In response to my comment on his post about TTRPGs lacking a real sense of exploration, noisms mentioned that doing narrative well is an underexplored skill on OSR blogs. At first, I was surprised by that opinion, but after letting it tumble around in my mind, it started making sense. Thus, I thought this might be a good opportunity to do something to help address that.
Before going any further, I'll also say that Justin Alexander and Scott Rehm/Angry GM have several posts about narration, story format, pacing, and similar topics on their blogs that are worth reading. I don't share their tastes in rules, but both have some degree of formal education in those matters (and Justin also has professional experience from his theater work) that can be helpful for any game. I just have the practical experience from having published a couple of novels, which is a questionable credential given the ease of self-publishing ebooks, though I take heart in the general positive feedback they've gotten. But I digress.
In order to limit the scope, I'll concentrate on narration of exploration, but the principles I talk about can be adapted to combat, social scenes, or anything else. I'm also going to focus on the moment-to-moment presentation at the table, leaving aside wider-scale topics like pacing of progression, strategic releases of dramatic tension, fitting fictional narrative flow within the confines of real time constraints, and so forth.
This should be obvious, but the first thing to do is decide whether exploring fantastic environments is even something that you (and your players) care about. Wilderness exploration is something that doesn't have a natural ending point (reaching impassible terrain like an ocean, a bottomless ravine, the edge of the world, or a vast empty reach of space are only ends insofar as the PCs may lack a means of continuing further, as the fact that each example went a step beyond its predecessor should indicate). It's a journey with no destination, there's no natural limit to how much time can be spent on it (metagame considerations like needing to end the actual play session are a separate concern), and there's not necessarily going to be any palpable reward for doing it.
It's not without reason why some GMs just pass it by with a statement that "you travel for <insert length of time> and arrive at <place>" or perhaps treat it with a series of skill checks that glosses over the minutia. If you just see exploration as something in the way of the actual adventure, don't feel obligated to include it in your games just because there are travel montages in many fantasy stories. The Fellowship of the Ring doesn't detail every day between leaving the Shire and arriving in Rivendell. This is fine; just be honest about it. Don't present your game as being a sandbox about exploring the vast lands around the Triboar Trail when it's actually about following a core plot line with the occasional stub of a branch, and you won't engender bad feelings about failing to deliver on expectations.
One step removed from that would be something like a wilderness pointcrawl (e.g. Slumbering Ursine Dunes) where there's a chance of encountering something unexpected along the paths (note that this need not be limited to pointcrawls; the hex map usage described in Beyond the Wall Further Afield is also along these lines). Exploration remains in the game in a fairly vestigial form that gets glossed over except when it poses a complication, and the main difference from the step above is some freedom to set out towards a destination that the players don't know. Once again, there's nothing wrong with this. Much of what I'm going to talk about below will still be useful here, but the actual presentation of traveling should be a supplementary touch, not a focal point, so when I get to talking about random encounter checks, bear in mind that you might have different issues of time scale to deal with.
Then there are the modes of play that embrace exploration fully, worrying not about the destination (or perhaps not even having a destination, in the most basic form of hexcrawling as a game structure) as long as there's discovery and enjoyment during the journey. This isn't without some drawbacks. There will be gormless fucking around. There will be time wasted in pointless endeavors. There will be nonsense that overtakes planned narratives. There will be tracking of resources and encumbrance (whether through simple accounting or more abstract means). If done well, though, there will also be wonders of taking in breathtaking (if imagined) vistas, invigorating terrors of dealing with the unknown, and a sense of pushing your horizons beyond their previous limits.
As nice as that might sound, I'm going to level with you: it's going to take work. To the best of my knowledge, there aren't any general game mechanics that can ease the GM's burden on this front, either (at best, you can fall back on some random generation tables as in OSRIC or Stars Without Number, but those work best when planned and shaped ahead of time). And while it might sound nice when stated as at the end of the previous paragraph, keep in mind that it has to be fun in actual play for it to be worthwhile, which means accepting the faults (aimlessness, time sinks, Plot Tumors) and tedium (tracking resources, tracking encumbrance). If you're willing to put in the work, it can be rewarding. If you're not, again, just be honest about that instead of making token half-assed efforts that are only going to end up as disappointments.
If you've decided that you do want to feature exploration, take some time to think about why. What's cool about it for you? Was it the oppressive weight of silence from being alone in a desert salt flat, with a seemingly endless expanse of white stones devoid of any signs of life? Was it the humbling isolation of moving through a cavern where your light illuminates a mere fraction of the space you're traversing? The majesty and wonder of looking out over a broken landscape as dawn brought warmth and color to it? The eerie otherness of a similar landscape being subsumed by the night? The grandeur, solidity, and agelessness of a forest that feels older than human existence? The simple joy of finding a hidden glade, with cute wildlife drinking from a brook as you inhale a heady natural fragrance? The amusement of getting distracted by a lizard detaching its tail before escaping? Be specific, concentrate on the moments that give you the strongest emotional responses, and take the time to write it down, because this will be your bed of inspiration for bringing it to life later on.
While you're doing that (or once you've finished, depending on how you prefer to work sequentially), fill in secondary details that brought the scenes to life in your mind: the little tufts of scrub clinging to life in the salt flat, the echoes providing a sense of scale in the cavern, the ravens roused with the rising sun, the sinister highlights struck by the moonlight, the sounds and smells and (mostly ambient) tastes of the woodland, the twitch of the animal's head towards your presence before it resumed drinking, the comedy of falling for a trick which you'd only read about previously. Once again, be specific, focus on the key bits that seem to trigger the most emotion, and get it onto a page/screen instead of relying on your memory.
In both of those steps, don't try to write well. Whether your natural expression is a blob of unsorted words and phrases, snatches of sentence fragments, bullet points, or any other form doesn't matter. The key is to get the pieces that make you want exploration into a record with as little filtering as possible, because once you start taking the details apart and begin to analyze them, they're probably going to start losing their magic. To copy a phrase from Bruce Lee: don't think, feel.
Once you've done that, put it aside and out of your mind, whether that means working on something else, taking a break to eat, sleeping on it overnight, or whatever else. Come back to it, read over what you wrote, and see if it brings some emotional reaction. If so, great. If not, take a moment to picture why in the world you thought exploration could be fun, and write down more based on that. Keep doing this until it starts to have an effect (go a step further and throw random bits of it at other people if you can, as a somewhat independent test).
I hope this won't be discouraging, but the real work starts now. You need to figure out how to take that jumble of words, which probably still relies in part on feeding off of stuff that's just in your mind, and communicate it in a way that catches interest instead of causing confusion or boredom. I can't give advice for that specific act, but I can give some broad guidelines that I tend to work with.
In general, when it comes to narrative description, you have a fairly small window of time during which people are willing to listen, so you have to try making the most of it. For example, if the PCs walked through a portal and emerged in a salt flat, I wouldn't have time to convey everything about that scene in glorious detail right off the bat (I'd think this would be obvious from, you know, actual experiences talking to people, but the copious boxed text that still pervades adventure modules seems to imply a significant number of GMs think it's acceptable to give a monologue to introduce every scene). If my key bits of that moment were an endless stretch of dirty white ground, complete silence, tufts of scrubby grasses clinging to life, a salty smell, and a soft breeze of hot air, that's too much to dump on the players all at once. Maybe I could get in four of those details (e.g. "once your eyes adjust to the blaring brightness, you see a silent expanse of dirty white rocks stretching in all directions, and a soft breeze carries a hint of salt") because I was able to combine some of them into single points, but my typical guideline is to limit myself to two or three details in between giving chances for the players to speak.
Don't feel like you have to give an explicit prompt for that, by the way. There's nothing wrong with doing that while you're learning how to do this, but it should come as part of the natural flow of conversation. Take a pause in your speaking, make eye contact, and get a read on the table. If they look like they're absorbing what you said and thinking about what to do, give them a moment to do that instead of piling on more details. If they start asking about certain details, embellish as appropriate, maybe adding on another detail if you didn't have much to say about their actual question. Only if they're caught up in the moment and hungry for more should you continue to elaborate, and stay on the conservative side with that.
How do you know what to say up front and what to hold for later? The only guideline I can give is to follow the natural progression of how you'd process that scene if you were there. Start with the two or three things that you'd notice immediately, and then flesh them out more as attention/timing allows. The only firm exception I'd make is clarifying if it sounds like the players are imagining something very different from what you were trying to communicate, and even that can be waived if what they're imagining doesn't break your sense of the scene (e.g. if I'd introduced the scene as "you're in a desert salt flat" and they start talking about it as if it was a huge space full of table salt grains, I might clarify that it's mostly solid rock with salt-stained white surfaces, but I might also just roll with it being a desert of white salt "sand").
This doesn't mean that all those fine details you might've written up before were wasted! Far from it. As the PCs continue to explore the place, you should be mentioning little bits and pieces of flavor to help give the players a more complete feeling of their surroundings. Make sure to also use a variety of senses; visual and auditory descriptions are probably going to be the bulk of it, but include the others. Smell has a strong link to memories, so at least drop a word on if a place smells dry, wet, moist, musty, acrid, spicy, stuffy, sweet, repugnant, etc. Touch can be described in terms of direct contact (e.g. slick ropes of hanging moss brushing faces as they move through the trees), indirect contact (e.g. the slimy bed of rotting vegetation under their boots provides unsteady footing), or just imagined contact (e.g. the wrinkled bark is stained with wet residues). Taste may not come up too often, but be sure to give it some attention if the PCs forage or swim, and there are enough links between tastes and smells (like the aforementioned moist, acrid, spicy, and sweet) that using the right words can cover both at once.
At the risk of sounding like a broken record, bear in mind that you shouldn't be overloading the players with this stuff. Think of it as something similar to describing empty rooms in a dungeon. Sprinkling one or two flourishes into every exchange will do you far more good than seizing an opportunity to take a verbal dump on their ears, and going overboard will tend to invite mockery or disconnection. Your goal in the general descriptions is to give just enough to maintain the mood of exploring a foreign environment.
Also, in preparation of continuing to a later point in my train of thought, try to include some foreshadowing of whatever else might be living out there.
Two things that can get in the way of that are exhaustive repetition and jarring moments of game mechanics, which have an unfortunate habit of coming together when it comes to random encounter checks. I have three pieces of advice on this: interpret the results of the check through your understanding of the context, make the check before providing context to the players, and give effort to narrating the result of the check regardless of whether it succeeds or fails.
In order to give a description concise enough to be digestible, you have to know what you're describing. This isn't too hard when you're working with a time scale close to what's passing in the real world (as in a dungeon) because results of a random encounter check tend to be sudden and immediate, but wilderness exploration is typically handling on a coarse time scale, with the basic unit being somewhere between several hours (such as Justin Alexander's 4-hour watches) and a day (as described in the wilderness section of Cook's Expert rules), inclusive. Similarly, distances in a dungeon tend to be much shorter and more immediate than in a wilderness, for reasons that shouldn't need elaboration. Short of being attacked by a small number of ambush predators, there will tend to be more grey area in between detecting signs of other creatures and actually encountering them in the wilderness.
Thus, when an encounter check turns up running into 250 bandits, it shouldn't be a matter of running right into them after rounding a hill. They're likely to have some sort of encampment (whether mobile or permanent), they're likely to be doing stuff (preparing food, doing chores, organizing activities), and they're likely to be detected at some distance (through tracks, trails, noise, sight, maybe even odors from upwind). If you've been including some touches of foreshadowing (as suggested above), build more off of that as a launching point. Also, make use of varied reactions (whether from reaction tables or contextual understanding) so that encounters don't feel like a mad-lib where you're just swapping around nouns and numbers.
Now, remember that all of this has to be balanced by the general limits of how much time you have to give descriptions. If you used your moment just talking about their traveling, you're going to be in trouble when you then have to take that time again to describe the encounter that just happened. Making the checks before describing anything to the players lets you take that into consideration and adjust your presentation accordingly. There's a world of difference between giving three sentences to your players as their prompt to act and giving six sentences; the former is more likely to capture the spark of the moment while the latter will tend to come off as an impediment to their play.
In my experience, most game systems have a fairly low chance of triggering a random encounter during any given time, usually no higher than 17% (1-in-6). In a dungeon, this isn't a issue because there should be plenty of details to check out in empty rooms or in dealing with keyed content to pass the time until an encounter happens. In the wilderness, though, when you're summarizing large chunks of time and space, you'll lose that moment-to-moment fine interactivity. This is where your record of inspirations can help fill in the gaps. If you break up your random encounter check into bands (e.g. for a d6, 1 = random encounter, 2-3 = evidence of random encounters, 4-5 = uneventful day, 6 = special moment), you can use those ideas to showcase something that made you think exploring would be cool.
That's when you can launch into a detailed scene. Bring your passion and energy! Display your excitement! Keep it brief enough to avoid having the players tune out, and keep a little in reserve to flesh out details as they get drawn in, but don't be afraid to show why you want them to explore and to hopefully give them a memory to define that experience.
Having said that, I wouldn't recommend having every special moment be something grandiose. I'd suggest having a table of natural events that includes both cool moments that exist purely to capture the wonder of exploration and moments to represent natural complications of being in the wilderness. Also, you may recall that some of the example cool moments that I gave were linked with time (specifically dawn and sunset). That's probably not the best practice if you're working with more than one or maybe two random encounter checks per day, so use your judgment on that front.
As for the days where nothing happens ("uneventful day" per my example list above), don't just say "nothing happens". Again, treat it like an empty room in a dungeon; give it a couple touches of flavor that the PCs can interact with for as long as they want to, but it shouldn't be expected to lead anywhere.
Of course, it bears mentioning that the wildernesses explored in TTRPGs needn't be limited to what's actually out in the real world. If there's fantasy involved (whether magical fantasy, science fantasy, or some combination), there will be elements that aren't real. Give those some thought, both in terms of their consequences (e.g. how does a dragon mark its territory, and how do other creatures interact with that?) and in terms of their implied symbolism (as I talked about towards the end of my Using Vampires post, e.g. how does the dragon's symbolic greed/sin warp their surroundings?). If you want to use a piece of Carcosa or the Wilderlands of High Fantasy, think about how that'd impact the peoples therein and nearby. If your giants are expressions of superiority complexes and systemic disparities, what would their impact on their surroundings be? What if they were instead expressions of colonialism and the divine right of kings? Or a mirror to reflect humanity's verminous aspects? Or a representation of Lovecraftian principles?
There are no right or wrong answers for those, but your answers can (some might argue "should") impact what the wilderness is like. If you want to go even further, would your inspirational ideas actually make sense in such a world, and would their impact be altered by it? Perhaps you don't need to think too much about this depending on the depth of play that you're aiming for, but assuming you've got the time to spare for it, it wouldn't hurt to give some thought to such things.
Those last couple of points are more about polishing content and developing thematic coherence, and I'd see them as stretch goals rather than foundational. Thinking about why you want to have exploration, bringing details of the environment to your narration in digestible pieces, and having a few special quality moments to let the players experience the high points of exploration are probably good starting points for improving the feel of exploration in actual play. It might seem like a lot, but so are many broad topics in running TTRPGs. With practice and experience, you'll likely start to find what works well for you (and your players, if you've got a fixed group), and there's just as much potential to save anything that goes unused to be recycled/repackaged in the future as there is for other parts of running a game.