Dungeon World West Marches Q&A Part 5: Did I Succeed? Would I do it again?

In this penultimate installment I answer the most frequent and general question I received: “So, like, how did it go in general?” I judge my success based on whether or not I accomplished what I set out to do, how much fun that thing turned out to be, whether I’d do it again, and if so what I’d change for the next time. I’ve added about 4 sentences for clarity and to give an example but otherwise this is verbatim what I wrote on the story-games.com forum like all other posts in this series.

I. What was I trying to do?

The general goal was always to make “something like Ben Robbins’ West Marches posts describe.”

Different people rank the features in Ben’s portrait of West Marches differently. You’ll often see, for example, people who think they could make a West Marches hex crawl. Ben himself has a split opinion. He stresses that hexes are inappropriate because they make it feel like the world is full of little binary containers but then also stresses that a hex map isn’t actually a deal-breaker for the style. For someone who thinks the exploration aspect of the campaign is key, however, that doesn’t make a lot of sense. So what West Marches is depends where you’re coming from.

My own core take-away is a campaign primarily about a rotating cast of players episodically exploring (and to some degree pillaging) from a centrally located, safe, and relatively mundane home base location in an immersive world that pre-existed the sessions that explore it; with players taking the lead on things actually happening in the game; and permeated by a general sense of ‘hands-off’ GMing and objective fairness. The overall setting of the game rules and world is something like “a place that doesn’t feel invented specifically for play purposes, but that nonetheless offers lots of deep challenge and mastery oriented content centered around players taking the initiative to explore the world.”

I had already given a first shot at that with my last big campaign, so I had a bunch of ideas I wanted to try out to improve on that first effort as well. Beyond hitting all those usual points, my own biggest concerns were making the campaign world bigger and better planned, with more players, wasting less effort so I wasn’t as frustrated, and improving on many meta-game elements I felt were lacking like the players’ attempts at making maps (especially in terms of immersion), and with a greater emphasis on “friendly rivals” style play between players competing to be the best (my prior campaign’s players joked they weren’t just ‘the party’ but were actually ‘the communist party’).

My best players from that campaign had also highlighted some things that I took under advisement when planning my DW WM. I could write a whole post about those, but for sake of brevity the highlights were that navigation could get tedious or was too difficult, the lack of GM-driven play could be daunting or feel aimless, and the overall difficulty of the game made it hard to play just for its own sake — a lack of success often felt more like a failure. The overall difficulty I felt could be addressed by advertising the game as “dark fantasy” at the recommendation of my co-GM (that way it would feel natural to the setting, since high difficulty is fairly non-negotiable itself according to Robbins), the lack of GM direction I thought might be solved with more robust player information-sharing systems (Almanac, map, full size Roll20 game forum, real-time chat room) and more explicit game objectives (in this case the Fame leaderboard and XP system), and the difficulties with navigation might be solved with new aids to help players make their own maps (something those players had been notoriously bad at).

Out of the needs listed above you can pretty easily see how the elements I’ve mentioned in my Dungeon World WM game came to be. The need to reduce wasted prep and increase player-base were both catered to by DW’s popularity and simplicity. The meta-game elements like maps, almanacs, fame leaderboards, and diegetic forum boards could be bolted on to any game system. Another GM could help increase player count and reduce workload. That just left a world that felt impartial, fair, and full of good gameplay that nonetheless felt prior to the sessions experiencing it, which is where the bulk of the work I’ve already described was done on the system modifications and world preparations.

So, overall, those were the sorts of things I was trying to accomplish.

IIa. Did I hit the main goals of West Marches?

The cast of players certainly rotated, and the game was definitely episodic, though the European crew never grew large enough to really break out into truly mixed up parties.

The new exploration-based XP system did encourage exploration in theory and was a great tool to aid world building for the GMs, but it had a natural internal tension: if you got XP for noting down important information about places and monsters, how would you know when you were done without a Doom-style “Secrets Found” meter that would instantly break immersion?

The town was fairly well fleshed out, but still suffered I think from the general problem of “we’ll do that later.” Events in the town often proceeded slowly because other things, like preparing new locations to adventure to, took precedent. And, while some players obviously had beef, overall I think the consensus was the world did in fact feel “hands off,” and therefore that our world building and running techniques worked.

The world was generally hailed as very impressive, immersive, and fair, so I count that as a pretty full success. The same for generally being filled with interesting challenges (the players particularly called out combats I ran as a good example of this).

I feel I did make sure players took the lead on doing things in the game world (though the game world was never passive – it bit back at the players’ interference and had its own agenda when the players weren’t around). They did mention that they still sometimes felt a bit directionless and that the game could have had a lower barrier to entry especially as it went on.

IIb. Did I hit my own patch goals from my prior West Marches campaign?

The short answer is “sort of.” I’d say overall that these were good first-pass fixes, but that they still have a ways to go in many cases. Two that stand out:

Firstly, DW’s simplicity definitely helped reduce GM overload and agitation at wasted effort in some ways, but it wasn’t as much of a help in the quest for players as I had hoped. While my co-GM built more adventure sites than I did, I ended up being the “player handler” GM — I was responsible for recruiting almost every player that ever played a session and I was nearly the exclusive moderator of the meta-game chat and forum.

While the player meta-game is one of the most interesting parts of West Marches to me, this also meant that I experienced an interminable cavalcade of introducing myself and my game to people (often two or three times a week) who ended up ghosting me a few days or weeks later. As I mentioned only about 1 in 8 players who submitted an application ended up as real players in the game. One especially brazen example comes to mind. One guy who played in a couple sessions also played some Hearthstone with me at one point. But when he stopped responding he didn’t bother to block me on battle.net or anything. He was just there playing Hearthstone and just ignoring me right to my face while I asked him if he was still playing in the campaign or not after a month of inactivity.

As someone prone to depressive thinking this constant player churn was soul-crushing work. The occasional conflict between the other GM and I and the fact that he also eventually joined the long list of ghostings didn’t help, either. I think the game was up-and-down largely on the basis of player count and interest, and once the other GM left it was obviously downhill from there overall, even if it did still have a few highlights on the way down like the assault on the lair of the Firebrand.

Secondly, many of the meta-game systems did seem to be a good first-pass solution to the problems I had identified in the previous campaign about things like not making maps or sharing information. But they could have been better in ways the players helped me identify in a post-campaign chat we had a year or two ago.

IIc. What would I fix going forward?

This naturally leads to questions about player convenience in general. All this meta-game stuff and the XP system were a bit of a hassle even for the core players, and this meant even core players sometimes felt at a loss for what to do next or where to find certain information. Often this was because we faced a trade-off between diegetic immersion and utility (“should the almanac be a plaintext Google doc or written on little sheafs of digital vellum?”). The players almost unanimously preferred accessibility to immersion whenever it came up, and made further points (which I agreed with) about this being important for player retention. While you can retain some players merely through having interesting game mechanics, game world, or personal social connection to the GM or other players, it seems like lots of people who play RPGs show up at the same bat time and same bat channel every week because their DM entertains them by putting on a one man show without asking anything in return. As you leave the realm of “core players” who play weekly (or, if you’re Gareth, twice weekly) and enter “marginal players” who play one or two times a month this kind of barrier to play might really get in the way.

Nonetheless I do think the changes to the map in particular – to make it a page on Roll20 in front of players’ faces and to provide them little terrain sprites to help those who felt they “couldn’t draw” – were largely successful. Some players noted that they didn’t feel they received enough of a reward beyond “not being lost” to justify the often-talked-about-but-never-performed solo map adjustment sessions to make the map more objectively accurate for new players’ sake, but in my prior campaign I had awarded the most XP of all for improving the map and nobody ever did it back then, so it’s something of a paradox of player motivation.

Moving into session play now, people generally agreed with me that although the combats were very good they could have been a little sharper and quicker. We agreed that a fight should stick around just long enough to present its “puzzle” (figuring out how to deflate a monster’s strengths and take advantage of its weaknesses), and if the combat puzzle is solved should go away quickly (in practice usually from a barbarian swinging their great-sword in a Hack & Slash that decapitated everything in one fell blow). In the actual events this mostly happened, but a bit less monster HP and a bit more monster damage would’ve probably made it a little better.

There was also a pattern of discussion where gameplay closely matched or only slightly deviated from players’ expectations, including how much exploring there was (one said I represented that it would be about 50% of play, and it was still high, but more like 40% in practice), how cohesive the world was (almost spot-on as-advertised), how fair and challenge-oriented things seemed (basically as advertised, especially in combat and thanks to my state-the-consequences style of delivering Moves), and that Dungeon World was largely a good choice because its fiction-first play meant it was simple (especially good for onboarding new players) and got out of the way of the play of actually making smart decisions and experiencing the world. Several more matched-expectations include remarkably flexible scheduling that was nonetheless not as flexible as perceived, and difficulty that was perceived to be very high in the moment of play but that in retrospect was actually not as deadly as they thought it would be at the time or in advance. So you can see how players whose expectations weren’t met are something of a conundrum. We all generally agreed I did all I could to make the expectations clear and then basically met them, so I’m not sure how that failed in the cases it did.

Speaking of on-boarding players, we had a long discussion about that. Ideas to improve this included assigning new players a kind of “mentor” player for a couple sessions until they had learned enough to become self-directed, using various scheduling apps to help make scheduling sessions easier, and using chat programs that had a couple more channels than a single Skype group chat to help keep things organized. So lots about lowering barrier to entry and improving information transfer. Despite problems of information transfer, all of the players I interviewed remarked that there were times in the game where they got to use their familiarity with an area to “lead the team,” and that these were really enjoyable. Such times lived up to that fantasy of playing in a world where you got to experience both sides of there being only one legendary explorer who knows the road to El Dorado, and sometimes it was another player and other times it was you, but it was never some NPC making you feel second-fiddle.

IId. Problems Common to West Marches More Generally

West Marches features a particularly interesting system of first-person overland navigation, like what many people are familiar from in D&D but applied to exterior environments like being in the woods. This is really hit or miss. Some players instantly get it and it’s great and no problem at all, others find it interminably confusing and it drags.

The player of the infamous Tork Jefferson was the former sort of player: he’d rattle off a list of directions based on the player map like “we head up the road keeping the river on our left until we hit Wenton, then we walk through Wenton and turn right about 90 degrees and walk down the valley until we see a lake with a village behind it, we keep that on our right and when we pass that we veer to the right about 30 degrees to head between the two hills to our front and when we hit the river we look for a bridge to cross it. When we see the burned village to our front after that we turn left and head up the valley until we hit the rocky arch, then scramble over that to cross the river on our right.”

This took about 30 seconds to say but sent them reliably half a day’s march from town. Other players would spend 30 minutes to go that far. (Remarkably, these same players often sped up considerably when the session had 5 minutes left and they had to get back to town.) There’s also commonly an issue with the navigation play where usually it’s one player who’s doing it and everybody else is just kind of waiting and watching (this is also a highly general problem in D&D exploratory play in dungeons and things like that). To give them something to do I’d often do “flavor” fills like “what are you talking about while you’re walking?”, but these were also hit or miss due to sessions typically having a hard deadline or 4 or 4.5 hours for some players and they felt a bit like wasted time especially as objectives got farther from home.

This relates to two more issues that commonly come up: location placement and episodic play re: returning to town.

Naturally, as you get farther from town (and difficulty gets higher) you’ll want to place things farther apart so players can get lost more easily. This leads to a density near town and a sparseness far from town that often leads to an “early game” of frenetic activity in the “newbie area,” and then a gradual deadening of the game as it shifts to a phase of far exploration and location restocking. If you don’t do this right it can lead to two ill-effects: slumming in the newbie zone as it restocks and having locations that feel disconnected due to their distance from other things. The latter can be fixed through tight narrative ties and designing locations in batches which our location system often (though not always) helped with. But the former I haven’t found a good solution for yet aside from steepening the reward/difficulty curve.

Speaking of the reward/difficulty curve, this is another problem so I’ll have to address that before getting to the promised returning to town problem. In general, if you make the characters’ power curves, difficulty curves, and reward curves steep, this is good because it rewards good and consistent play. On the other hand, it can also bifurcate your player base and lead to situations where new players have to accompany veteran players to areas they are totally unsuited to and vice-versa. I went with a flatter curve in this iteration and it worked OK. My recent thoughts are that a steeper curve is also OK, but that the rewards should be extra steep so newbies who go to the death zone actually get something worth it out of that. Then you should just be extra-careful about curating your player-base, and perhaps include things that let people restart characters at half their prior level or some such thing.

Anyway, returning to town was a bit of a point of disagreement between the players and I. I always felt it was a bit of a mess, though some of them disagreed. My perspective is that players tended to push the end of the session to play more (a good thing), but that this often left them with an uncomfortable amount of time to return to town (a bad thing, especially if unforeseen circumstances like a random encounter arose). My solution to this was to simply abstract the return to town as a Defy Danger roll (after all, it was dangerous!) if time compression was ever needed due to player bungling. But they would always beg and cajole me to give them 5 more minutes instead. Some players thought this wasn’t a big deal, but I can remember many exasperated session finales where the falling action took longer than planned, and there is a fundamental tension here: do you go back early and risk missing out on extra play time, or do you go back late and risk being obnoxious? One player was adamant that solving this problem should involve a system of “fast travel” to places they were familiar with, but I’m not entirely convinced. We did have a few hooks for such systems in the game, but none of them ended up panning out (like the Wizard’s shadow walk spell or whatever that was called). Worth thinking about, I guess.

Speaking of being fixated, that’s probably the last major common West Marches problem. Sometimes players will become obsessed with certain things in the game world, which can be good, but can also be bad — such as in the case where they’re obsessed with figuring out the ultimate secret to a piece of window dressing or in exploring something they’re clearly not ready for. In more character-driven games that aren’t as prep-forward you can always just roll with this, but in a West Marches game there’s not much you can do except tap their head with the cluebat and hope they get it or you’re in for some boring sessions.

Finally, for some reason players who apply to your game as a group of friends who already know each other are nothing but trouble. Perhaps it’s a small sample size, but this has consistently been my experience. Especially in an open-table game they tend to form a little clique of their own, and they tend to like fighting the GM on things in ways that ordinary singletons don’t. I suppose maybe they feel that their own group expectations have extra heft because there’s several of them. Either way, though, would not recommend, no idea why.

III. Conclusion

So, that’s pretty much it. Do I see the campaign as a success? I see it as a few steps in the right direction and a few stumbles. What I tried fixing got more fixed, though not always solved. And the new things I introduced to be better than ever also did good things while introducing new problems of their own. Would I run something like it again? Probably, with all the fixes and caveats mentioned (and probably also a hefty helping of trying some new stuff that’s as likely to crash and burn the same way a lot of the new stuff in this campaign did compared to the last). Would I recommend it to others? I dunno, read Ben’s blog posts and the last 40 pages I wrote and see if it sounds like something you’d enjoy. If so, then I guess I’d recommend it. If not, then probably I wouldn’t.

Ultimately, it’s like that model train reviewer said: you’ll like it if you’re the kind of person who likes this kind of thing. If you read West Marches or my posts here and go “that sounds awesome!” you’ll probably think it’s awesome, no matter which approach you take to it. If you read it and go “that sounds kind of obscure and I can’t understand why you’d do it that way” then you probably won’t find it to be awesome, but rather a lot of confusing and unnecessary work. Some of my players thought the former, others the latter.

To Be Continued…

Next time: commentary from one of the campaign’s players.

Dungeon World West Marches Q&A Part 4: Freebooters & Follies

This is part four of a Q&A I did on the now-defunct story-games.com forum about the West Marches campaign I ran using Dungeon World. Find part one here. I’ve re-arranged the order of the questions to make it more functional as a blog post and included a few very small edits denoted by [square brackets].

I’m learning about Freebooters of the Frontier which seems to be more of a blorby take on DW?

– User “2097”

[2021 Hydra here: this user uses the term “blorb” to refer to a kind of game with an emphasis on “what would really happen in this situation?” rather than “what would make the best story here?” or “what would make the most interesting challenge here?”]

It’s more like “I want to do an OSR hexcrawl with lots of random generation, but in a Dungeon World way.” It’s very cool and I highly recommend it, but not really for the purpose of a prep-forward game.

Also, are you familiar with other attempts to do this kind of thing? I’m thinking of games like Freebooters on the Frontier (a Dungeon World hack designed specifically for this style of play).

– User “Paul_T”

I don’t think Freebooters actually is designed for this style of play. I have made a couple short comments on this so far, but like you I had a thought process something like “Freebooters? That’s that OSR hexcrawl wilderness exploration thing for Dungeon World, right? I’m doing a OSR wilderness exploration sort of thing, maybe it’ll be useful.” It wasn’t useful, it turns out, because it was really designed to accentuate the Dungeon World where players are making stuff up all the time and you’re improvising everything and so on and so forth as it is marketed. Just about the only things we ended up using from Perilous Wilds and Freebooters were 100 coin weighing 1 and rolling over somebody’s body to grant them Last Breath.

For example, Freebooters’ suggested process for making the map of the region is to start with a blank map, ask someone to draw the border region and home steading, ask a player to start defining the steading, then another player to continue defining it. The players name the steading. They name the place in town they call their headquarters. The players define the predominant terrain of the wilderness, a lost civilization, a region, a landmark, a creature. Players players players. Players don’t make any of these things in Robbins’ deep history model, it’s 100% backwards.

The new Moves are by and large equally pointless. Perceive is just Discern Realities without a question list, Establish is just Spout Lore by another name. Negotiate is no different from Parley. We didn’t need to add a new stat (Luck) because it didn’t do anything for our setting: in a grim dark fantasy frontier you don’t get lucky, if anything you’re in a perpetual state of un-luck instead. And you don’t have hirelings because the PCs are the only ones crazy enough to go out into that horrible mess, so Rise to the Occasion doesn’t serve any purpose and neither do any of the hireling Moves. We didn’t need to replace Moves and Playbooks that already worked just fine — we were trying to maximize approachability for player count reasons, so each unnecessary change was actively bad for us and we tried very hard to avoid them.

The system of Discoveries and Dangers in the base supplement Perilous Wilds is cool, but it’s similarly all improvisation and generated randomly. You build a stock of locations (or perhaps roll them on the fly) and pull them out whenever the dice say the players found something.

But there’s a reason there was a trail-sized gouge in the ground from just north of Wenton all the way up to Crescent Lake, and it’s not because Gareth rolled an 8 and I thought “you know what’d be cool? If he found this huge gouge in the ground, I’ll figure out what it means later.” And it couldn’t have been in any other place than it was, either, because the thing that made that gouge in the ground only would have done so if certain events took place at Crescent Lake first.

There are some things you can improvise, and some things you can weight on a random table, and it won’t matter when or where in particular they happen so much and that’s fine. If somebody rolls a 6- on Spout Lore and you want it to start raining, well, that’s not Deep History ™ but what’s the harm? Sometimes it just starts raining, you know? But you can’t just die-drop entire locations onto your map at random and expect the same kind of result you could get by placing them deliberately.

Do you think the game style works better in-person rather than online?

– User “Deliverator”

Well, this is speculation on my part because I’ve only run West Marches games online (in fact, I do the bulk of my RPG playing online in general, so I’m not the best person to ask about the advantages of in-person games). I’d say each has some advantages and disadvantages.

One advantage for online play is that your pool of potential players is larger. Disadvantage? They tend to flake easier. Plus, scheduling across time zones is never easy. Just because you could play together in theory doesn’t mean your schedule will ever line up in practice.

Another advantage is that if you’re sitting at a computer you can refer to things like your vector map file in real time without as much of a problem as referring back and forth to a laptop in a meatspace game. Disadvantage? You lose a lot of social connection and body language and stuff. Miscommunication becomes more likely, and your game has to stand on its own quality a lot more because social ties are practically non-existent for your first few sessions together.

Something that could be advantageous or disadvantageous is also the kind of between-game stuff you have. Even Ben’s in-person game had a mailing list, so you’ve got to have some kind of player communication technology (accessible to everyone publicly, lest the social monster attack!), but stuff like the table map has ups and downs with being exclusively online. Lots of people will doodle or write in a notebook in a way they won’t break out MS Paint or edit Google Docs. My attempts at replicating physical artifacts of this sort online with the Roll20 map page and Almanac pages were kind of hit or miss — they often involved a lot of extra effort just to “be like” writing or drawing in a composition notebook, and as I mentioned lots of players didn’t find that trade of utility for immersion to be worthwhile.

This seems like something a player perspective would be particularly helpful with, so hopefully Viv or that other player of mine who’s thinking about stopping by could also answer this.

What were the problems you had that you would like to fix in a second iteration?

– Paraphrase of user “Paul_T”

One problem was definitely something to do with player expectations. I’m not entirely sure how to solve this, but it was a real struggle gaining and retaining enough players to have a real big tent feeling, which to me is one of the leading features of an open table game like West Marches. Typically you were looking at around 1 person in 8 making it from application to first session, and a bunch of those people left too.

One part of the problem is obviously just the fact that we were running an online game on Roll20. Random online players have never been the most reliable people.

Many of them probably saw “great for someone with an inconsistent schedule” and thought “low commitment game!” But West Marches is actually a higher commitment style than meeting once a week because in addition to playing in the game the players have a host of other responsibilities: scheduling sessions, talking with other players to tell them how their session went, saying what their character does between sessions, and so on. The GM isn’t leading you on a fun-house tour whether you want to go or not, you need to make the game happen.

Another possible part of the problem is we used a system “against type,” as it were. I’m not sure how we could have been clearer about this — online players, again, not being well known for reading a 250 word explanation of what this game is like and why it’s not like most Dungeon World games on the application page. We required a small application to weed out this sort of thing (name, skype contact info, what interested you about the game, nothing crazy), but I’d still get applications by people named JewStomper (yes, really) that consisted entirely of: “imma need in”. No, JewStomper. I’mma need you out, actually.

Finally, joining the game mid-stream may have been too overwhelming. By the time there were 40 unsorted almanac pages and a janky paint-and-tile map covering a few hundred square kilometers that was a lot to take in for a new player. And without enough knowledge about the state of the world they often couldn’t get time to shine on their first expedition to Everybody-But-Them-Knows-This-Place Gulch.

The core of veterans at the end all pretty much said the same thing: “I knew it wasn’t story game style Dungeon World going into it, and I think it was a pretty good choice of a system given how you handled it. I got exactly what I expected.” As for why all these other people weren’t getting what they expected I have no idea. As mentioned, players who leave online games usually don’t stick around to do exit interviews, so I’m basically reduced to guessing. So it seems like something ought to be done about this, but I honestly don’t know what.

I’m musing about some sort of long-term goal involving the players establishing strongholds farther and farther from the original base, and perhaps even a specific in-fiction goal which, when reached, will trigger the end of the campaign. One of the problems with this model, after all, as with many other campaign frameworks, is that it tends to just peter out, rather than reaching a conclusory “ending.”

– User “Deliverator”

It’s certainly something worth musing about and I have mused about it myself! I originally intended for the campaign I’m describing here to have a specific end condition, but my co-GM explicitly preferred to just play until it fell apart and so that’s the direction we went because that wasn’t a deal-breaker for me. Certainly the original West Marches just went until it petered out, and so did my earlier imitation campaign, so it has a lot of precedent. But I’m not sure if it’s for the best. Maybe it’s something I’ll test next time.

I can’t recall much of what I thought the ending conditions ought to be, but I imagine it tied into the Fame system I had developed. In theory a higher Fame score meant you were winning more, so perhaps it could have been hitting a certain threshold of that.

Something to also think about is that “clearing hexes” and building fortresses works better in an AD&D hexcrawl than in the West Marches paradigm because of certain assumptions about hirelings and who’s willing to go into the woods and stand guard at a construction site against monsters (spoiler: nobody except you!). Although I’ve had players interested in expanding civilization and establishing stockades and loot stashes out in the wilderness, I haven’t yet found an interesting and assumption-consistent way to do it (and so the efforts put into it have been a bit ad-hoc and impermanent). In this campaign the reason the town was safe was due to mysterious runes etched on the city walls, so taking the wall down to expand into the monster-infested farmlands wouldn’t have made a lot of sense.

Reconciling all of these things in a way that’s satisfying and makes sense I guess is a direction where more research is needed!

To Be Continued

Next time: a complete retrospective. Do I think the game worked overall? What would I do differently? Would I do it again?

Dungeon World West Marches Q&A Part 3: Players, Priorities, & Fairness

This is the third installment of a re-host for a Q&A about my Dungeon World West Marches campaign I had on the story-games.com forum a bit over a year ago. You can find the first part here.

This part is a collection of the correspondence about what makes a Dungeon World game fair as well as the player side of the campaign more generally.

Some of the text is lightly edited to clarify things since I’ve moved these questions and responses out of their original order from the forum thread.

I’ve always wondered what the purpose of the “messy” tag is. Does it make a monster more dangerous somehow?

– User “Paul_T”

Well, the description in the DW book is:

It deals damage in a particularly destructive way, ripping people and things apart.

Dungeon World Core Book p. 397

I take that to mean basically anything it hits starts losing body parts immediately. I’d say that’s pretty dangerous! DW’s dragon can either deal 12 damage or bite your head clean off your body when the GM Deals Damage — you can lose through fiction a lot faster than you can lose through hit points, which is why we were so particular about this sort of thing.

People who got a 7-9 on Last Breath was one of our intended windows for inflicting Messy damage without necessarily needing the Messy tag. One PC lost their leg below the knee that way.

Would you sometimes kill PCs before they ran out of hit points?

– User “2097”

Well, I wouldn’t kill anybody — that’s the rules’ job 😉

But yes, sometimes fiction death preceded hit point death. A particularly good example is the guy who lassoed a giant bird. It did not particularly like being lassoed, so it took off and did a tight bank — he lost his grip (Defy Danger 6-) and fell hundreds of feet to his mangled, instant death.

Overall, though, the game was surprisingly low on PCs actually dying or being subjected to Last Breath given the tone and difficulty. Three early on, then a few in dribs and drabs over the following months. That said, probably about a score came within a roll of one or the other, so it wasn’t exactly a cakewalk.

[Many questions about serious bodily harm and healing]

– Multiple users

We had developed a few in-world solutions for this sort of injury, though the players never ended up pursuing them so I won’t say more in the interest of secrecy.

The one major injury of this sort that I recall was a badge of honor for Handel the Skirmisher, who was the previously mentioned character missing a leg below the knee. He was one of the ones who got maimed by Last Breath, in this case as the result of – what else – falling from a great height.

He replaced his missing leg with a hook leg (long story) his compatriot had stolen from a rascally pirate (an even longer story), and spent several in-game months basically doing physical therapy in the town rather than the normal town moves to eventually recover his lost mobility. Maintaining it was no problem since he took the Advanced Move in his class that basically made him a handyman. Aside from the occasional stumble in janky terrain he made a full recovery, and I like to think he resembled one of those people with the modern high tech curved runner’s prosthetics, but, like, medieval style.

And there was somebody else who complained the GM used the “deal damage” move too often in combat, rather than offering them interesting choices.

– User “Vivificient”, who played in the campaign

The funny thing about this incident is that nobody even hit 0 HP. Since nothing even happens to you in Dungeon World from losing HP per se until you hit 0 HP I had a hard time understanding how this could be using Deal Damage too much. For all the color and flavor in that battle I may as well have not even hurt anybody at all.

Perhaps it was just because the fight involved too much fighting and not enough… whatever else you do in a fight. I don’t recall that they specifically asked for more interesting choices, though, just that I was doing Deal Damage too much. So we’re left to guess at their motivation.

Either way, that player made a principled stand to never play in a session I ran again. I couldn’t understand it.

One of the things I love about DW (and, really, pretty much all PbtA games) is how there’s no separate “combat” phase of the game, really. There’s no, “now roll initiative.” You can freely use the combat moves along with trying to parley, or interact with the environment, or escape, or whatever. I think, perhaps, that one of the emergent features of Jon and his co-GM’s modifications was less of that feeling: combat became more rigorously adjudicated, and thus more of “its own thing” in the game, separate from “normal” play.

– User “Deliverator”

I can’t say I lean towards this as the definite explanation, though since I don’t have a great explanation myself I have to concede that you may be totally right for all I know. For the record I also like the system for this reason, just like you do.

Did we use Deal Damage enough so the players would notice? Yeah. But, I mean, if the monsters aren’t damaging you then what are hit points even for? As for other decisions based on standardized procedures, if we imagine 5 alternate universes in which I’m faced with the same situation I’d imagine I make the procedure-based decision in 2 or 3 of them anyway, so I don’t think things were particularly rigid or unnatural. We had plenty of talking turn into fights and plenty of fights turn into talking, sometimes by the baddies’ initiative and other times by the players’ (especially considering the game setting was a hostile wilderness full of hungry abominations). As the cherry on top I practice Renaissance martial arts so I pay special attention to fictional positioning in combat compared to most GMs. As far as I could tell we weren’t hung up on rigid procedures and numbers.

Speaking of fictional positioning and consistency in and out of combat, I remembered another guideline I developed partway through the campaign that I believe is actually fundamental to the style and in fact is applied to most adjudications in a given session.

DW is very forward about rolling in the open and the GM is encouraged to not even roll the dice at all. That’s cool, and it certainly prevents traditional fudging of results by fudging the numbers on the dice. But changing the numbers on your dice isn’t the only way to fudge the result of a dice test. In traditional games the numbers on the dice mean certain things, so you use the mechanism of the face-up numbers to change it to the result you want. In DW, though, you don’t have to fudge the numbers to get a desired result because you can fudge the thing the numbers mean by executing a softer or different Move after you see the 6- hit the table than you would have done in a vacuum.

While technically rules legal, and not always committed consciously, I call this a kind of fudging and I think it’s bad for two reasons — I believe in general, but certainly at least under the traditional challenge/sim-oriented playstyle.

One reason is that I think in practice you’ll find the Move you would pre-commit to is much harder than the Move you would actually make after the 6- hits the table. Your only concern when pre-committing is what would reasonably happen, and preferably be interesting, as the result of failure. You don’t have to concern yourself with the fact that with the 6- already on the table your Move choice is all that stands between a PC and death and that you will be personally responsible for killing them if you arbitrarily choose to do so.

The other reason is that pre-committing gives the players more information about their action, and having more information leads to more informed choices, and more informed choices leads to better play because something something a game is a series of interesting choices (thanks Sid Meier).

So the delivery I developed is that whenever it would make any sense at all I would tell the player what would likely happen if their attempt failed (so sometimes not for snap reactions or if the character would reasonably not know what would happen if they failed like on Discern Realities in an empty room or whatever).

In addition to circumventing the two problems mentioned above (inconsistent/wimpy Moves and lack of information), it also has two further benefits.

One of those benefits is foregrounding fictional positioning. Often players will make a prudent plan in Dungeon World but not feel like it had any effect on the outcome because it wasn’t good enough to skip the roll entirely and they rolled a 6- and bad stuff happened anyway. With this method if they were trying to climb a cliff with no equipment I could say something like “OK, but if you fail in you’re probably going to plummet to your death on the jagged rocks jutting out of the ocean down below.” And then the player would go “WAIT, I HAVE ROPE!” and they’d anchor themselves with rope and I’d go “OK, now if you fail you won’t plummet to your death, but when you’re caught by the rope that’ll jerk you around and might bang you into the cliff or knock something loose out of your pack or something like that.” This makes the advantage gained from smart planning in the fiction very apparent in a form like “without this plan, X would have happened if you failed; but now with this plan Y will happen if you fail instead, which is much better.” This idea can also apply to successes if needed.

The other benefit is that the players become immediately involved in making a fair, impartial game as your helpers. Where Ben saw lots of potential in 3e’s fairness-as-bureaucracy (everyone gets the same rules, no exceptions), this method offers fairness-as-democracy (everyone gets a say about whether something is fair before it happens so that if the GM is about to make a genuine error they can catch themselves better). Obviously the final say must rest in the GM as the referee, but fairness-by-group-consensus is one powerful way to take the arbitrary sting out of fundamentally fiat-based systems like selecting a GM Move. And with legit players calling you on wimpy Moves, in a weird way you get the same kind of fair result by consensus rather than by writ.

So that was a long digression but long story short I don’t think I ran combat really differently than I normally would.

One thing about this project though… I remember one of my friends asking me about DW and I said “lol you really want us to switch to a game where there’s around 40% chance of fumbling on every roll?” 6- rolls happen often

– User “2097”

One thing to keep in mind is 6- is “make a GM Move,” not necessarily “the players fail.” We erred on the side of being jerks on a 6- for consistency with our setting and tone, but you don’t necessarily have to do that all the time. That quote from the infamous Dungeon World combat primer you had in the other thread is all about how fudging the difficulty of a 6- is technically rules legal, and to be fair it is. You could also just hand out extra +1s to stats or whatever more frequently, but from experience Dungeon World characters are tough and can handle a lot of failure unless you decapitate them every time they take damage.

It can also help just how you frame failure in the narrative. A trick I picked up from Stalker: The SciFi Roleplaying Game is that unless the plan was really bad and bound to fail (in which case they probably shouldn’t even get a roll, since rolling implicitly requires the possibility of success) you should narrate player failure as not really being their fault – the situation imposed on them. Anybody can effectively 10+ Hack & Slash on a target dummy (because you don’t even have to roll), but against a live target your 6- isn’t just you sucking: the baddie’s trying to stay alive, too! If they Defy Danger 6- to jump a chasm they don’t trip, their impressive leap was just not far enough. If they Parley 6- they don’t stutter and mumble, the target just doesn’t react to the leverage in the way you might have expected – who could have guessed?

Incidentally, “a missed roll is an embarrassing fumble” probably has its memetic origin in a poor reading of D&D’s combat system if I had to guess. A poor reading because actually a “miss” in D&D combat doesn’t mean a failure to connect, it means that there weren’t any effective hits. Maybe the opponent dodged your blow that was otherwise well-aimed, maybe their armor shrugged it off, maybe they parried (not that D&D gives an AC bonus for having a weapon, though it probably should given how hard it is to block a sword with your bare hands). And originally a to-hit roll represented something like a minute of scuffling, so to say that you swung wildly at the air and missed like a buffoon for a full minute is kind of ridiculous if you think about it critically, nearly as much so as treating HP as “meat points” where you take full axe blows right to the nose to no effect 3 or 4 times before suddenly keeling over dead. But that’s not really on topic I guess, just a personal theory.

One reaction I’ve been having reading this that I hadn’t been expecting going in was that I’m not sure the math works out? […] I usually love [math like 2d6+x vs. 7] (it’s so close to 4dF vs 0) but… seems like it was kinda lethal?

– User “2097”

Regarding lethality it’s been mentioned a few times now that PCs rarely died. I think maybe 5 or so over the course of about 60 sessions? I can only think of four specifically off the top of my head, so call it probably no more than 6. And three of those were from a single early session, so the rest of the campaign was actually quite the opposite of lethal.

On the lethality issue: in DW, it’s really not an issue of “how often do dice rolls fail”. After all, D&D5e is set up so that most rolls end up roughly in the 50% ballpark (maybe 60%/40%, or something like that?), as well, it seems to me.

– User “Paul_T”

I don’t want to keep harping on about this, but I do want to emphasize here just how much the math actually favors Fighter-types in straight fights that are about trading hit points back and forth in Dungeon World. With a +2 to hit, 2 armor, and d10+2 damage (pretty common for a level 1 Fighter) you’re going to be looking at an average damage result for the fighter of around 7 damage each time they throw a Hack & Slash against a return of like 2 damage from a monster throwing 1d8 back. You could expect a Fighter to kill ten monsters like that in a row before even noticing their HP was getting low.

But if you crank this monster up to 20 HP and d10+4 damage to make it an even shot against the Fighter, this absolutely creams anybody else. A Ranger with an unusual +2 strength bonus trying to go at such a monster in a fair fight would die twice in a row and the monster would still be alive. And this problem only gets worse as levels go up and the Fighter accumulates more armor, damage, and a +3 Strength.

This is why I make such a big deal out of ‘standardizing the fiction’ – it just wasn’t even remotely an option to balance the game around the numbers alone, because if we had then our only options were Fighters being unkillable gods of death or everyone who wasn’t a Fighter dying in one hit. We chose instead to challenge people with dudes on the roof throwing rocks down at them so they couldn’t fight back, or backstabbing them because they were on home terrain and knew the shortcuts — not duels in empty 10×10 foot rooms against guys with big damage dice. The difference between somebody fighting against 3 of those d8 damage goons to their front in a hallway (d8+2 on a 9- Hack & Slash per the group combat rules of DW vanilla) and 3 guys to their front and 1 guy bonking them on the head behind them (2d8+2 on a 9- H&S and still 1d8 on a 10+ from the free uncontested hit from behind) is pretty enormous.

In this way the challenge was almost exclusively in coming up with a good plan, so as long as the players weren’t truly bone-headed the game wasn’t actually that difficult (but the pressure to come up with such a plan made it feel pressing in the moment).

[…] sometimes in [my own game system] I don’t have them roll at all, it’s just “ok you can do that”; does that happen in [your way of playing Dungeon World here]?

– User “2097”

My impression so far is that the answer is a strong “yes”, and that positioning yourself so you didn’t need to roll was a major part of the game. (Because rolling could always turn rather sour for you.)

Is that close?

– User “Paul_T”

I’d say Paul’s got it. The game was designed and run to encourage players to do things that would ‘just work’ without ‘accidentally’ triggering a Move as a theoretically optimal way to play.

An example that probably took this a little too far was their half-serious-half-joking refrain of “don’t ask questions.” Players often sought additional clarification about what their characters saw from me in order to help make decisions, but not to the degree that they were consulting their character’s intuition and rolling Discern Realities or their character’s memory and rolling Spout Lore. They were OK with sometimes not having important information because they believed missing some information was often preferable to risking a 6- result.

Personally I think they took this too far sometimes. Knowing what I did, it was frequently the case that the solution to a puzzling situation in front of them was information they had previously learned, or perhaps even had access to in their almanac, carousing stories, and player reports, but for whatever reason had forgotten or neglected. If they had risked a failed roll they also would have ventured to crack open the whole scenario. But they chose to play it safe, and that was designed to be a valid option.

There are plenty of arguably more successful examples, though. Climbing things slowly and carefully using teamwork, rope, and harnesses was very common. Navigation was done entirely in this way: they said “I walk towards the rock” and so they did, no Move required. Many players successfully took advantage of the fact that if you attack a helpless opponent you skip straight to dealing damage by creating a situation in which their opponent was rendered helpless.

I think that a huge part of this kind of play is a disciplined and consistent GM and the trust of the players in them. If the players have sufficient trust, the mechanics don’t HAVE to be very impartial at all. (I think of the Spicy Dice Game, as an exemplar – those players would happily play a “West Marches” style campaign with that GM and those rules, even though they provide basically zero support for consistent adjudication.)

– User “Paul_T”

Yeah, that sounds like the kind of thing I’m talking about here. Not that I’d ever bet on finding 12 players interested in Spicy Dice West Marches, but if they were into it and found it believable and fair then that’s 90% of the way there.

It’s not like 3rd edition is even that hands-off if you think about it. If you get into a fight the DM could just have all the monsters run around and do double movement every turn instead of fighting back. That’s 100% rules legal. But it’s also unbelievable, so nobody would ever do that.

The key thing here is that you need a system that gets you to the point where the decisions you have left to make won’t end up being unbelievable ones. And if 3rd edition D&D does that, great. If Spicy Dice does that, also great. I found that – if I was careful about it – Dungeon World did that, so that was fine by me.

What does it mean to “have enough players to spin up a strong meta game”, and what kinds of numbers did you find were needed?

Does it depend on the engagement level of the particular players, as well? (So, maybe, one really highly engaged player is worth four regular players in this hypothetical algebra 🙂 )

– User “Paul_T”

Basically to get the open table effect you need enough players so that they start mixing together. Once they start mixing together they’ll have a need to share information with everyone so their teammates are all up to speed, and that kickstarts the session summaries and map drawing (if you’re lucky). My favorite part of West Marches is actually the dynamic between the sessions where players talk about the game. It’s rare to see players as interested in the game as the GM is (even if only as a means to an end), so it’s a real treat. It’s a bit of a self-reinforcing loop with the prep: prepping in advance makes things deeply connected enough to be worth talking about, and them talking about it helps to make you want to prepare even better things to make them talk about them more.

I think you’ve also got it right that players who are more engaged can be worth more than less engaged ones for this purpose, though only to a certain point. If your West Marches is just four super engaged players then for all their engagement they can never really make a meta-game because there’s nothing new to share, they were all there!

In Ben’s game he listed a pool of 10-14 people and we shot slightly higher for about 16. I’ve found previously that running a game like this online you tend to get a bunch of players who want to play around twice a week, so we expected to have a hardcore group of about 8 who wanted to do that, then the remaining 8 to be a mix from once a week to once every other week or so. Much past that and you kind of lose track of what’s going on and don’t feel like part of the crew anymore I’ve found.

In practice our group chat occasionally got up that high, but the usual numbers were more like 5 or 6 hardcore players, then 6 or 7 more in varying states ranging from “once a week” to “actually this person has already quit the game but you won’t come to realize it until 3 weeks from now.” So probably around 10 players actually playing at a given time, plus a few malingerers. It was a little on the low end for my tastes, and it definitely had too much turnover.

What did the players use for a map?

– User “Paul_T”

This, roughly speaking:


It’s a page on our Roll20 game made to look like a piece of paper on a table. Players had access to the usual drawing and text tools, plus terrain hex tokens for the artistically inclined, with an invisible grid offering a map scale with the ruler tool. They had run off the edge of the paper and transitioned to a bigger map but that was near the end of the campaign so this version is more representative.

The players didn’t take advantage of some organizational tools we offered them, like having tokens with A through Z to represent locations each referring to a specific Almanac page. Then they complained that their Almanac and map were disorganized. Oh well.

How much correcting/management did it require from you, as GMs?

– User “Paul_T”

Once we set up the page the only management we did was dealing with wonky Roll20 image permissions. It’s a key feature of West Marches that the GM should never provide the players with meta-game materials like maps or session summaries — the players need to make them for themselves.

I’m musing about some sort of long-term goal involving the players establishing strongholds farther and farther from the original base, and perhaps even a specific in-fiction goal which, when reached, will trigger the end of the campaign. One of the problems with this model, after all, as with many other campaign frameworks, is that it tends to just peter out, rather than reaching a conclusory “ending.”

– User “Deliverator”

It’s certainly something worth musing about and I have mused about it myself! I originally intended for the campaign I’m describing here to have a specific end condition, but my co-GM explicitly preferred to just play until it fell apart and so that’s the direction we went. Certainly the original West Marches just went until it petered out, and so did my earlier imitation campaign, so it has a lot of precedent. Whether it’s for the best is another story, though.

I can’t recall much of what I thought the ending conditions ought to be, but I imagine it tied into the Fame system I had developed. In theory a higher Fame score meant you were winning more, so perhaps it could have been hitting a certain threshold of that.

Something to also think about is that “clearing hexes” and building fortresses works better in an AD&D hexcrawl than in the West Marches paradigm because of certain assumptions about hirelings and who’s willing to go into the woods and stand guard at a construction site against monsters (spoiler: nobody except you!). Although I’ve had players interested in expanding civilization and establishing stockades and loot stashes out in the wilderness, I haven’t yet found an interesting and assumption-consistent way to do it (and so the efforts put into it have been a bit ad-hoc and impermanent). In this campaign the reason the town was safe was due to mysterious runes etched on the city walls, so taking the wall down to expand into the monster-infested farmlands wouldn’t have made a lot of sense.

Reconciling all of these things in a way that’s satisfying and makes sense I guess is a direction where more research is needed!

To Be Continued

Next time: questions all about two commonly recommended “old school” / exploration supplements to Dungeon World (Freebooters on the Frontier and Perilous Wilds) as well as a few problems with Dungeon World West Marches and running games online.

Dungeon World West Marches Q&A Part 2: Monsters, Masters, and Mysteries

This is the second installment of a re-host for a Q&A about my Dungeon World West Marches game I had on the story-games.com forum a bit over a year ago. You can find the first part here.

This part is a collection of the correspondence about how we changed the monsters of Dungeon World, how we ran a campaign with two GMs while still keeping it feeling like one game (usually), and why I won’t spill the beans about some of the content in the game.

Some of the text is lightly edited to clarify things since it’s no longer in its original message board context. Editorial work is indicated by [square brackets].

Is [discussing how you’d adjudicate a common game situation for a while until you agreed] an illustration of the way the two GMs spent some time getting on the same page, or would you actually make a point to do this before running the game?

– User “Paul_T”

It’s a typical description of how all of our actual rules got made. We’d get together on Friday for an hour and think about common situations in the game and wonder about stuff like “hey, when would you use this question in Discern Realities versus this other one? They seem really similar.”

We were both fairly argumentative people, though, so I imagine it wouldn’t have to take as long as it often took us. It’d also be faster if you didn’t need two people to agree and were just doing it by yourself. We didn’t run a game with two GMs in the same session, so this was all prep work.

How much of a range was there between your two “adjudications”, and what helped you get on the same page?

– User “Paul_T”

Sometimes it was as easy as “oh yeah I do that the same way,” other times we had to come up with a whole procedure because we had never really thought about it before. One example that comes to mind is we spent some time talking about different circumstances in which a player would just get hit by something in combat. Like, if you’re fighting two guys and one is on each side of you does that count? Both in front? One ahead and one behind? What about doing something else while there’s an archer overlooking the whole situation 50 yards away? Over the course of an hour or so we basically exhausted all the common scenarios we could think of where somebody might just take damage without getting a Defy Danger or anything, and then we made a big list of the ones where we agreed it would definitely happen.

One example of something we didn’t bother to standardize was how we selected players to spotlight. I tended to just go with whatever made sense as the next most pressing thing based on a combination of speed, importance, and making sure nobody gets left out for more than a few minutes at a time. He preferred to pick an arbitrary order and just cycle through it. But because of how Dungeon World works regarding how consequences rarely happen to you unless you first take action we thought it would be equally fair either way we decided to handle who sits out and when.

Do you remember any examples of [when the you two GMs failed to adhere to the same rules of execution]? I’m curious what “level” of the game they came in on (e.g. prep vs. adventure design vs. MC moves vs. adjudicating particular player-side moves).

– User “Paul_T”

The one that comes to mind was to do with maiming PCs. A player had got their finger caught in a mimic and my compatriot GM had it get torn off, which I thought was in line with our standard. But then he said it was OK to re-attach it with Cure Light Wounds, although it would not regain its functionality. I thought that violated our common understanding of Cure Light Wounds not being able to re-attach lost limbs at all, regardless of whether the limb worked or not after. So we had a bit of a spat over that, and the fate of the finger was in limbo ever since.

It sounds like you either don’t remember or didn’t document [all the standards, monster stats, and running away], but they sound really vital and important to making this work. Any chance you have some of it somewhere (in email records, perhaps)?

– User “Paul_T”

Well, I want to be careful in answering some of these because one of my rules for running West Marches is I never reveal any of the world content to the players other than what they find, and I have one of my players running around this forum =P

I can answer to a pretty large degree stuff about systems and rules, though, insofar as they don’t give away any secrets about the world.

Regarding fleeing monsters I think we never wrote it down and I don’t remember the specifics of what Moves got made when from actual cases from the game. I could contact some of the players who were involved in the more famous chases but there’s no guarantee they’d respond or remember anything.

I can say some things about the monster system, though, because that was recorded for reference purposes.

We built the system number-wise to balance around two goals: that a monster shouldn’t just gib a wizard with minimum hit points (in our game 12) and that an “appropriate encounter” ought to badly hurt but not kill a “typical party” of Fighter, Ranger, Cleric, and Wizard if they just trade back and forth. The assumption was that players should be finding cleverer ways of fighting than just trading Hack & Slash back and forth with the monsters if they wanted to get by with more than the skin of their teeth.

In practice this happened sometimes, so go us, but more commonly the Barbarian and Fighter players got into it with the monsters and the other players supported them by holding the flanks and dealing with unexpected problems so they could keep at pitched combat with their superior damage, armor, and HP. This is because Barbarians and Fighters actually have a very substantial advantage in a straight fight compared to the other classes on account of how Dungeon World’s numbers work, so monsters calibrated to challenge a “typical party” overall are not nearly as much of a challenge to a Fighter or Barbarian in particular. This led to a funny dichotomy where, depending on their preferred method of dealing with monsters, some players treated Defend as the Second Coming while others treated it as patently useless. Some enemies totally dismantled this style, though, so it wasn’t happening every time.

Anyway, we basically made our balance rules around running sims of a few “boring combats” like that. In the end we decided a peer monster (a monster meant to square off against one PC) should start around 8 HP and with d8 damage, and to adjust from there depending on the intended level and number appearing. Unlike Sage I think “stuff gets less HP if there’s more of it in a group” is more about game balance than realism, but I like having good game balance even at the expense of realism in edge cases (armies of human soldiers, prides of lions, etc.) so that’s fine by me. In retrospect I think slightly lower HP and slightly more damage may have been more prudent since fights took a little longer than I wanted (quick fights being another reason to pick Dungeon World over, for example, 3rd edition D&D), and fulfilling the goal of not gibbing the Wizard while challenging the Fighter on raw numbers proved to be tenuous, and I’d rather challenge the Fighter than not gib the Wizard.

On top of that monsters got various Strengths based on the intended level of their PC opponents. A higher level monster had slightly bigger stats, but also things like the Messy tag, some narrative advantage that tends to require Defy Dangers to be made, a specifically advantageous environment, or handing out Debilities. All monsters also had one or more Weaknesses, which were fictional positioning elements that gave a dramatic advantage over them in a conflict if utilized.

Monsters also had the usual DW Instincts, which we treated as their overarching goal in life, and Tells & Behaviors, which covered telegraphing that the monster was around and what it was trying to do on a moment to moment basis. The monster’s Moves were the particular things it would do to fulfill its Behaviors, which in turn would lead to fulfilling the Instincts. They also all had Origins, which were a major source of XP if discovered (they often weren’t).

To give you an example of all this, one of the first monsters we designed was a kind of ten foot tall bog monster covered in rusty chains and padlocks that lived in a lake far from town (we referred to him affectionately as Rusty). He wasn’t the highest level monster that existed, but he was near the middle of the level 1-10 spectrum. We chose his extra levels to grant a bit of extra Damage (we thought this made sense as he was physically quite big) and a fair whack of Armor (we thought this made sense because of the giant metal chains). Being a solitary monster grants some additional survivability, so we opted for a bit of extra HP and also a Defy Danger type defense (the chains acted as Reach+ weapons you had to get past to close the distance).

Some of his Strengths the players experienced but never noted were that he was very intimidating (basically an aura of fear that caused DDs to approach), he was extremely strong (Forceful tag), and he could wrap people up in the chains and loved dragging them into the lake with him since he didn’t have to breathe air (so that’s a strength of being adapted to its environment as mentioned earlier). I recall a few PCs almost drowned this way.

I don’t want to talk about one of the Moves or the Tells & Behaviors because those contain a few things I don’t think anybody ever experienced, but the other two Moves make obvious sense: Enwrap in Chains (from its strength of the chains) and Do a Massive Punch (from being Forceful).

Finally, they did note the monster’s main Weakness, which they had exploited: you could unlock the chains if you dared to use the keyed padlocks, rendering him much less potent (being without weapons or armor).

[…] I am a bit confused by your attempts to avoid “spoilers” about the campaign. You speak about it as though it’s in the past, but you’re also careful not to give things away. Why is that? Are you planning to run it again with the same players, or something like that?

– User “Paul_T”

This is for two reasons. One reason is that Ben is notoriously resistant to giving away specific information from his campaign, so I figure it’s traditional for West Marches games to do that and so I had better do it, too.

A less obtuse reason is that I believe missing information has value in many cases. Handing it out can change the value of the original work in much the same way a sequel answering certain questions one way rather than another can change the value of the original work by modifying its context in retrospect. I’m not interested in modifying the context of my campaign in retrospect in terms of the contents of the world, so I refrain from answering those kinds of questions.

After all, if I can’t keep a secret to myself in such a low pressure environment how will I maintain one when it actually matters during the game? And keeping secrets to oneself is vitally important to running a game in the exploratory prepared style. At the end of the day, West Marches is about giving player actions real weight. The players are the only ones willing and able to explore the dangerous world beyond the wall, in the game or out of it, and to give away the answers they worked so hard to get for free at a later date is to cheat them of their labor.

So basically I think it would make my game be worse and less immersive in retrospect, and at this point in time retrospect is the only way to experience it at all and I am not interested in making my game be worse and less immersive. I am aware how you would get a lot of value out of knowing these specifics, but it is a non-starter for me.

Monsters that are devious and hard to pin down isn’t always the case in D&D style games!

Was this a conscious design choice, and what efforts did you make to maintain it? How did you agree to do the same way or reliably?

– User “Paul_T”

Well, it’s not like we sat down one day and went “you know what, we should have all our monsters be tricky bastards.” But a bunch of things came together and that ended up being the way we tended to go with things.

One reason is, as I keep mentioning, a monster that pastes a Ranger at Hack & Slash in turn gets pasted by a Fighter, and that only gets worse as the levels increase. The numbers just happen to work out that way. So you simply can’t balance Dungeon World combat by numbers alone, even if you wanted to. We weren’t playing GURPS here so we weren’t trying to balance it completely by numbers, but even our best result in that respect was still only close enough for horseshoes.

The other GM and I were both very aware that Dungeon World’s combat difficulty basically revolved around how much of a dick the GM is with their Moves, and that we would need to do something in this area given the realities of the numbers discussed previously (besides, balancing by pure abstract numbers is boring and misses the main selling point of the system, which is that it’s fiction first).

Robbins’ WM was very much in the vein of “prep according to fair guidelines, then play as hard as you can within the rules and prep knowing that you’ve been constrained fairly by your past self already.” The question became: how hard could we play “within the rules” and what would our constraints from prep be?

To know the limits of “as hard as we could within the rules” we invented all of the rules of thumb for adjudicating combat situations. Monsters in this world were often intelligent, and intelligent things don’t want to die. They’ll use whatever advantage they can scrape together, like ambushing with projectiles. What’s a fair way to damage people with projectiles from ambush? We had to consider that kind of thing. Once we had the rules in place for what would and wouldn’t happen in a certain situation, we could move on to playing the monster as hard as we could within our prep certain that the outcome would be fairly well understood in advance for those actions.

So then that’s the other question: what would we prepare in advance in terms of monsters being cunning? This is where our system of building monsters, and most especially Instinct, Tells & Behaviors, Strengths, Weaknesses, and Moves came in. All monsters tried to fulfill their ultimate Instinct, and they did that by making their Moves, which were telegraphed and portrayed by their Tells & Behaviors. Rusty liked to wrap people in chains and drag them into Crescent Lake. There’s a reason for that in Rusty’s instinct, strengths, moves, and behaviors. I won’t discuss what it is because I don’t think anybody ever figured it out, but we both knew from each and every monster entry what a monster was trying to do and how it would try to do it no matter which person was running it or which person had created it. There were times when one or the other of us would look at a monster entry or location and go “uh, what? I don’t know what you mean here, what would happen if the players did XYZ?” and in the few days before a session the other GM would clarify things. Usually that wouldn’t take more than a couple minutes.

We would always do our best to explicitly take advantage of a monster’s listed Strengths whenever possible. Rusty was strong (and, mechanically, Forceful), so naturally he liked to get into tug of war because he would always win that game (except that one time he didn’t, but he’s intelligent not clairvoyant). He had Reach weapons and an aura of fear, so naturally he played zone defense to keep those pesky barbarians with the big sword away from him. And if someone did somehow get too close and present a serious threat he’d boot them away (Forceful, remember?) and they’d have to run the gauntlet all over again. And if that somehow failed, well, he could breathe underwater and the PCs couldn’t, so he’d just retreat into the lake. In this respect a monster’s tactics kind of played themselves if you just took 5 minutes to read the entry, look at the location, and think about it a little. We were both on the same page from the beginning about the general style of challenge-oriented play, so there weren’t any issues in that regard. We both wanted to play monsters as hard as we could within the prep and guidelines and we both did a pretty good job of being creative (while not violating prep or guidelines) with that I think.

We were also always doing our best to at least display a monster’s Weakness(es). A player can’t exploit something they don’t know about or can’t infer. We were obligated to portray monsters as the whole package of goals, information (tells, weaknesses), behaviors (as well as specific moves), and strengths at all times. Lots of locations were designed in tandem with their inhabitants, and we took that interaction into consideration while imagining how the scenario would probably go in advance.

In fact, locations were all designed with an explicit progression for what would happen there if the players didn’t interfere (just like Fronts in the base game) just so the world didn’t feel like a location was waiting in a time capsule for players to show up, and this progression typically was a result of their inhabitants’ Instinct and Behaviors. Just ask them about the infamous tornado bridge. So this further incentivized us both to be very familiar with how the inhabitants of locations would behave if it came to a fight. Every fight had a goal they were trying to accomplish, and they’d use the whole gamut of their monster entry to try to achieve it.

So basically our priorities were already aligned, it was just a question of working behind the scenes to shore up Dungeon World’s willy-nilly nature to something more consistent and, for lack of a better term, ‘legible’ to someone who didn’t create a location or monster in the first place but who had to run it.

Could you share an example of a monster write up and combat goals, as an illustration of what you’re talking about there?

– User “Paul_T”

Well, I’ve basically shared all of Rusty that matters, so here is what combat with him tended to go like (not based on any particular encounter but more of a pastiche of what I remember in general from all their run-ins with him):

Players are exploring near the lake he lives in. He is interested in drowning them, so he stalks them from the water (instinct & behaviors, roughly speaking). I describe ripples in the water (tells). Players don’t attend to the ripples. I describe one character being grappled by rusty chains suddenly shooting from the water (golden opportunity, they ignored the ripples). They describe pulling and resisting, and somebody helps them. Aid Another or automatic fail (Forceful makes it impossible to plausibly beat Rusty at tug of war by yourself which precludes Defy Danger). Aid Another succeeds, they pair up to pull on the chains and haul Rusty to the surface of the water and just barely to the shore (Defy Danger + STR). I describe him as a giant human-shaped monster covered in lake gunk (physical description) and rusty chains (Strength) affixed with keyed padlocks (Weakness).

The other two players make an attempt to attack the creature in melee by running up to it. It whips chains at them, and one thinks better of risking being hit and rummages through their bag for some kind of ranged weapon (player taking action that automatically succeeds due to not triggering a Move). The other player dodges the chains (Defy Danger + DEX). They succeed, but when they get up to it they see its imposing size and physique and suddenly feel cowed (Defy Danger + I forget, whatever resists fear). They succeed and maintain courage enough to attack, rolling Hack & Slash. 7-9, their attack deals some damage to Rusty (mitigated by his armor from the chains, descriptive narration involves striking the chains to highlight Strength), and Rusty boots them like 6 or 7 feet away onto their face (Forceful + HP damage, Monster Move = Do A Massive Punch (also Deals Damage)).

One two skip a few, players get free from the chains with some more Defy Danger or perhaps an automatic success by e.g. coating the grabbed person in oil or some such thing, the players all retreat to a fair distance and threaten Rusty with that ranged weapon the player rummaged their pack for and so he retreats into the lake (a kind of ‘automatic success’ of sorts, since I know how Rusty will react to being threatened by ranged superiority and great distance from the lake and so there’s no need to Parley with the leverage of being turned into a pin cushion due to his instinct and behaviors).

How was your prep recorded and organized? Was it fully shared across the two GMs, and, if so, in what format? Or were there some informal agreements that, say, such-and-such castle is your “territory”, and therefore you’ll be GMing any adventures there?

– User “Paul_T”

To clarify, we did share all of our prepared locations. We had a Google Sheets file with all of the mechanical nitty-gritty for monsters and locations (sometimes appended by a Google Docs file with short summaries for locations, though this was less standardized), and of course a major benefit of having a vector image file for a map is that you can write all the important information for running a location right next to it!

Typically we’d also have a 5 or 10 minute chat before the session anytime one of us was going to run material the other had developed, which was actually pretty much every session. For whatever reason the players in all my games went to places the other GM had made, and the players in all of his games went to places I had made. There were 2 or 3 instances I can think of where what the players got wasn’t what one or the other of us had originally envisioned in our preparations, but by and large it worked well for the first few dozen sessions. After that things started getting a bit raggedy because both of our coursework was picking up and we had less time to talk and the short version of locations (being only a couple paragraphs long at most) wasn’t always as bulletproof to interpretation as we might like.

To Be Continued

That’s the end of part 2! Tune in next time for questions and answers about players, their priorities, and what makes for a fair game of Dungeon World West Marches.

Dungeon World West Marches Q&A Part 1: Philosophy, Playbooks, and Principles

Last year or so a friend of mine who goes by the username “Vivificient” told me that some users on the story-games.com forum were interested in how I used Dungeon World to run a multiple-GM but otherwise roughly textbook West Marches campaign in Ben Robbins’s original style. Since story-games.com is no more I thought I would re-host the Q&A here for anyone who’s interested in how I got a game that’s supposedly all about story and improv to work in a style that’s supposedly all about prep and following the rules.

The following conversation isn’t exactly in its original order. You know how forum posts get all mixed up with people quoting different parts of messages and so on. So for clarity I’ve tried to group it into questions or remarks about one or two topics at once to make it easier to follow. Otherwise the text is basically unedited. Understanding a lot of what is said here will require you to already be familiar with Ben Robbins’s posts on his West Marches campaign concept I linked above.

To start off, had you modified the GM moves and the outcomes of the playbook&basic moves?

– User “2097”

Yes, somewhat. As a general rule we tried to leave unmodified as many things as possible, because part of a good Robbins-style West Marches is in having enough players. We didn’t want to use D&D because it tended to be too cumbersome (I’ve had bad experiences with West Marches games and wasting lots of time explaining to players how to build characters, and how long it can take to prepare adventure locations), so Dungeon World was a next best option with lots of popularity, and we didn’t want to ruin that by homebrewing the entire game from the ground up.

That said, the default picture of Dungeon World (as well as some of its rules and contents) did not mesh well with a prep heavy game. We tried to be excruciatingly clear that this would not play like many improvisation-heavy story-forward Dungeon World campaigns people might be used to, and that a few changes had been made where necessary to support a planning-based play experience.

Can you give some examples of how the new moves looked?

– User “2097”

I actually made a little Google Docs hyperlink adventure booklet for precisely this purpose with new players. That includes all the Dungeon World rules, though, including the ones that weren’t changed, so I’ll discuss some of the ones that were specifically changed here.

As mentioned previously, we tried to avoid changes where we plausibly could. So this typically led to 3 kinds of changes: clarifications (the most common), tweaks to existing Moves (less common), and the creation of new Moves (least common).

Clarifications were often of the sort like how most Moves actually implicitly involve the term “relevant” or “makes sense”. From experience I’ve found lots of people who interpret the Moves without this idea of “makes sense in the fiction so far” in mind have trouble playing the game, so we made these clarifications where they were needed. Here are four representative examples.

As a kind of ur-example, if you gain +1 Forward from acquiring knowledge of a monster’s weak spot on its belly this would reasonably apply to your next Hack & Slash or Volley Move. But probably it wouldn’t apply to Last Breath, because why would it? Many clarifications of this sort were made throughout the Moves — if it doesn’t make coherent sense, obviously you can’t do it.

A similar clarification was made to Hack & Slash. To be in melee means you are in a conflict at short range where you and something else are trying, and have the ability, to hurt each other physically. If you can’t hurt it or it can’t hurt you, that’s just GM Moves deciding that based on what makes sense (generally: it bops you in the face or you bop it in the face, respectively).

Spout Lore was clarified to point out how you are consulting your character’s knowledge, and your character reasonably wouldn’t know a lot of the things you wish they did in a game about exploring an unknown continent.

Discern Realities clarifies two things that I believe were present in the original rules but many people miss. One: basic sensory details are always provided by the GM in their description of what is happening or as answers to basic player questions. The Move is for learning more than that, including relying on your character’s in-world reasoning capabilities beyond those of the player. The example I gave in the document was that the GM will tell you about the six thumb-sized holes in the wall, but “What here is not what it appears to be” could reveal that they’re probably part of a dart trap and “what should I be on the lookout for?” would mention you should be looking out for those holes, they seem dangerous. Two: question 3 and 4 look like they ask the same thing, so we decided “what should I be on the lookout for?” meant dangerous things while “what here is useful or valuable” means good or useful things.

So there are lots of clarifications and minor alterations of those sorts.

For bigger alterations, Last Breath and Make Camp are good candidates.

Last Breath was a rare case of finding a bit of inspiration in Freebooters on the Frontier (or that other one that’s almost exactly like it, I forget the name). It now reads like this:

Last Breath
When you’re dying and someone checks your body within a few hours, you catch a glimpse of the true face of Death (the GM will describe it). Rumors have long said that Death will offer the strong a game of chance against their life. The rumors are true. Take Death’s Black Dice and Roll+NOTHING. ✴On a 10+, Death allows you to continue: you’re in a bad spot, but alive. ✴On a 7–9, Death offers you a bargain. Take it and stabilize or refuse and pass beyond the Black Gates into whatever fate awaits you. ✴On 6-, your fate is sealed. You are now marked as Death’s own, and must die. The GM will tell you when and how. Usually ‘right now, messily.’

Checking the body was from whatever source that was and I thought it was cool, because it’s like “did that 200 foot fall really kill them or did they miraculously survive somehow?” (Hey, it happened — three times, if you can believe it…). It adds just one more mystery to investigate in a game about investigating mysteries, and it adds a lot of gameplay. At least once a sole survivor rallied a whole new crew in town within the hour to embark on a new expedition to save their fallen comrades.

The other major change is that the 2d6+nothing roll is framed diegetically as literally playing a game of chance for your life with the avatar of Death. We decided early on to make the game as diegetic as possible, which extended to the Skype discussion group (which represented in-character conversations in the Cock & Whistle Tavern and Taphouse — actual in character conversation was rare here, but it was an abstracted form of something actually happening in the game world), the forum for game listings and town happenings were locations in the town (you’ve heard of Virtual Bulletin Board, now get ready for Virtual Real Bulletin Board), connected in-game and out-of-game time (each week out of game contained a month of time in game, leading to a new season every 3 weeks), and we even gave the GMs characters in the game world (two scribes who were hired to record the adventures of the players as they were reported to them and generally serve as their secretaries by providing the players with all kinds of important town information — I regret to say that experiments with playing as this character while running an actual game session met limited success and was dropped, but they were more successful outside of the game session).

In Dungeon World vanilla Make Camp basically heals you to full. But in this campaign nature is fundamentally hostile, so that didn’t make a lot of sense. We reduced the healing to 5+CON HP, and required you to carry rations or suffer serious consequences like accumulating debilities.

Some new Moves included things like Forced March (roll to resist fatigue), Carouse (a total re-write of the existing party move for spending money in town to throw a big party celebrating yourself and become famous, which awarded a kind of meta-game leaderboard points), or Return to Town (a total re-write of the Move for getting XP and resolving Bonds). There were also Moves specific to downtime in the town.

Regarding playbooks specifically, some required more re-writing than others. We decided at the outset that since it would be a dark fantasy setting and we would have a lot of players (we aimed for around 16, 8 to 10 of which would be playing at least once a week) we would have all the core classes (except Druid — being at one with nature in a game about exploring hostile unknown nature didn’t really fit), the official Barbarian, plus as many of the classes as we could justify from Grim World the fan supplement. We ended up adding one more (a Monk we cobbled together from various sources plus our own brains), which left us with 13 character classes.

The most major rewrites were probably of the Bard (who would you talk to? it’s just monsters and trees out there for the most part; we re-wrote it as more like Dandelion from The Witcher, a person focused on learning and knowledge from stories who accompanies the characters to record their exploits, plus a bit of charm and magic on the side and a special focus on the Fame meta-game), Monk (as mentioned it was created roughly ex nihilo as a blend between eastern-style martial artist and sage and western-style literate academic), and Ranger (more of a woodsman or hunter and less of a living compass and pet-owner to keep with the ‘nature is hostile’ theme). We did make minor changes to most classes, mainly in keeping with the theme of “players often can’t just make stuff up” and the needs of long term campaign play.

Finally, GM Moves largely remained the same. The main change to those was in how we had a system to standardize their usage. For example, under certain conditions we would always damage characters in certain ways in fights. This was partly for consistency that benefited the players trying to figure out the way things worked, but it was also partly because we had two people running sessions and we didn’t want the game to be noticeably different depending which GM you had. So a lot of the changes “under the hood” were playstyle things more than rule things. That said we did also make some systems of our own for things like exploration-based XP and things like that.

Did you document all the “under the hood” changes to MC moves, and applying them consistently? How did this work?

– User “Paul_T”

I assume we documented it, though I don’t know where that piece of history went so I’ll be relying on my memory here.

Before I get to the Moves I’d like to have a quick sidetrack into the Agenda/Principles, since all Moves ought to be made in accordance with the Agenda and Principles. In general, there’s nothing wrong with any of these for running a more traditional style prep-oriented game, they just require a certain point of view. For example, here’s what Adam Koebel told me when I sent him an email about my plan to run a more traditional or prep-heavy Dungeon World:

So, the big thing about the whole ‘collaborative storytelling universebuilding blah blah blah’ is that it’s a dial. Our marketing for the game is all about how you can include your players and stuff but honestly, you don’t have to. You can dial the player feedback meter to zero, if you really want to. Ask them ONLY what their characters are feeling and thinking and leave it at that. Don’t feel bad about it, either. I’ve run completely traditional style games of Dungeon World with a published adventure. The point is, then, to never play to guide the PCs in a certain direction. To let them explore the world and get into trouble wherever they like as they learn what you, as the GM, already know. It’s still fun, it’s still awesome and the core rules of the game are still there. Just ask questions whenever you feel like it and build on what you do ask about. Don’t feel like you have to hand over the authoritative control […]

Adam Koebel via eMail many years ago

So, that was emblematic of the sorts of ways we interpreted the principles and agenda. For example, to ask questions and use the answers doesn’t mean to ask players what’s in the chest they just opened (in fact, Koebel, LaTorra – and relatedly John Harper – all pretty much walked back on that kind of question ever being a good idea; these days they all prefer to ground player-established world details in things the character ought to know, at least).

The biggest changes in implementation in the GM Moves were probably related to combat stuff like Deal damage and wandering-monster-type uses of things like Show signs of an approaching threat.

I don’t remember all of the specifics, but for the former we had lots of rules of thumb like “on 9 or less Hack & Slash if it makes even the remotest bit of sense be sure to deal HP damage in addition to whatever else happens” or “warn PCs of potential bodily mutilation on the first hit before actually doing it.” Things like when players got hit automatically, what triggered Defy Danger before they could perform this Move, and how to handle fighting while outnumbered were all fairly standardized and all threats were designed to be strongly telegraphed.

Since combat was a big place where GM preference can really alter the difficulty and experience of the game we thought this area of standardization was one of the most important. We also re-did how monsters were generated pretty substantially, and we got a lot of mileage out of all of our monsters having specific weaknesses, somewhat beefier than vanilla DW stats (so we could rely less on being mean with our Moves and more on numbers), and building their Moves together so we both understood how they were supposed to work.

We definitely also came up with something about how PCs escape from monsters but I can’t recall what it was.

Regarding stuff like “Show signs of an approaching threat” used to indicate approaching monsters or something, we did also have a random encounter system we cooked up. It ended up being a bit overcomplicated and unfinished (a rule for anyone making a West Marches campaign: if you’re tempted to say “I’ll finish that later” be prepared for it to remain unfinished). When we did our post-campaign discussion, though, nobody had noticed anything wrong with it, so I guess it was sufficient.

We were also both very on board with the general way to play Dungeon World in a player-skill-challenge oriented fashion, namely that players should be actively seeking to get what they want by doing something that obviously works but doesn’t invoke a Move wherever possible (and only then resorting to Moves they’re good at, Moves they’re not good at, and doing things that will be obviously bad for them but with no corresponding Move, in that order). In post-campaign interviews the players seemed to get this idea pretty quickly if they hadn’t already, so I’d guess that meant our GM Moves on 7-9s or Misses were pretty punishing and pretty consistent, which is exactly how we intended it to be.

You can imagine the general case of how these changes were made by imagining two Dungeon World GMs coming up with a common game situation and then being like “how would you adjudicate that?” and then arguing about it for 45 minutes (or an hour… or two hours…) until we came to a rough consensus. That way we’d run the game consistently, and that’d go on to benefit any players who were interested in figuring out how the game worked “under the hood” for their own benefit.

We did end up having a bit of a fight over whether one or the other of us had adhered to these decisions properly at least once that I can remember (probably more like a small handful of times). But by and large it worked well, and the impression I got from the campaign veterans was that it felt pretty consistent and predictable. So the bottom line is probably just that deciding to follow certain rules of thumb to standardize your approach to the game works, whether or not you mess with the Moves themselves or what the rules of thumb you develop happen to be.

From my point of view the underlying consistency / prep / no-fudge attitude was obvious. Telltale signs included (as Paul would predict) occasional boring sessions where we went out exploring somewhere and didn’t find much of anything, and the occasional player death due to bad decisions and bad dice rolls. This was very much what I had signed up for and I thought it was obvious that this was what the game was going to be like from the GMs’ pitch.

User “Vivificient” (Player in the campaign)

One thing we tried to do to minimize sessions that were a bust was to ask for not only one plan of action, but a backup plan of action in case the main plan was a bust. The risk with not doing that is either a) the session is a total bust or b) they get there then try going someplace you hadn’t prepared yet when they realize it’s going to be a bust. By having the backup plan in mind you need to hit two busts in a row, which is quite rare, so you won’t have to worry as much about players running off the edge of the map.

[Responding to a case where players didn’t find the world to feel fair and engaging even though West Marches is supposed to feel very fair because of its neutral arbiter of a GM.]

Huh. So not enough transparency of method…? When it’s not clear to what extent the gloracle is responsible, the GM takes the blame?

User “2097”

[2021 Hydra here: This user uses the term “gloracle” to refer to game outcomes generated by the game system, like rolling dice, rather than player decisions.]

I have a story about this, not from this game of DW WM but from a prior West Marches game I ran in 4th edition D&D (you can see now where my bad experience with explaining how to build characters comes from). It involves a player, let’s call them D.

D and some pals were exploring some crypts near the town when one of his pals has to leave unexpectedly. The guy who had to go unexpectedly was a cool dude so he said “you guys can run my character to help out in a fight or whatever just don’t get me killed” so the session could continue. D is a bit of a power gamer so this is no problem, he can run any character in a combat. As the session goes on they run out of healing surges, which in 4th edition means they basically can’t recover hit points anymore and this is a dangerous state of affairs to be in. D’s compatriots want to return to the town, wary of losing their comrade’s character in his absence, but D says (and this is nearly a direct quote): “You can go back if you want, but I bet the boss is behind this next door so I’m going in whether you come with me or not.”

You can imagine what happened next.

This was the singularly most deserved PC death I have witnessed in thousands of hours of RPG play except perhaps the time a Barbarian drank from a clearly labeled fountain of acid with a skeleton in it. They had already guessed successfully that a big scary monster was behind this next door, their friends urged them not to, and they ran ahead into a trap I was in the middle of describing. There’s really not much I can do to help stop that series of events if I’m supposed to be a hands-off referee.

Nonetheless, I was blamed (called a “Killer DM,” in fact, although this player had played in over 20 sessions of a prior campaign I ran without objection). Technically it was my fault: I built the whole world they were playing in, and I could have given every monster 1 hit point and 0 damage if I wanted. But I don’t think that’s very reasonable, so instead I mostly tried to follow the DMG’s encounter guidelines. The lesson I took away from this is that some players are just determined to see the GM’s hand in everything no matter how strictly they follow design guidelines and telegraph dangers and all the rest.

Long story short, I’d imagine there are parallels between [this case and the case of a player in DW WM finding the game unfair]. But that’s just a guess. The unfortunate thing about players leaving online games is they usually don’t stick around to do an exit interview.

“We want to do DW… but is WM really the best fit for it?” I think you’ve answered in a way that makes me say yes.

“We want to do WM… but is DW really the best fit for it?” I’d rather personally use another rules engine for it. What’s your take, in hindsight?

User “2097”

It’s interesting that you think I’ve answered that first question rather than the second, because from my perspective I’ve been framing all of my answers so far in terms of the second question — probably because that’s the question I was trying to answer myself at the time! I had already run a West Marches style campaign and decided I wanted to run another, but bigger and better. And looking at all of the games I could use to run it, I settled on Dungeon World as the best fit for my needs.

Dungeon World is a poor fit for West Marches on first impression, no doubt. One time I tweeted Ben about that and he basically said the same thing. I believe the reasoning is that from his perspective what makes West Marches unique isn’t the rotating cast, the wilderness exploration, or even the episodic gaming on demand. More-so it’s the revolution of 3rd edition being a game that can sort of “run itself.” 1st edition AD&D was meant to be to basic what 3rd edition ended up actually being to everything that came before it: a set of rules that could be run “without interpretation.” If the rules are laws and DMs are judges, then West Marches and 3rd edition make the DM an historical literalist: strictly what the text says – no interpreting author’s intent and no active revisionism, thanks.

To the extent that I want to run “a campaign like Ben’s” I appreciate and try to imitate that, but it’s also not 2001 anymore and I’ve had different experiences than he’s had, so I have my own take on the canonical West Marches as reported and what would make for a “good fit” with it.

For example, recall my story earlier that showed me how playing 3rd or 4th edition D&D strictly by the book is not necessarily a shield from accusations of “activist judging.” John Wick (the game designer, not the action movie character) once related an even more brutal story of how running his group through the Tomb of Horrors when he was a kid led to one of his friends punching him right in the face. That definitely wasn’t from being a biased judge in executing the combat with the gargoyle. Hell, people get mad at video games for being “unfair,” and they’re literally arbitrated with perfect consistency by a computer.

At a certain point it’s up to the player to suspend disbelief in order to achieve Ben’s vaunted “detached puzzle solving” effect. I decided that the way I played Dungeon World and the changes I had made to it would suffice for as far as I was willing to bend over backwards to facilitate that, and players would just have to do the rest. The game had tools for the players to access GM-provided hints if they wanted them (things like Spout Lore and Discern Realities). My tendency to provide the results of actions before rolling, if anything, was even more openly accessible than an inviolate player’s handbook because, as Vivificient pointed out, a player wouldn’t forget relevant things in the way they often forget about feats they have or adding circumstantial bonuses in modern D&D. Ben spoke disparagingly of “mother may I?”, and while I’m sure my players asked me clarifying questions sometimes, by and large they mostly just did what made sense to them and saw plausible results.

Here are all the things Ben said were key to a West Marches game listed explicitly: No regular time. No regular party. No regular plot. Exploring a fantasy wilderness using first person turn-by-turn navigation referencing a GM-kept vector map. A table map. Session summaries. Random encounter tables and restocking locations. Things getting more dangerous the farther from town you go, with pockets of danger nearer by that are well telegraphed. A safe town and wild wilds. No NPC adventurers. Disclaiming responsibility for the world’s contents. A treasure map to get them started. No town games. Players who try to out-do one another. A meta-game scheduling location like a forum or mailing list. The rules are available in a player-facing document. Combat is a tactical puzzle with little hidden information. Your world should have a fleshed out history that explains why things are in the state they currently are, but which you shouldn’t share with your players except as they discover it through play.

None or few of these things are actually system-specific, so you’ll be bolting stuff on to just about any game you play. And of the things that are system-specific, any game that looks even vaguely in the direction of system mastery and challenge play basically has what you need to succeed in principle, and from there it’s just a matter of degree as I’ve just explained about players and suspending disbelief. What made 3rd edition revolutionary in 2000 — having clear rules that anybody can look up whenever they want and expected wealth by level tables and encounter building rules — is just normal practice today for a whole swath of games. As long as you avoid games like Fate that explicitly model stories instead of physics you’ll be just fine (assuming your West Marches setting isn’t in Faerie — then having physics be stories makes a lot of sense, actually!).

For my specific needs, I found that the non-negotiables in a game were having enough players to spin up a strong metagame, being mechanically light enough so I didn’t burn out, being immersive and supporting challenge/mastery-oriented play to some acceptable degree, and that I was familiar with and just enjoyed playing it. Any game that didn’t offer those things was a non-starter, and any game that did I could bolt on the rest.

Dungeon World offered, I bolted-on, and the rest is the history you’ve been reading.

To Be Continued

That’s the end of part 1! Join me soon for part 2 where I cover monster design and the challenges of running West Marches, which needs consistency, with multiple game-masters who might be anything but consistent.

Discourses on GMing Series 1 Post 7

Today’s discourse will feature chapters 5 and 6 from Jamison’s “Gamemastering” eBook. It’s free, so I encourage you to read along – else I might not make a lot of sense!

Notecards, Notecards Everywhere…

At first I was convinced Jamison’s advice about notecards was similar to his advice about player-characters of a non-player gender or so on. It seemed like taking a minor problem and exaggerating it into overkill proportions. Who needs this many notecards?

Bluntly, I was wrong. The notecards are more important than they appear, and it’s because of something psychologists call cognitive load theory.

You know that saying that goes something like “you can only think about 5 to 9 things at once”? It’s true, except it’s actually even more restrictive than it sounds. If you’re manipulating the information, like how you would combine “character is good at jumping” with “a pit obstacle requires jumping” during prep, then the number of items you can keep in mind drops to between 2 and 4.

There’s good news and bad news here. The good news is: the more familiar you are with the information you’re using the more of it you can use at once. That’s why this list of letters –


…is a lot harder for most Americans to memorize than this list of letters –


It’s the same list of letters, but by re-grouping them in a meaningful way you can remember them much more easily due to the familiarity. The same goes for things like the player-characters’ lists of skills, goals, NPC verbal tics, and so on. So that’s the bad news: we don’t have eight spare hours to become familiar with (read: memorize) all this stuff.

That’s where externalizing the information in the form of notecards comes in. By moving the information onto notecards instead of holding it in your head you can manipulate a lot more of it at once without having to first memorize it all. Think of it like that cork board filled with news clippings and photos and red string connected by thumb tacks in police procedural TV shows. Those detectives know how important it is to externalize the vast amount of information in their cases, and the same goes for you and your campaign!

Giving ’em What They Want

This is the core theme of the book: to have a fun game, give the players what they want in a way they don’t expect. This is why assembling all that information before the game began was so important. The players have by this point told you what they want: the NPCs, the kinds of challenges, and so on. Now it’s your job as the GM to decide how they get it. Changing game prep from inventing new content to instead merely inventing a new presentation is how Jamison can say with a straight face that he can prepare a 4 hour game every week in only 15 minutes.

The result of the method is what I consider a nearly air-tight adventure design. Here is a dumb example to bluntly highlight how it works:

Jumpman (PC) is a superhero who can jump really high (skills). He needs to defeat his nemesis Runman (Foe NPC) who is wrecking Peaceburg (starting location), but he can never catch him and needs some help. Jumpman’s teacher, Mentorman (Friend NPC), offers to help him. However, first he needs Jumpman to go retrieve the Widget of Goodness from the top of a mountain (obstacle resolved with skills) so that he can recharge his super powers and come out of retirement (woe).

While hopelessly boorish by actual writing standards I think it illustrates my point nicely. If you’re Jumpman and you’re presented with the adventure to the top of Mount Highjump in return for help in defeating your nefarious rival why wouldn’t you take up that offer? You get to do something challenging, show off how cool you are, and achieve your goals if all goes to plan.

This is why “giving ’em what they want” is such a successful method of preparing games. The players tell you what they want, and then you spend a few minutes a week shuffling notecards around a table and – voila – out pops an irresistible, bespoke adventure better than any you could have ever bought at a store.

Just, uh, make sure your players are more creative than “Jumpman” here.

Next Time

Next time I’ll be discussing the finer details of building adventures and some more of Jamison’s pointed advice from sidebars that completely miss the point.

Discourses on GMing Series 1 Post 6

Character Creation

There’s a saying you may have heard if you know a teacher: “I taught them but they didn’t learn it.”

Of course, no teacher can force a student to learn – they do have free will, after all. But often it isn’t enough just to tell students something and expect them to learn it – “teaching by telling and hoping” I’ve heard it called.

“Game design by telling and hoping” is the exact thing Jamison warns against in the introduction to this chapter, and it’s one of the most important things he says in the whole book.

What is the thing most GMs do when they first make an adventure and only then get their friends to make some characters? What criteria are they using to design the adventure so that it’s fun? At best they’re wildly guessing what their players will find to be fun. More commonly they just make something they think is cool and hope the players think so, too. In other words, they’re running their game by crossing their fingers and hoping the players just go along with it.

When is the last time the players “just went along with it”? The last blue moon, maybe? “Campaign that ran off the rails in the first 30 seconds” is possibly the most popular D&D story genre out there. In an online game you can advertise your game as-is and players will self-select: if they think it’s cool they’ll show up, if not they won’t. But in a more traditional game your campaign will only ever be fun by luck. That’s not acceptable. So, let’s give ’em what they want instead.

Jamison’s Rules for Good Characters

They must be likable. They should not be created randomly (e.g. rolling stats).

These are not wildly controversial opinions, and they make perfect sense given Jamison’s insistence on games about the story of a core cast of competent player-characters. I think there’s room for a randomly generated character in that kind of game, but they’re more like a challenge than a standard. They can easily be too weak, or too strong, or so on. Much easier to build them by hand.

Regarding likable characters this is much less obvious advice than it appears. I have heard a lot of stories over the years of chaotic evil loners, and even seen one or two myself. You know who doesn’t hang out in a tight-knit team? Chaotic evil loners. Again, maybe playing against type here could be a fun challenge, but it just isn’t the standard.

Finally, Jamison insists that characters begin relatively weak. Why? It allows for a wider variety of challenges across the length of a campaign, players like becoming more powerful (the SAPS – Status, Access, Power, Stuff – model describes this particular form of player motivation). Sounds reasonable to me.

The Character Interview Process

Core to Jamison’s method of adventure preparation is the idea of building it around the characters. Naturally, this rules out the idea of “discovering about your character in play” (the GM needs those details to build with!).

Before we get to the method itself, however, first I want to make a comment about the “Difficult or Impossible Roles” sidebar. I think a lot of what Jamison wrote in this box is, to be blunt, extremely inconsistent. Is it very hard to play a character of another gender or with a mental illness convincingly? Yes. But here’s the thing. We expect the GM to play dozens of characters like this throughout the course of a campaign. So then why shouldn’t a player be allowed to play just one? Be cautious? Sure. But disallow it? Seems invalid to me.

The character creation method, however, is both very consistent and very useful. It has four parts: Rights & Wrongs, Friends & Foes, Goals, and Quirks & Traits. These are all useful information, both for the player and for the GM.

Rights & Wrongs are critical information in the creation of challenges. Baby Orcs in particular may be overdone, but utilitarian dilemmas are a classic of drama: will you do the wrong thing for the right reason? How about if the characters don’t agree on what’s right and what’s wrong? The other key point here, I think, is how Sins & Virtues uses a six point scale. Six point scales force people to pick sides, sides lead to conflict, and conflict leads to a good story.

Goals are necessary because they are literally what the character (and thus the player) is interested in doing. It will reduce your plot hook rejection rate to nearly zero, which is ideal for Jamison’s method – no need to prepare extra material in case the players don’t bite.

Quirks and traits are mostly helpful for the player because they are a list of things the player should keep in mind while play-acting as their character, like accents, temperament, and common behaviors. A particular favorite of mine is to lift the concept of weaknesses for this section from the STALKER Scifi Roleplaying Game.

Finally, consider Friends & Foes. Together with character goals I think this is one of the best ideas in the whole book when it comes to minimizing wasted prep work. The reason for this is because of, as I mentioned in the last post, the IKEA effect.

How many times have you tried to introduce a villain and the players all just made fun of him? It sucks and it happens all the time. How do you get instant player buy-in? Easy. You get them to make the villain themselves! Nobody thinks their own creation is stupid, after all. With a player and the GM instantly bought in that makes buy-in from everyone else a lot easier to come by. Of course you don’t want anyone in charge of their own opposition, so the advice to hand the foe off to the GM for further work after coming up with the skeleton is sound.

The reason I like this trick so much is that it completely generalizes. Anything that really needs player buy-in should be something you try to get the players to have a hand in making. It’ll save you a load of headaches, believe me.

Next Time

The preparations really heat up! Jamison’s method for fleshing out NPCs and creating the core of an adventure – the obstacles!

Discourses on GMing Series 1 Post 5

The End of Chapter 2: The Rules

To wrap up chapter 2 I want to talk about something Jamison wrote that reminded me of my theory of Game-as-Training-Wheels. I have noticed the trend he references – heavier systems being deemed less useful for experts – in myself as well as in many people I know. This seems to support my idea that rules are useful as guidelines for newbies, but they are more cumbersome than informative for experts who have already developed a deep understanding of “what works”. That said, it’s not easy to find players for a game listed as “homebrew,” so making your own custom light RPG after 20 years of play experience is really advice that only applies to those who have a group of players already on board, I think.

Even just with house rules something I tried to do in my last campaign was to make them as GM-facing as possible. Nobody likes to be confronted with a 20 page errata booklet. Learning rules from a book is tedious and hard and takes too long. So it’s best if you can make the changes invisible to the players as you strip down the rules and customize them to your heart’s content. Other than that I am on board with Jamison’s advice here.

Chapter 3: Creating the Universe

Jamison is right. Most GMs do spend too long making the universe. Nobody is going to buy the 300 page setting book you crank out, unfortunately, and most of it probably won’t even feature in the game you run. That’s a lot of wasted effort. That said, that’s not necessarily a reason to just wing it all at the table as we’ve established earlier. So let’s see what Jamison thinks are the high-impact-low-time elements.


Jamison recommends making at most two societies: the one the characters are from and the antagonist society. This is not a hard and fast rule in my eyes – an easy way to make more societies is for a player to decide they want their character to be from somewhere you haven’t made yet. But when it comes to the GM making societies, I think it’s fair to say “the place you start” and “the place a lot of your obstacles come from” are pretty high-impact.

Additionally, Jamison’s tables of suggestions here and throughout the book are very handy. They aren’t anything an expert couldn’t think of themselves, but they are useful story stuff. You need to know who’s in charge, why they’re in charge, and what rules they’ve decreed to inform your cast of NPCs and creation of conflicts in the world. You need to know how to tell different people apart (in terms of social class and beliefs and so on) so you can describe them, and how they speak so you can act them out.

The economics is similarly basic: what does the society make and what do they need on the whole? But it’s similarly very useful. Conflict is created by scarcity, and good stories are driven by conflict because conflict is action. So knowing what’s scarce and what’s in abundance is going to be a key ingredient in creating interesting conflict. Is every conflict going to be about mustard smuggling? Well, that’s up to your personal taste.

(Me? I like a little mustard on my adventure once in a blue moon or so…)

The map-making procedure is similarly utilitarian. You can make a visually brilliant map if you like, but that really brings us to two major understandings of world building that are critically important:

Most GMs do most of their world-building because they think it’s fun, not because it’s useful.

As demonstrated, you can crank out the important details of a society (village, town, city, empire, these things scale up and down as you like) in, like, an hour. Then you can sketch a map in another hour or so. Maybe spend a third hour on an antagonist faction and boom your world is ready.

So why do we spend weeks on this stuff? Frankly, it’s because we like it. And if we don’t like it? Well, it turns out we don’t need to spend very long on it. So remember that the next time you’re feeling the burn(out).

Most GMs like their world a lot more than their players do.

Additionally, if you look at how people value things one important fact that stands out is the IKEA effect. Basically, people strongly over-value things that they make compared to identical items they did not make. By about 50%, if we’re being specific. As the GM you probably care about the world you’re making 50% more than your players do, so try to keep that in perspective.

Similarly, the same team that discovered the IKEA effect went on to do more experiments in a similar vein. For example, having people construct LEGO robots for 2 dollars a piece. In one condition they were disassembled for parts and you could build another if you liked, while the other you could build another out of entirely new parts while the original remained assembled. Participants completed over 50% more robots when given new parts – when their existing work was valued.

The point is that this goes both ways. GMs overvalue their work and will put in extra effort if they see their work being “properly” valued by the players. Similarly, GMs, try not to value your world too much – and know that players value their own creations as much as you do, even if you are putting in more work than they are overall.

This effect is part of why I think a particular step of Jamison’s character creation method is brilliant, and I’ll say more about that next time!

Next Time: Character Creation

Jamison’s rules for this are a bit idiosyncratic, much like some of his “how to play RPGs right” preferences were. Should be interesting!

Discourses on GMing Series 1 Post 4

All About Players

So, here we are: the very beginning of a campaign as according to Jamison. Probably the most important thing he says in this chapter is his introductory note, I would wager – that people like to classify things, but that commonly when it comes to categorizing people we tend to not do a very solid job. Probably because people are complicated and don’t fit neatly into categories. Nonetheless, Jamison offers his reduction of players into the categories that he finds to be most useful in preparing a game.

He divides players into three dichotomies leading to eight possible combinations: chaos vs. balance, acting vs.action, and storytellers vs. realists.

Chaos vs. Balance

Chaos vs. balance is the player equivalent to the traditional chaos vs. lawful axis in D&D’s alignment system: one likes to break the conventions of the fictional universe and the other likes to fit in with them. While he suggests these two kinds of players won’t get along, I think there is an important caveat. Consider beer’n’pretzels style D&D: Sir Bearington fighting alongside Prissy Pants the Paladin fighting alongside the Eladrin Corsair Calico Knifegutz because somebody’s gotta stop the end of the world and loot every village on the way there. The less the players care about immersion and realism the less they’ll care about mixing law with chaos. So I am not sure this axis is as distinct from the “Acting vs. Action” axis as it may seem.

Acting vs. Action

Speaking of which, in this dichotomy the former likes to play-act as their character (which often involves dialogue and emotion) and the latter likes to treat their character as an avatar for doing Cool Action Stuff ™.

This reminds me of the first time I played D&D online. Word got out the DM was running two separate campaigns, one that was a bit more action and the other that was a bit more “serious roleplaying.” None of us really had any experience besides the DM, but for whatever reason we all suddenly wanted to be in the serious roleplaying campaign.

I’ve found that, given the choice, most people will say the same. But that’s not actually true. This was the great lesson of Howard Moskowitz’s work in the food industry (as described nicely in this TED talk by Malcolm Gladwell). When you ask people what kind of coffee they like they will usually report that they like bold dark coffee. But if this was true then how could Starbucks possibly survive when 90% of its menu consists of weak coffee-flavored milkshakes? It turns out that people will report liking what they are “supposed” to like according to their culture regardless of their actual preferences – hence blind tasting panels being the food industry standard these days.

I have similar worries, sometimes, about player preferences. How do we most often report them? We just ask the player. Giving them what they ask for is good, but it’s not necessarily the same as giving them what they want. Perhaps that’s another point in favor of the convention/online “take it or leave it” model. For what it’s worth, I usually ask new players in my games to tell me a story about a time they played an RPG before. Their go-to story tends to be related to their preferences in a way a simple self-report would fail to capture (but telling a story is a lot less obnoxious than a questionnaire).

Storytellers vs. Realists

The final dichotomy presented has a similar historical pedigree to the first: it’s basically narrativism and simulationism in the GNS (gamist/narrativist/simulationist) theory from The Forge. I seem to recall a particularly sharp expectation mismatch from my last campaign where a player wanted more “cinematic” description – in the literal sense of describing camera movement and portraying scenes to the players that their characters were not present for. The game was advertised as emphasizing simulation. That combination did not last long.

Other models of player desires?

In my time researching game design I have found numerous different taxonomies of player desires. Raph Koster suggested in A Theory of Fun for Game Design that really all games are fun because they are about learning in a particularly interesting way.

Richard Bartle’s Hearts, Diamonds, Clubs, and Spades: Players Who Suit MUDs suggests there are two axes: acting-interacting and players-world. Killers like acting on players (PvP), achievers like acting on the world (PvE), socializers like interacting with players (cooperation/guilds), and explorers like interacting with the world (finding secrets).

Hunicke, LeBlanc, and Zubec’s MDA Framework suggests there are possibly eight kinds of things players experience that they enjoy:

  • Aesthetics/sensory input (novel or pleasing figures, graphics, etc.).
  • Fantasy (immersion).
  • Narrative (story).
  • Challenge (winning/losing/mastery).
  • Fellowship (other people).
  • Discovery (learning something new).
  • Expression (making something new).
  • Submission (as an identity for the player, for use in relaxing).

Nick Yee’s Quantic Foundry video game model breaks gamer motivations into six associated pairs:

And his board game model breaks motivations into four categories with seven sub-categories:

My point here, really, is that we’ve been trying to classify gamers for a long time. Rigorous attempts have existed since at least the 1990s. And I think something I’ve said before about unreliable personality tests applies here as well.

Lots of people have heard of the Meyers-Briggs personality test. Those of you keeping up with your academic psychology will know that it has poor test-retest reliability, which makes it a very poor test of character traits (which are supposed to be stable across time). These gaming motivation tests are probably pretty similar in that regard: they are not very reliable or highly accurate.

Nonetheless, if you gave a group of people a 50 question quiz about how much they like Star Wars (which by all means does not sound like a very useful personality test) and then grouped them into “likes Star Wars” groups and “does not like Star Wars” groups you would wind up with two groups who get along better than the whole group overall. Even if the box you use is stupid and arbitrary and not very reliable, the very fact that you introduce boxes into which to sort players provides a sort of service. At the very least it will get you the equivalent of the Robbers Cave experiment and get you a feeling of being an in-group together.

Now, do I recommend throwing personality surveys at your players? No. I’ve tried it and it wasn’t as helpful as the effort warranted. But you can certainly ask your players a couple questions and, assuming their self-reports are accurate, you can get a pretty alright idea of what they want based on what they say they want, or what they have enjoyed in the past.

Miscellaneous Commentary

Finally, Jamison ends his section on players with a few small asides. First, he claims wargamers care too much about winning to be effective RPG participants. I don’t agree with this on two counts. First, the original RPG inventors and players were all wargamers. Maybe Gygax and Arneson and their crew of 40 friends were all terrible at RPGs, but that seems extremely unlikely. The second count is the idea that to win is to have losers. Players win RPGs all the time – in fact, that is their expected condition. Jamison himself said the referee cannot be “neutral” because they should be ultimately giving the players wins. RPGs are cooperative games – you can win 100% of the time with no losers if you like. So how is a player set on winning – who will be doing so constantly through the nature of Jamison’s system – going to disrupt the game? Wasn’t our creed to give ’em what they want? I think the more likely explanation is that Jamison has simply had some bad experiences with people he met playing Warhammer at his local game store or something. I know the kind of problem player he is describing here, but I would not call them a “wargamer” – I would call them a socially underdeveloped person. Plenty of people play wargames and make perfectly fine RPG players. Wargames have little to do with it.

Second, he suggests the fewer the players the better (to a certain point). Collaborating is fun, but being crowded out is not. The longer I have played RPGs the more I find myself agreeing with him: more than 4 players starts to be a bit of a mess in a modern story-laden game. An exercise I strongly recommend for any among you who haven’t played in a while: give playing alongside 5 other people a go sometime and take note of how much time you spend doing nothing. Since the GM is often the portal through which all gameplay must flow that keeps the GM very busy. It’s easy not to notice just how much time players spend waiting on other players. Empathy in this regard is very important.

Finally, everything he says about pitching a game is great advice. Separating the discussion of whether to play from what to play is probably smart given his premises, and it’s just good life advice to have your ideas polished down to an elevator-pitch-level sheen. I was very careful in my last online game, for example, to keep my pitch to a single page of text.

Next Time

Picking the game system & creating the universe. Yay world-building!

Discourses on GMing Series 1 Post 3

Time to actually get to the heart of the matter. Well, chapter 1. But chapter 1 is the heart of the matter in my opinion. Let me tell you why…

Brian Jamison’s Only Immutable Law of Gamemastering

“The more the Gamemaster plots, the less the players will follow the plot.”

I think this phrase is the foundational wisdom of Jamison’s system for preparing and running RPGs. He’s right, I think, when he says “Traditionally … the Gamemaster labors for hours … working out plot and story, then springs the pre-written adventure on the team. The Gamemaster then hopes the players will enjoy the adventure and has given himself little opportunity to change things if they don’t.” I see this on RPG forums all the time. Somebody posts asking for help with a plot point, but the issue is that their plots assume the players will behave a certain way. They are acting like a writer – who has total control – not a game designer, who doesn’t.

Once I read a post on reddit where someone was asking for help with their first session. It was a GURPS game where the PCs would be modern university students thrown back in time during a visit to the large hadron collider. That’s a cool enough premise, right? But the entire thing was prepared completely backwards: first the PCs will do this, then the NPCs will talk to each other, then the PCs will be captured, then they will be rescued by more NPCs… it was classic “GMing as writing” that assumed the players would stick to the script they were (not) given.

For all of his idiosyncrasies from the introduction (and more yet to come), Jamison’s entire method is fundamentally based on this one gem of a paragraph:

Plotting or scripting outcomes are totally against the spirit of roleplaying! Playing an already-written story is boring for the players because they won’t have any input, and boring for the GM because the story is already written. In fact, any game that has a predetermined outcome isn’t a game. GMs sometimes write a story in advance because that’s what they want to do. These folks should write a book/play/movie and get it published or produced. Likewise, if the players want to act in a scripted environment, they’d probably have more fun with a theater troupe.

Truer words have never been written about roleplaying games. Games have players, and players make decisions. If your game doesn’t have players making decisions then pretty much it’s not a game.

There are two elements from this idea that make Jamison’s preparation method so lean. First, he only tells you to prepare things you probably won’t be good at improvising on the spot. Second, he only tells you to prepare things that will be useful in a game environment where you cannot be 100% sure that your prepared material will be useful in a given instance. The entire method is about “preparing to improvise,” as I have come to call it. You can’t have a game without improvising – games need decisions to be made on the spot. But you can’t make good decisions on the spot unless you’re prepared. So you need to prepare to improvise.

The Jamison Order

Much like the Machete Order re-arranges the Star Wars movie viewing sequence, Jamison attempts to re-arrange the sequence of launching a new roleplaying campaign to be more optimal. He claims the traditional order (which I agree is often what happens) is:

  1. GM chooses game system
  2. GM readies the adventure (buys/reads/prepares it)
  3. Characters are created
  4. Start playing

Now, this method has a few merits online that Jamison doesn’t consider because of the meat-space limitations of his scope. Online it’s very easy to start a pick up game, most commonly by somebody willing to GM (due to supply and demand). There are also great tools for sorting through games to join in places like Roll20 so that you’ll basically get the game you want. Simply because of the immense scale of the online RPG campaign market it’s actually quite practical to present a game “take it or leave it.” This is also how convention games operate – there are so many people (and so little time) you can afford to advertise a pre-prepared game and have it work.

However, especially at an in-person game, I think Jamison’s order makes a lot of sense:

  1. GM chooses the players
  2. Everyone agrees on a setting
  3. GM chooses game system
  4. Characters are co-created
  5. Adventure skeleton is written
  6. Start playing

In teaching there is a lot of hubbub about “student-centered” education – teaching students things they will believe are relevant and interesting. Jamison here presents a sort of “player-centered campaign design” – pick some people you like first, then construct everything around the group’s shared preferences. It’s pretty hard to argue with the “give ’em what they want!” approach. What’s the alternative? Give ’em what they don’t want? Seems impractical. At least students legally have to be there – players have no such compulsion to stick around a game not meeting their preferences.

Next Time…

We’re on to chapter 2: Choosing Players and the Game System.