One of the most visible differences between DnD (5e) and OSR-style games is lethality: in DnD, player character death is rare. This is perhaps surprising, considering the amount of extreme violence the average adventuring party gets up to, but what bothers me isn’t necessarily the lack of realism, it’s the lack of death as a storytelling tool. I understand that this is partially my own taste: I don’t like happily-ever-afters, and some of my favourite moments from movies and TV are character deaths (the end of season 1 of Game of Thrones! Either of the heartbreaking deaths in Fellowship of the Ring! (the best Lord of the Rings movie)). However, personal taste aside, I do think that lots of stories are improved by the inclusion of tragedy, and I do think a lot of stories told through DnD are lesser because the system avoids this outcome so much.

Pictured: left, a DnD player having a good time. Right: a DnD player also having a good time.

A TTRPG has a difficult line to walk, because although the high-level goal is to tell a compelling story, the moment-to-moment gameplay usually involves players pursuing goals that are much more in line with their characters’ goals, and often those two goals come in conflict. When a player is one hour deep into a four-hour combat session, all that player (and the DM, usually) wants is for the PC to win the fight; what both people should remember, however, is that failure can be just as interesting and compelling an outcome as success, and that success is only satisfying if failure was really an option. It’s important, then, to avoid the temptation to go easy, or fudge rolls, because it will make the game less fun in the long run.

This tension between goals will be present in any TTRPG, but (to my taste) DnD’s balance leans too heavily in favour of giving the player what their character wants, to the detriment of what the story wants, especially when it comes to character death. Part of this is because players expect a certain structure to a DnD campaign: it’s usually zero-to-hero, told over 10+ sessions, over the course of months or years. The system is designed to facilitate that kind of play, which is fine! It can be very rewarding to follow a character through a long arc of development. However, the negative effects are clear: in the small scale, the rarity of death takes a lot of the feeling of danger out of combat; at best, I find that DnD combat feels merely challenging, and at worst, it can feel like a grind. In the larger scale, it can rob the story as a whole of stakes, and of the kind of memorable moments that stick with you because they were upsetting at the time. My favourite season of Dimension 20 (by far) was A Crown Of Candy, the only season I have watched which allowed early player-character perma-death. Dimension 20 always had excellent world-building, and compelling characters; this season differed from the others in that I found myself getting more engrossed in the story as it proceeded, because the stakes felt like they were getting genuinely higher. It was also the only season where I found combat just as enjoyable to watch as the out-of-combat episodes.

Since DMing myself, I’ve realised the biggest part of the DnD system that I dislike is actually how it can sometimes feel video-gamey. In particular, it can feel like a mass-market AAA video game, with a big focus on numerical stat increases and loot as rewards. Video games that lean on that kind of mechanic to improve player retention are now often derided as “Skinner boxes”, but I feel like the same negative aspects of these kinds of rewards can show up in TTRPGs as well.

Again, I want to be clear that this kind of thing can be a lot of fun. Your character getting a new sword is cool, and sometimes the easiest way for a DM to reward a win is giving out some loot or bumping a stat.

However, I can’t shake the sense that it’s a little cheap, as a mechanic. Whenever I have noticed myself doing something in a video game that feels like grinding for a stat improvement, I make a conscious effort to re-evaluate if I’m actually having fun, or if I’m just a little addicted.

Both the rarity of character death and the primacy of mechanical stat-based progression seem to me to be designed to make it easier to run the “ideal” DnD campaign. For example: players want to tell long stories? Make it harder for player characters to die. Players want to see character progression and development? Well, one easy way to simulate that is through mechanical progression, so let’s facilitate that as much as possible.

In this way, the OSR as a contrast to modern DnD reminds me a little of the way that Dark Souls games are often seen as a reaction to modern AAA bloat. By returning to the conventions of older, cruftier games, by removing the affordances that were supposed to enhance narrative, we’ve rediscovered a leaner, meaner system that ends up being more narratively satisfying. (I realise I have now basically written “Mothership is the Dark Souls of RPGs”, but this is my personal blog, so I’m allowed to be trite.)

One of the big tenets of the OSR is the idea that dice rolls should be used sparingly. There are two aspects to this, as far as I can tell. First, it’s important that the GM only calls for a roll when all outcomes of that roll are interesting, and push the story forward.

Take, for example, the following situation: it’s the beginning of a campaign, and the players are investigating the office of a missing merchant (presumed kidnapped). Hidden in a secret drawer of the merchant’s desk is an enigmatic note full of tantalising leads, plot-thickening revelations, and hints of a conspiracy that goes right to the heart of the government!

Except that the players all fail their investigation check, so they don’t find the secret drawer. Now they have no leads, and nothing to do, so the plot grinds to a halt, while the DM scrambles to try and figure out how to give them the note anyway. Oops. Now, the OSR wasn’t the first community to recognise this kind of failure state: in DnD, “don’t hide plot hooks behind dice rolls” is a common mantra. What happened in this situation is that the DM didn’t actually want to roll the dice to see if the players could find the drawer or not: the DM just wanted the players to find the drawer, and they called for a roll because that’s just what you do when investigating.

The OSR, though, takes this tip and generalises it. If dice are being rolled, all outcomes had better push the story forward. For an especially deft GM, this can mean designing scenarios such that situations like the above don’t arise; for me, it means that when situations like the above do arise, I just don’t ask for a roll.

The “failed investigation check grinds the story to a halt” is such a common pitfall that most DMs seem to know to avoid it, but the generalised problem still shows up, and I think that the game has evolved a much worse solution than the OSR’s approach. In DnD, there are lots of ways for a DM to make it so that a roll basically can’t fail. Think about group history checks, where everyone around the table has (say) a 50/50 chance of passing the roll, so a table of 6 people has a 98.5% chance of passing at least once, or the rogue for whom lock-picking has become a mild inconvenience. These ways to put a thumb on the scales evolved because sometimes the DM just wants the party to know something or get through some door, and dice rolling is so integral to the game that it seems more natural to try and modify that mechanic rather than removing it entirely. The result, however, is that dice rolls lose their weight.

Dimension 20 has a gimmick where, if a roll is especially important (maybe a death save, or a pivotal counterspell), the DM will bring out the “Box of Doom”, which is a dice tray with a skull on it. It’s a great prop, and it works well to heighten tension. An extreme version of the OSR tenet of “only roll if it matters” might be “only roll for Box of Doom rolls”.

The second way that rolls are used sparingly in the OSR is by trying to take the gameplay away from the dice as much as possible. The classic example of this is social rolls, which are not present in Mothership: if one character wants to persuade another, instead of rolling a relevant stat, the player will roleplay persuasion; then, the GM will judge if the character was persuasive or not, and rule based on that. I can remember one moment in particular in an actual-play I was watching when a player role-played an extremely cool and intimidating speech to threaten an antagonist; then, the DM asked the player to roll for intimidation, the player got a 4, so the dice said the speech didn’t work. It felt like a real bummer in the moment, and I could see it making players around the table want to avoid doing fun roleplay in the future, and instead doing the dreaded “Oh there’s a guard? Yeah my character tries to scare them. I’m going to roll intimidation.”

Social rolls aren’t the only thing that can be turned into strategic gameplay. I like to include lots of little puzzles in my games; where a character might roll against their intelligence to see if they can hack a computer, I think it’s fun to instead present a wordle-like minigame that represents them trying to guess the password. You can even blend the dice rolling/character stats and the minigames: maybe a successful roll grants the player another attempt at the puzzle, or a failure sets off the alarm system. In general, the principle I try to follow is that I want to encourage players to be creative and strategic as much as possible, and then the dice should only come into play when some randomness makes things more fun.

I don’t want to sound dogmatic about this style of play, by the way: there are definitely times when it’s more fun and better for the flow of play to just roll rather than whipping out another convoluted puzzle or roleplaying some interaction that isn’t actually that important. However, on balance, I prefer to avoid most social rolls, all information rolls (history, insight, etc.: I just give the players the information I think they should have), and some intelligence rolls.

One of the parts of TTRPGs that I have actually enjoyed the most is preparing for sessions. I really like making handouts and writing NPC backstories, even if they never come up in the game itself.

For example, for this game I decided to add a bit of Cthulhu flavour to the scenario: I made one of the guards a cultist, and I added in a lot about the “cursed language” of Cthulhu (R’Lyehian). I made it so several of the doors in the facility were locked by R’Lyehian codewords, but there was a R’Lyehian-to-English dictionary in the library. The catch was that players couldn’t keep the dictionary between loops, so they’d have to run to the library, read it quickly and try and find the relevant words, and then remember them to use them on the next loop. As an added twist, I had secret roles for each player, one of which instructed them that speaking R’Lyehian was incredibly dangerous, and they were to be suspicious of anyone who did it.

Here’s all the material I made for the game:

• R’Lyehian-to-English dictionary
• Secret Roles
• Scenario Notes
• Locations
• Facility Reference
• (Handout) Todo List
• (Handout) Diary

One last change that I made was that I had the timer start at 25 minutes, and I reduced it by 3–5 minutes every loop. I think this improved the pacing a little: it meant the first loop could be a little slower and more exploratory, and as the players got more familiar with the environment, and quicker at navigating the challenges, the later loops would speed up to match. It also put a limit on the overall number of loops: by the time the timer got below 10 minutes, the players knew that they only had 2–3 loops to solve the mystery and escape, or they’d be stuck looping one second in the elevator for all of eternity.