Against Torches

You ever tell someone you don’t like [food], and they tell you “no you just haven’t had GOOD [food]”? So you go to the restaurant they tell you and order the thing they tell you and, surprise surprise, you still don’t like it? 

That’s me with tracking torches. For years I’ve considered it a waste of time. And whenever I tell anyone this they say “torches are good, you just have to make tracking light matter!”

I came to this conclusion after running seven sessions of His Majesty The Worm. HMTW makes tracking light matter, and it still sucks. All it does is fill the players’ inventory spaces and hands. It adds an extra step to every interaction where I ask the players who’s holding the light source and then they get to continue doing whatever they’re doing. Technically you’re supposed to impose numeric penalties on players who stray too far away from the light, and have the guy holding the light narrate getting closer to help with any fine detail work, but I very quickly stopped doing it because it was such a gigantic pain in the ass. And that's with the overloaded encounter table (Meatgrinder in the game's argot) removing the need to track how many actions each individual torch lasts by hand.

 

In a way it’s freeing. I no longer have to feel like I’m missing out. I played a game that took torches seriously and gave them the best possible showing, and now I can confidently say it’s not for me. But what are torches supposed to do, and how can I replicate those functions so that getting rid of them doesn't accidentally get rid of something I care about?

Filling Inventory Space/Opportunity Cost

Every torch is an inventory slot you can't stuff something else in. At least one person has their hand occupied carrying the light source. 

This is important in His Majesty The Worm where player inventories are completely full most of the time. To be frank I don't like the inventory management aspect of that game. Every time the players find an item it immediately triggers a round of hemming and hawing over who should pick it up and what they should discard to make it fit. But I also understand that the tedious inventory management bullshit supports the "slice of life" element of the game that the players genuinely like.

If you get rid of torches you should reduce the size of the player inventory commensurately. Eyeball how much space/weight the players usually spend on torches and slash it by that amount.

Atmosphere

Dungeons are dark, so we should mechanically simulate the dungeon being dark by making the players carry a light. Dungeon crawling at low levels is supposed to be a horror game. The players should be afraid to run out of light and be subject to whatever "lost in the dungeon" table you're using.

I think this is a fundamental misunderstanding of how atmosphere in RPGs works. I think you’re never going to create the real life feeling of being in a dark confined space by monitoring the depletion of an imaginary resource.

One aspect that we can copy from visual media (films, video games...) to create atmosphere is the use of music. When I run a game in a space I control I pick out backing tracks for each dungeon areas, and sometimes additional pieces of music for keyed monsters. I can't do this in a public place, but I've moved away from running in game stores, cafes and breweries because I don't like shouting over roaring drunks, screaming toddlers and howling dogs for four hours at a time.

Light Spells/Special Vision Types

d20 fantasy is full of player special abilities that interact with the lighting system. Light, Daylight, Infravision, Darkvision... Circumventing the lighting system is an important form of mechanical differentiation for different character classes, fantasy creatures and monsters. The players are intruders in the underworld and can only operate in a tiny bubble of light, whereas the underworld monsters can move and act freely in the dark.

If you don't make the players track torches then you should just eliminate the Light spell. The players shouldn't be able to pick or randomly roll an option that's just useless. More complex variants like Daylight or Continual Light can stick around since they serve functions besides illumination (such as triggering monsters' sunlight weakness). Typically your playable fantasy races have some other trait besides darkvision, IE the OSE Dwarf getting saving throws and construction tricks or the Worm Dwarf getting a bonus camp action.
 

But my italicized caricature of a torch defender has a point about the asymmetry between monsters and players. I really like the feeling of the dungeon as a hostile alien ecology where the adventurers start at a disadvantage. Many iconic monsters can see in the dark while the players can't (Drow and Duergar), or have special senses that circumvent the need for light (Umber Hulks and Bulettes), or even create darkness (Dark Creeper). But I don't think any of them need torch tracking to stalk and hunt and move freely while the players stumble around. The Drow and Duergar set vicious traps in their tunnel networks and use spells to spy and stealth up on adventurers. The Hulk and Bulette can burst through walls. In the post 1e books the Dark Creeper creates magical darkness which holding a light source doesn't defend against. 

 

(Though I do like the version of Dark Creepers I used in His Majesty The Worm, based on their OSE Advanced Fantasy version, who target a light source carried by the players and then take advantage of their temporary blindness to steal items. Monsters targeting light sources isn't something I see often but it's the most interesting and fun thing you can do with torches and you do lose it if you stop tracking light). 

Expedition/Personnel Management

Light management, if tracked strictly, is a strong motivator for the players' initial foray into hiring NPCs. If holding a torch is such a chore, hire a linkboy to do it. Hire NPCs and now you're plugged into the loyalty and morale rules, wage rates, provisioning expeditions once the group grows large enough to need a supply train, you've got a stack of backup characters... Wage rates in Basic derived games are typically low enough that you can hire a pile of henchmen with a fraction of the wealth necessary to reach Level 2, allowing you to initiate the expedition management phase early on.

My experience is that players bring in hired guns because they're outnumbered and they want to even the score vs the monsters they're facing, because they recognize a weakness in their lineup of abilities, or because they encounter a non-hostile monster and recruit it for fun. They don't need torches and lighting to prompt acquisition of henchmen. I've never run a full blown hex adventure game where the players had a mule train with teamsters and logistics to support the expedition. I have run a game where they hired armies of mercenaries.

Time Pressure

The central tenet of the dungeon crawling genre is that spending time in the dungeon is costly and dangerous. Giving actions a cost in time, measured in the depletion of a resource, forces the players to balance caution with decisive play that maximizes the use of that resource.

On paper my preferred method of applying forward pressure is wandering monsters. They make your dungeon feel alive without exhaustive tracking of keyed monsters as they patrol around the map. Combined with encounter distance, reaction rolls, monster motivations and morale a good wandering monster system can create a broad array of friendly, violent or tense encounters. 

The big problem with wandering monsters is pacing, and this is something torches do better. Even with a breezy battle system like in OSE you're looking at a substantial table time commitment, especially if you use the number appearing given in the book. You really feel this when a fight breaks out toward the end of a session. And in His Majesty The Worm? Forget about it. The combat rules are far too chunky and involved to organically slip combat into the flow of play. 

Whether in Esoteric Enterprises, OSE, HMTW or my own personal heartbreaker Begone FOE I usually run dungeon crawls with the explicit instruction that the players must end the session in a safe area, typically exiting the dungeon by the end of play (three or four hours). It's up to the players to budget their time, turn around and head for the surface with enough time on the clock to deal with any random encounters that pop up (I add stoppage time if there's a delay which is my fault, like having to take an important phone call midgame). This is actually the primary form of time pressure I use, more than anything in the rules or game world. I know I'm not the only one to identify table time as the real limiting factor of dungeon exploration because last year's indie darling Shadowdark did something similar, tying torch depletion to real time elapsed rather than a clock in the game world.

Using session length to apply time pressure works most of the time but there are some problems. After seven sessions of open table play I can say confidently it does not work for His Majesty The Worm because the game's resource economy assumes multi-session dungeon exploration with rare trips to the surface city Regardless of system, chaining everything to the real world session clock means the players' affect toward monsters is primarily determined by factors outside the game world. Not "can we take these guys" or "have they done anything to deserve our anger" or "do they have treasure", but "do we have enough time left in the session for this?" In the Maintenance and Preservation campaign this was fine because the players were maintenance workers with a job to do. In a game where faction play and treasure acquisition are supposed to be paramount it can run counter to my design goals. 

There are a couple ways I can address this. The first is to stop running open table games and just do everything with a closed group of consistent players. If I can be reasonably sure I'll have the same cohort every week (and that's a huge if) then I don't need to ensure we end the game at a good stopping place every time. The other change is to build my dungeons with "safe" areas where players can end the session without navigating all the way back to the entrance.

(Even outside of dungeon crawling games, I wonder if I've got the arc of play backwards. For a long time I've warned people that anything done at the end of an RPG session risks being cut for time or compressed into irrelevance. The traditional arc of an RPG scenario is that the most exciting thing happens at the end. This can be a boss fight, something that encourages pregenerated characters to turn on one another, a shocking revelation that sends everyone reeling... But putting the most exciting thing in your scenario at the very end risks a group of fatigued, cranky players disengaging at the exact moment you need them to be on the ball. Especially if there's a mechanically dense combat).

Backwards Compatibility

Taken together, the above factors make a strong case for continuing to track lighting in order to keep the game compatible with the past fifty years of dungeon crawling content. The old ways were far from perfect and mostly done on an ad-hoc basis rather than according to any coherent philosophy of design, but every change you make to the formula erodes your ability to pick up and play anything written using that formula. At the end of the day people care more about having cool modules to play than having a perfectly designed game system. 

This is the toughest nut to crack. Moreso than the pacing issue, which is important to me personally but strongly tied to factors outside the game rules. I'm in Revision 15 of my personal fantasy heartbreaker Begone FOE, a game that exists because I want to capture the parts I like about retrofantasy dungeon crawling while fixing my problems with other games. But my game is still full of stuff I don't like, such as tracking torches and rolling a d20 vs AC and whiffing over and over, because changing it would make my system less useful to other people who want to use monsters and dungeons and magic items from the last fifty years of D&D. My Dwarf still has darkvision, defining the race primarily in opposition to a light subsystem I don't like or care about. I swapped from tracking torch time to an overloaded encounter die like the one His Majesty The Worm uses, but if I think tracking torches is a waste of time why include it at all? For all my mean spirited strawmen of torch defenders I'm out here doing the same thing.

Maybe I need to worry less about compatibility and stick to my guns. There are already a million retroclones out there for people who want 1:1 compatibility with the current crop of system-neutral-but-really-B/X modules. I don't care about marketing niches or commercial viability but I care a lot about creating something which has its own identity. And I don't think it's that hard to convert things from one system to another. I suspect that a lot of dungeons were and still are written without much attention paid to lighting at all, and won't break if you run them that way. Though that still leaves the Dwarf and Orc in need of new special abilities...

(Parts of this post is taken from my ongoing FATAL and Friends review of His Majesty The Worm)