Torchbearer In-Depth

In RPG discussion circles in places like Reddit and Twitter, there has been a fair amount of hay made over the relative absence of ‘critique’ in the RPG realm. As someone who reviews RPGs fairly regularly, this is something which is difficult to parse but ultimately fair. A review aims to go over the traits and writing quality of a game in order to answer whether it is worth buying and worth trying to play. In RPGs (though honestly in all media) reviews skew positive because negative reviews get more negative attention, and because honestly writing 1500 words about something you don’t like simply isn’t fun.

So then what is critique? I think the simplest answer here is that a review attempts to summarize the quality of a product, whereas critique aims to investigate what the game attempts to do, and how well. Due to the mutability of the role-playing game as a form, what critique really looks like is still evolving. That said, I think there’s one constant in how RPG critique is defined: While it is possible to review an RPG system or supplement merely by reading it, critique must be rooted in play, as RPGs are meant to be played. Which brings us here. While I read and then discussed Torchbearer a while back, earlier this year I actually ran a short game of Torchbearer. And Torchbearer as a game, especially as the game I ran immediately after concluding a D&D campaign, serves both as a great foil to D&D as well as an example by contrast of how embedded “D&Disms” are in the hobby.

The Game

Torchbearer is a game designed by Thor Olavsrud and Luke Crane, built on a modified version of the Burning Wheel ruleset. Thematically, the game focuses on adventurers who must delve deep into underground dungeons and face dangerous monsters in order to collect enough treasure to continue to fund their meager existence. If this sounds familiar, it’s because it’s a direct thematic overlap with Dungeons and Dragons, especially earlier editions of Dungeons and Dragons. This is deliberate, as Torchbearer is a love letter to old-school D&D as well as an extension of Burning Wheel. Character creation emphasizes this with a version of the race-as-class character options familiar to players of Basic D&D, and the somewhat tortured overlay of levels onto the starkly different advancement system from Burning Wheel drives the point home.

After you make your level 1 adventurers, though, the mechanics of the game completely depart. There are two core gameplay loops, joined together in an overarching timekeeping mechanic called The Grind. The gameplay loop my group spent the most time with (and based on Burning Wheel, the one which is supposed to be dominant) was the Testing mechanics. Bolted onto that system are a set of rules for conflicts, using the same core as Burning Wheel’s conflict mechanics but pared down to be easier to use.

Testing mechanics are the game’s skill system, and this skill system says a lot about the dice-rolling philosophy of its designers (a philosophy which is also conveyed in Burning Wheel, but muddied a bit for reasons both the designer’s fault and not). If you are rolling a test in Torchbearer, it better damn mean something. First, there’s The Grind, which means that you are exhausting resources every four actions you make (generally tests or conflicts). Second, there are Twists, which give the GM license to make something bad happen should you fail a test. How this played out for us was a lot of planning. As is true in Burning Wheel, raw dice quantities are biased towards failure, heavily. The game’s math centers around a raw pool being equal to the obstacle number, which for a pool and obstacle of 3 implies a roughly 13% chance of success. Of course you aren’t supposed to roll with a raw pool, you’re supposed to have equipment (there’s a die) and friends who can help (one or two dice). This makes every test an event, whereas in games like D&D it’s expected that there’s a relatively quick cadence of d20s hitting the table.

Another thing that’s interesting here, which I didn’t play with as much as I could have, is how failure is defined in this game. The players are never required to walk away from a roll just not getting what they were going for. A player can always get what they want, but with a Twist or with taking a Consequence. Twists are meant to be mean, they’re meant to hurt. That’s one reason I didn’t really lean into them…there are circumstances in the game where an appropriate twist is “and then the character dies”. Now, this is a game where the consequences are supposed to be broadcast, and the player will know this is a possibility, but…I personally am never going to feel comfortable making character death the outcome of a single die roll.

Character death isn’t supposed to feel like it hinges on a single die roll in Torchbearer, but to get where it succeeds and fails at that setup we have to talk about consequences a bit. There are six core consequences in the game: hungry/thirsty, angry, afraid, exhausted, sick, and injured. There are two others that sit on either side of the continuum: Fresh, which is basically an inducement for players to spend their treasure on a die bonus, and Dead, which due to the (current?) lack of resurrection mechanics in the game doesn’t really fit as a condition. Each of these six conditions penalize the character in some way, and once a character gets two or three there is a bit of a death spiral that starts. At the beginning of the game, not understanding how the math would fall, my players accepted conditions for failed rolls fairly readily. By the end of the game, even characters who were fairly healthy were loath to do it. Conditions hang over everything and are hard to get rid of. An interesting thing I noticed, though, was that even though The Grind is theoretically the pacing centerpiece of the game, I think there was maybe one time that any character actually got a condition by not having rations as The Grind came around. The Grind as a pacing mechanic is more of a fail state than something to be interacted with.

So there’s The Grind, and there are Tests. There are also Conflicts, which are way more interesting than combat in most systems. My favorite thing about Conflicts in Torchbearer is the stakes-setting. Your players get to define what they’re trying to do with the Conflict, and in doing so set the stakes. Trying to drive off an animal? Well if you fail you might be driven out of the woods instead. Trying to convince a crowd? Well if you fail then people may start to hate you. Trying to kill someone or something? They might take one of your party members down with them. That last one got my players to instantly reframe how they thought about combat. The group was a little more cavalier once they realized what they could do to regain disposition (a proxy for hit points which tracks the combat), but the idea that at the end of the conflict you simply choose someone who dies is weighty.

The conflict system is neat, but a place where I wish it was possible to leave more of the detail in from Burning Wheel without overcomplicating it. There are four actions: Attack, Defend, Maneuver, and Feint. There’s still some strategy involved, and the three action design of each round means that players will be coordinating and discussing strategy just as much as if you were running combat with minis. Still, though, as each opponent has basically the same four actions to choose from, it becomes difficult to differentiate encounters. This is made worse by the fact that the same mechanics are used for combat and social encounters. D&D might not be all that much more differentiated, but it has way more window dressing.

I should at least mention the structure and phases of the game as well. The Grind is featured during the Adventure phase, while the core game also consists of the Town and Camp phases. The Camp phase combines a version of D&D’s rest mechanic with some meta-currency. If you earn ‘checks’ by roleplaying to your mechanical detriment during the Adventure phase, you can spend those checks to try and recover conditions in Camp. My players immediately grokked this, though Camp was often just a place to remember how difficult some of those recovery rolls can be. The Town phase gives you more free reign to recover conditions, as well as shop, learn spells, find rumors, and other things. The rub here is that all of those things cost money. Never before have I seen a system so brutally effective at making characters part with their coin, something which D&D, for better or worse, is completely useless at.


The thing for me which is so striking about Torchbearer is how vastly different a game than D&D it is with the same basic design goals. For all the OSR crowd’s talk about difficulty, Torchbearer is a much better design example of a difficult game than D&D ever was or ever will be. Monster challenge affects the rate of resource burn but (for the most part) not optimal player decisions, so the only lever for difficulty in D&D is how difficult the rolls are. Introducing player skill into the game via puzzles or remembering to say the DM’s expected incantation of “I check the room for traps” are irrelevant to the game mechanics themselves, as well as antithetical to how I play games. In Torchbearer, however, you must engage with the game’s systems, and it presents a lot of decisions which require strategy. “Twist or Consequence” is one my players puzzled over a lot; stakes in Conflicts and choosing approaches to the Town phase are also fraught.

Unlike many games that we play, Torchbearer is intended to produce story from the stakes rather than the other way around. The players provide characters, the GM provides a dungeon, and the conceit is constrained, at least at first. The drama comes not from scenarios which frame the mechanics (like in PbtA, which is all about fictional positioning determining the necessity of moves), but from what happens when the mechanics have been engaged. One of the most interesting arcs in my short game was that of Winslow the dire wolf. The characters had just finished clearing the undead from a manor which belonged to a deceased town official…a circumstance which arose from rolling on the Entering Town table. After successfully clearing the manor the players decided to hunt for food in order to prepare for the journey back to town. This test was a failure, but the players still needed the food…so I gave a twist. The hunter was chased back to the manor by a dire wolf, bringing the characters into another resource-draining conflict. But one of the players had an idea…they had had a chance to recover from the undead, so how about the goal of this conflict be to capture the dire wolf? The conflict went well, so now the party had captured a dire wolf. Next step of course was to train it. I described the factors of the roll, so the one character in the party with the skill opted to go for the easiest possible roll. While he succeeded, this resulted in a poorly trained, barely loyal, and irritable dire wolf. Of course, the level of training of an animal is itself a factor in any future rolls to command the animal, so this was going to go poorly for the party. And go poorly it did. Winslow the dire wolf became my favorite twist, and by the time the characters made it back to town there was a warrant out for the murderous wolf.

What’s great about this example is that I prepped nothing. When the dire wolf did something, it was because a character failed a dire wolf-adjacent roll. The fact that the dire wolf was there at all was itself triggered by a roll. It drives home the point that the mechanics are intended to work simultaneously as a game mechanic to be engaged and as a fiction generator. And beyond that, even with a clear difficulty curve and clear failure states, the game is still built to ensure that failure moves the story forward rather than stop it.

Beyond the mechanics, Torchbearer has different meta-assumptions than D&D. D&D charts inexorably upward in terms of power and resources. Very early on in the game, gold ends up meaning almost nothing, a player has so much of it. And the core mechanic of the game is to consume larger and larger quantities of resources by piling on larger and larger threats. Torchbearer still charts upward in terms of power, though more slowly and to a more humble end. Resources, though, are where the game stays relatively flat. Only through luck and grit do characters manage to hold on to a little snatch of gold, and it takes a very long stretch of luck to get to anything resembling stability. This is a detail of the mechanics I really appreciate, by the way…up at higher skill factors are all the tasks you need to have your characters build a homestead or a stronghold, just like upper level D&D adventurers. All the rewards are there…but it’s going to be a really tough journey to access them. I appreciate that Torchbearer acknowledges the importance of the arc, even while also acknowledging how few of us played all the way through it, in Torchbearer or in D&D.

Player’s Perspective


Playing Torchbearer feels a bit like a good roll in mud: In the moment everything is sticky, wet, and slimy, but that’s kind of the point, right? The game achieves its purpose handily. Every action comes laden with risk, and every condition stacks making your life all the harder. The Grind is ceaseless. Sometimes I felt clever for packing my backpack perfectly, hauling just the right amount of light, and food, and treasure. Most of the time I felt like the game wanted me to find a safe place in the wilderness and just camp forever. I think my highest praise is that it made playing the party’s cook a mechanically rewarding role. No matter how tired, scared, and angry we got, I always knew how to scrounge up a bite to eat! Torchbearer will never be my go-to system, but sometimes it’s nice to get real crunchy.


For me, my experience with Torchbearer found it at odds with my current playstyle. Years ago, I was given a copy of Play Unsafe, a guide to playing and GMing that prioritized story over mechanics. More to the point, I had seen in a number of different games where players (myself included) worried so much about what the risks were, and if their choices were mechanically optimal that they wouldn’t take a cool opportunity when they had one, even if it would make their character better. I think that my first character in my current group wound up being far too timid and open-minded for someone who was supposed to be reckless and orthodox because he might have been killed. That’s not to say that those were bad things for a character to be, but I didn’t play out a good reason for him not to change, and that made him weaker in the story, especially when the other members of my party were continually building their narratives.

In Torchbearer, mechanics are as integral to game as the story, if not more so. They are so innately baked in to everything the players do, and the scarcity of goods and need to prepare for winter is so dire that the mechanics spur players without any story prods. I thought that I had been prepared for that. In the end, even knowing that things were supposed to be gritty and dire, that we were supposed to be just barely staving off death and that conditions were meant to be a downward spiral…I still found myself getting caught up in the whirlpool without and it felt like I couldn’t contribute anything to the party or story. Perhaps it was just a bit of bad luck, or I had not properly steeled myself for the type of universe I was entering. It was just a rough first entry. I would certainly go back to a similar game, but I think I would want to try a different mechanical set before then.

Torchbearer is a deliberately designed game; one reason it’s easier to run than Burning Wheel is that there are constraints on the type of game it can truly run. What’s interesting though is that it shoots across the bow of old-school D&D while having nearly no mechanical resemblance to it. It is both a testament to stretching mechanics (in this case Burning Wheel’s) as well as a direct opprobrium to using mechanics just because they’re there. Torchbearer is a better game than Basic D&D, and I don’t really think it’s a stretch to say that. It interrogates what makes a more difficult old-school style interesting, and comes up with a very different set of answers than D&D, informed by decades of game design. If you’re even for a second serious about your nostalgia for “3d6 line roll” D&D, for “ten foot poles for every character” D&D, for “oh shit I didn’t remember to tell you I checked for traps” D&D, you owe it to yourself to play or run Torchbearer. It’s a game designed around those experiences, without any compromises to wargaming that were embedded in D&D (and still are). If you want a hard game more than a story, Torchbearer is great. If you want drama to arise in an emergent fashion, Torchbearer is great. If you want to play an empowered character, or spend your time somewhere other than the wilderness or a dungeon…let’s just say Torchbearer wears its limitations on its sleeve.

Torchbearer is available at DriveThruRPG. Header is art from the Torchbearer rulebook, by Kurt Komoda.

3 thoughts on “Torchbearer In-Depth”

  1. Great article Aaron. I think Aki’s quote really lands for me. Torchbearer limited how I tend to play characters. (You’ve seen me in Genesys, I like stirring up trouble). Torchbearer made me feel like defying the “optimal playstyle” is a mistake, where other more story-focused games embrace the chaos of a sub-optimal character.
    I still like it. It made me appreciate game mechanics in a different way than I had before. It also felt like a necessary crash-course on the old school ethos that I never got the chance to experience. Long live Shortbread.

    Liked by 2 people

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.