OSR Wisdom Distilled
This post is both to help me clarify and organize in my head what I think are the most useful pieces of advice I've gotten out of OSR spaces, but also as a useful help to anyone else reading this that's not already familiar with them, specifically aimed at DMs coming in from non-OSR spaces (read: 5e).1
1: Focus on the World, not on the Story
What I mean here is when prepping, instead of starting with a story, start with the actual foundations. What are the points of interest, the geography, settlements, towns, ruins, lairs and dungeons, what factions are present in the region, how much and what kind of power do they have, what resources do they possess, what are the key people to these factions (these automatically become your key NPCs). Of course, while doing this, you can have the story in the back of your mind, but focus on getting this legwork down first. What this means is that you are no longer chained to working linearly, laying out the story before the players. The world is set, the events are in motion, your players now only need to engage with it and you are ready for it to go in any direction. There is a big caveat, though. This type of play demands that the players act proactively and not just react to whatever is in front of them and get dragged by the DM where the plot needs them. Talk to your players to make this clear to them. 5E culture conditions players to play very reactively, placing a lot of burden on the DM as the entertainer and the one to push the game forward. To me at least, this is wrong and leads to worse games, not only because it's unfair towards the DM but also because you unnecessarily put all the work on one person.
Now, this might sound like a lot of work, which brings me to point 2 and 3.
2: Don't prep needlessly
Now, I did say to make detailed prep, yes. But this applies initially only to the immediate starting location and the broad strokes of the setting as a whole. The guiding principle is this: prep in detail only this, which your players realistically can reach during the session. A key part of this is asking your players every time at the end of the session where they are planning to head next and exactly what their plain is. That way you don't need to do guesswork, and don't need to waste your efforts. Naturally, with spare time, you can invest it in additional prep, but this should be your main focus.
3: Use random tables
I think people are wary of using randomness during prep and play, with random encounters seen as time-wasters that don't contribute meaningfully to the story and using tables during prep being seen as lazy. I'll address both.
Random encounters are great in a way that they create situations that not even the DM could predict. Each session becomes unique in that way, and they lift the burden of meticulously preparing scenarios and adjusting them to the party's strength. The party might be ambushed by bandits along the way to a location, or they might stumble on a dragon which completely changes how a session might go down, and with encounter distances, i.e. giving the party a chance to retreat, try to circle around or ambush an encounter, these can all exist without TPKs.
Something to be wary of is that small, weak encounters are fine in systems with quick combat like B/X (they create attrition), in 5E or PF2 you might need to take a more hand-crafted approach to the tables to not create time-wasters. Encounters need not be combat only. Traveling NPCs and events are all on the table.
Using random tables during prep has a huge strength, in my opinion. Generating results not only helps you "get going" and avoid creative blocks, but every result is in it's own way a creative prompt for you to elaborate on. In that way you not only have an easier time doing the actual legwork needed to make a detailed play setting, but are still using those creative muscles, creating connections and reading into implications. Another advantage is that random tables force you to work outside your "home turf", suggesting things that you wouldn't come on your own (I personally can get stuck in my own ways all too often). Even seemingly inane results, once you build up enough of them and start drawing connections, become complex and interesting and touched by your own unique voice.
4: Be an arbiter
The ideal form of a DM is one as the neutral arbiter. This ties into the prep-heavy approach and letting random encounters be random. The first principle is to never fudge rolls. Doing this, in my opinion, ruins the meaningfulness of the player's actions, making success feel undeserved and actions devoid of consequences. This extends to letting encounter results play out undisturbed. The party should almost always have the option to retreat, so this does not mean "TPK your part if they roll a poorly on the encounter table", but what it does mean is that events happen naturally and can go in any direction, in this way the game creates surprises to even you, the DM. Resist the urge to create Deus Ex Machinas to save a player from death, but also when an encounter might go easier than you hoped for. Let the players feel smart or powerful instead of adjusting the power to even things out.
Fudging can also mean having the result be the same no matter what choice the part makes, something people have dubbed the "Quantum Ogre". You shouldn't do this either. Think of it like a story video game, where all choices lead to the same result. What this means in effect is that the choices were meaningless.
A harsher version of this is as follows: you can change things before a session, but once it begins, it is beyond your control. As an example, before a campaign, you might opt for a system that lets players make a reflex save to avoid a trap's damage, and a save to resist it's effect, since you want to lower lethality. Once the game begins though, you may not intervene and change those mechanics or results mid-session. The world details and the rules are set before the session, and once the session starts, you are only the Arbiter.
5: Let encounters be deadly
As touched on before, it's okay for the world to feature enemies that are above the party's capabilities in straight combat, the same way it's okay for some to be weaker. Encounters should be what exists in the world, and how the party handles them, whether by avoiding them, putting them against each other, or finding a way to overcome something above their power level and reap the rewards is fully up to them. In this way the world feels like it has an existence of it's own, and isn't tailored to the party's strength like a video game. Some locations should remain out of reach of the power level of a party even if they are aware of it. Naturally you should still plan beforehand to make play and progression possible, placing weaker lairs and dungeons closer to the starting settlement and always giving the party options.
6: Ground your party in the world
Make your player characters tied to the starting region. These are the kind of questions you should ask your players during character creation: What is your character's backstory that relates to their chosen Character Class and their starting resources/money.
If they are a Fighter, how did they learn to fight? Are they a veteran of a war, a mercenary or enlisted? Or maybe they are a son of a noble house, why can they no longer rely on their family's fortune? Can they still use their good name among the commoners?
If they are a Cleric, which denomination? Are they on a mission? Many Clerics are nobility as well.
A Thief could have been an orphan raised in the slums, or a trained member of a criminal organization, or maybe they were pushed into crime. Did they steal their starting gold? A Magic User could be learned in a School, taken up by a wandering practitioner looking for someone to pass their art, or maybe a prodigy.2
Wilderness travel procedures, dungeon delving procedures, torches, lamp oil, ropes and food. These are all things that while small, ground your party in the world as well and make them feel like a part of it. Hardship and planning ends up making success feel more significant. These procedures should only come up when relevant, though.
7: Make less rolls
This one is pretty simple. If there is no time pressure, the character simply succeeds. Let players succeed in noticing things more often than not. Players having more information and being inquisitive is not a bad thing. If something is made difficult by time pressure, but the character has aptitude for it, they may roll. If there is time pressure and they lack the tools or skills, or the task is simply impossible (technically this also counts as the former), they fail.
8: Put focus on social advancement, not class advancement
Modern DnD puts a lot of focus on leveling up, to the point where you characters become essentially fantasy superheros. I personally don't find that as interesting as it doesn't really align with the fantasy fiction I enjoy, and I do think it ends up coming too fast, where the player character's amass enough power to completely ignore the setting and it's people and laws without meaningfully engaging with them. Dungeon Meshi, Earthsea, Lord of the Rings, characters in these stories, while capable, are in the end always at the mercy of the world even at the peak of their power. A low level adventurer PC is still cut above the regular populace. Staying at lower levels and encouraging player's to engage with the social and political side of the setting and world make for more interesting games, in my opinion. Quick leveling also makes diagetic advancement in other forms, like magic items, or as mentioned, social advancement like allies, faction favor, fame, followers, etc., less encouraged.
9: The case for dungeons and gold as XP
The advantage of a dungeon, which need not be an actual dungeon but simply a constrained area with distinct zones, is that it gives you as a DM as well as your players a more constrained, and therefore more manageable area to let the player's loose. This benefits you as there is less going on at once, but also your player's since it limits the decision space and let's them focus better. Now, the art of making dungeons is a whole other thing and unfortunately something that is very poorly understood in 5E spaces leading often to the perception that dungeons are inherently boring and a slog to get through. 5E combat taking longer does not benefit this. But they are a great tool in themselves. Dungeons can be powerful tools to flesh out your setting.
As for gold as xp, the benefit here is that the treasure the party manages to obtain is not only an interesting goal as opposed to killing monsters, but a good measure of their struggles and success. It provides opportunities for interesting situations, from stealing treasure from under a sleeping monsters nose, taking a gem from a trap, to figuring out how to carry a door made of adamantium out of a dungeon, as well as the regular logistic problem of carrying bags of gold coins. This isn't the only way to do it, though. Money often is a none-issue in 5E games, so players don't feel the need or a reason to get treasure. Give them things to use treasure on and they will seek it out. As for progression, milestone leveling isn't so bad so long as you do actually keep a tally of the party's hardships and achievements and not simply award them levels because a couple sessions have passed.
10: Let deaths happen
Death in RPGs is a scary thing. It should be, but it leads people to push it away so that it becomes a non-issue. That's fine for some games, I've played in one where there was an unspoken agreement that death was not on the table. But I think there is value in it being an ever-present threat. There is a balance to be struck so as to not make investing in a character feel like a waste, the type of play where one-hit-kill traps exist has it's own charm but isn't what one is usually looking for, but it is also up to the player to do actually put that investment in knowing there is a risk of death and loss. That's what makes it so captivating, the player's and their character's motivations and feelings align strongly in the fear of death, which make daring and heroic acts even more significant, and, in my opinion, character's personalities can really shine under pressure.
All books have their end, and in a way I think a character dying in a meaningful way is a better ending than simply surviving the whole of a campaign without much effort. Same can be said for choosing to retire and become an ally or patron to the party.
11: Don't overthink it
All that said, I think the best advice is to not get caught up too much in what should be done or what a good game looks like, and to just have fun and throw a dragon at the party once in a while and do what you think is cool.