and I can hardly having him drawing up battle plans for Koln, or directing him in combat. The result is that he is a lot less earnest than the original incarnation, since he needs to be fairly relaxed about letting Koln take command. In terms of class, it is much easier to get what I want in 4e than is was in 1e, as a result of which he is now a Cunning Bard, on his way to becoming a Resourceful Magician.

Descriptive Text: Deconstruction

I was thinking about the post I wrote yesterday about descriptive text, and I thought I might post an example using Alexis Smolensk’s descriptive text as a starting point.

I should start by saying, I don’t actually think that Smolensk runs his narratives like the example on his blog. In the single, short video I have seen of him running a game, he did not. I know he doesn’t use the first person “we” in his narratives, because I asked him. He said the intention was to write the narrative as though the player was thinking it.

The blog is a written medium, and he obviously enjoys written language and is good at it. In fact, if he had written the description in effective oral language, it might have undermined the overall point of the article.

However, I see examples of this sort of descriptive text all the time, from GMs with influence, and I suspect it leads to the impression this this is appropriate to oral narration in your game. I would like to defray that impression.

I think the important information Smolensk is trying to convey is this:

  • It is night.
  • You need to find somewhere to rest.
  • It is raining, which makes you cold, wet, and miserable, and you do not want to sleep in the open.
  • The shantytown outside of the city walls does not have any obvious inns.
  • There is a gate to the inner city ahead of you.
  • The gate is open, with a raised portcullis, but it is guarded.
  • There are people travelling ahead of you, who take a turn immediately before they get to the gate. (Note he does not say what direction they turn, which, given the position of the gallery, may be important. He should probably also make it clear whether those people are the same as the people who are travelling in the gallery. I think not, but I could see players being confused.)
  • The guards have noticed the party

In addition, he is trying to convey the following additional information, which may or may not be important:

  • There is a gallery to the left of you. He did not say it was raised, so I am assuming it is at or near street level, like it is in the accompanying picture.
  • There are some people travelling along the gallery in the same direction as you.
  • The guards are carrying a yellowish-green light.

Note this is quite a bit of information for players to absorb all at once, particularly if it is conveyed orally. I would probably break it into chunks. Start with the need for shelter, and the ramshackle outer city; when they are having trouble finding an inn, then have them stumble on the gate.

If any of the information above is a red herring, I would leave it out. It is hard enough for players to catch all of the relevant information a GM is providing, without also providing irrelevant information. Irrelevant information is a barrier to communication.

Another thing I want to suggest, is that every object or feature, or group of objects or features, should have its own sentence. The pause that follows a period (and make sure you do pause) gives the players an opportunity to absorb the information and fix the feature in their mental map of the area.

If I was to set the scene that Smolensk is describing, and I had the time to plan it prior to the game, it might look something like this:

Step 1: Setting the Mood

At this stage you can effectively ignore everything I have said so far, because here you are conveying mood and need, as opposed to locating objects in space. So, to start, I would plagiarise Smolensk in part:

  • Night has fallen, and you are tired. It’s been raining for the last two hours, and that has put a hard finish on the day. It would have been hard enough to travel in the dark without the rain, but now the clouds have blotted out the stars and heavy mud clings to your boots. You do not want to sleep in this.
Step 2: Describe the Outer City

Now we start describing the physical space:

  • The road you have been travelling on becomes a dark lane in a strange city.

Note this is still really mood setting.

Also note that if your players are like mine, they might object to you immediately placing them in the city, paranoid creatures that they are (I am still trying to cure them of the paranoia taught to them by past GMs, but that is another story). They might not want to take the main road, and instead scout through the shantytown, or stay out of the city entirely and try to find a farmhouse. I am less likely to invoke their paranoia if I let them know when the buildings first become visible, and avoid triggering their fear of being railroaded into an unwinnable situation.

So if you choose to skip ahead like this, be prepared to back up and let them make another choice.

Now let’s transition between mood and physical space:

  • The buildings are cluttered and ill-kempt.
  • From what you can see the side streets are narrow, twisted and dark.

The last point was added so the players know a bit about the environment if they choose to get off the main road. It could be left for player questions, but if you make something look unpalatable after they ask, it can look like there are rails even where there are not.

Note those two statements would probably count as a single sentence using the Angry GM’s method.

Now we underline the need:

  • You see no signs of an inn.
  • But there must be some in a city this size.
  • You think you may have better luck if you go deeper into the city.

Now, pause, and let your players decide to go deeper into the city. Or not, as the case may be. If they do follow the road, go on to describe a new scene.

Step 3: Describe the Gate

Start general, and move to more specific or urgent elements. Let’s start with a reminder of the darkness and the rain, because they are related to mood, and because they have an impact on visibility, and combat conditions if it comes to that.

  • It is still raining.

Then move on to the most prominent feature, the gate:

  • You see a gate ahead of you.
  • The gate is open, with a raised portcullis.
  • Three guards stand in front of it.
  • Each guard carries a yellow-greenish light.

Again, this would count as two Angry sentences. Note that while “before” is more elegant than “in front of”, I chose the latter as being easier to process.

Then move to the gallery:

  • To your left, there is a gallery, running parallel to the street
  • You can see several figures travelling along the gallery, in the same direction as you

Many players will assume the figures are malevolent. Placing them early in the description lessens their importance from likely assassins to likely witnesses.

Now move to the possible alternative entrance:

  • A pair of figures are walking on the road between you and the gate.
  • They turn [left/right] before they get to the gate and disappear.

Since these figures appear later in the description, they take on a greater importance. Note I added a direction that they are turning, because if they turn towards the gallery, the players will assume they are interacting with the people in the gallery. I would probably not comment that there is another way in: I leave it to the players to come to that conclusion, or not. For the same reason, I would not include the comment that they “might be citizens”. .

Finally, the most urgent issue:

  • The guards spot you, and wait for you to approach.

This is still longer than the certified Angry Gm™ method would suggest it ought to be. I don’t know that I would actually have all of that going on at once; I might break it up by giving the players an opportunity to react before the figures disappear around a corner, or the guards notice them. But if I did it all at once, it becomes even more important to convey it in short, easy to digest pieces.

So let’s try this instead:

  • It is still raining.
  • You see a gate ahead of you.
  • The gate is open, with a raised portcullis.
  • Three guards stand in front of it.
  • Each guard carries a yellow-greenish light.
  • To your left, there is a gallery, running parallel to the street
  • You can see several figures travelling along the gallery, in the same direction as you
  • A pair of figures are walking on the road in front of you, heading in the direction of the gate.

Which is plenty of information. Maybe too much; I’m tempted to leave out the reminder about the weather, but I feel weather is ignored all too often. Pause for player action, if any (Smolensk’s players apparently stop their horses), then:

  • The figures in the gallery continue to move in the direction of the gate.
  • The figures ahead of you turn to the right and disappear before they reach the gate.
  • The guards turn and spot you.
  • They wait for you to approach.

You have ended with the most immediate potential obstacle (the guards’ attention), while giving a rough sketch of what is going on around the characters. The description may not be beautiful, but hopefully it is at least clear, providing the players with enough information to make decisions, or at least ask relevant questions. If you can pull that off, you have done your job in setting the scene.

Descriptive Text

I was reading a recent post by Alexis Smolensk of the Tao of D&D, which was talking about describing scenes.

It begins with a description which is really quite elegant, and then proceeds to discuss how it is difficult to know how the players are reacting to the description, and not to get upset if you can’t see their reactions. If it sounds trite, it is because I am not doing it justice; I am paraphrasing in the most general way, because that is not, in fact, what I took from the article, and it is not the issue that I would like to discuss.

What occurred to me when reading the descriptive passage was, first of all, how good the choice of language was at setting the mood, and second, how poor the choice of language was for describing a space, or the physical actions taking place within it.

I think this is a common mistake GMs make. We tend to have a bit of a literary bent, and we often love language. We want our descriptions to be pretty. But the problem is that, by and large, the techniques for accurately conveying the physical are quite different from the techniques for conveying the emotional, or the abstract, and the better you convey one, the more poorly you are likely to convey the other. Put another way, the mood obscures the action.

The problem is exacerbated because western culture is, at this stage in out history, a written culture, and descriptions in D&D are delivered orally. GMs, I suspect, usually prepare their descriptions using written language, and effective written language differs from effective oral language. Moreover, players are used to absorbing information either through the written word, or through video, which is only partially oral. Absorbing information from oral language is its own skill, and one that has atrophied in western culture.

The Angry GM has an excellent article called How to Talk to Players: The Art of Narration, which creates a nice template for descriptive text, as well as explaining why less can be more when it comes to adjectives. He breaks these rules in examples elsewhere in his website, but then, the website is a written medium, so the same rules don’t apply, do they?

This is the addition I would make to Angry’s article (and if you do any GMing, you really should read it). When you prepare your description, always keep in mind that it will be spoken, not read. If you have developed the skill to speak from bullet points, use those instead of writing it out. If you have difficulty speaking from bulleted points, write out the text, but make sure you write it as though it is to be spoken. Read it out loud to be sure you have it right. It is, in a sense, the dialogue of the narrator.


I'm sorry, but we no longer support this web browser. Please upgrade your browser or install Chrome or Firefox to enjoy the full functionality of this site.