A couple of weeks ago, Matt Colville published a new video about prophecies, both talking about the concept and soliciting feedback from the community as to whether anyone has ever successfully included prophecies and visions in their games. (For the record, he hasn't ever gotten it to work.)
I'm not in the habit of spraying my opinion across the internet, so I'm sharing it here.
I don't watch Colville's videos religiously, but I'm always a little surprised by them when I manage to watch them because I usually agree with about fifty percent of his content. (Let's say 40%-70%, probably depending on my mood.) He runs a very different game than I do, and so a lot of his discussion and advice isn't particularly applicable to my table. (He also doesn't seem to play or run terribly often, so I get the impression that his games are way more planned and plotted than mine. I'm embarrassingly running five-ish games right now on a baroque, rotating schedule, so I frequently have to be comfortable with a good answer now rather than a perfect answer derived from five hours' planning.)
Colville tends to see role-playing game scenarios as an alternate form of literature akin to short stories or novels: linear plots with plot points, themes, moods, etc. These things are all knobs that the author (in this case, the GM) sets and constantly tweaks in response to player action. Astute observers will recognize the core of '90s game design, with its metaplots and linear adventures, probably traceable back to Maliszewski's so-called "Hickman Revolution" in the early-to-mid-1980s.
My initial and anarchic foray into RPGs way back in the late 1990s may have started with White Wolf, but I quickly surmised that their Storytelling advice wasn't going to work for me. How can I establish a theme and mood for the game without knowing the energy the players are going to bring to the table on any given night? They have a say in how the game runs, too, and they're probably not going to explicitly tell me what they want.
In retrospect, it's hardly surprising that I fell in love with old-school play. Playing to find out what happens, the "story" of the game is the emergent story at the table, and all that.
Which brings me back around to the point: I think Colville's approach to prophecies and oracles failed because he was still thinking about role-playing game sessions like novels and not as their own genre of art, as I will argue again and again in this corner of digital real estate.
In a novel (for example), a prophecy usually serves two purposes: it acts as exposition and foreshadowing. If a prophecy is somewhat vague, as most of them are, then the audience might get the shape of it — enough to know that something big is coming and maybe even with some idea of what shape that thing is going to take — but readers won't know the outcome until later in the story. The more explicit the prophecy or the more deft the writer, the greater the likelihood that the audience will be "rewarded" by figuring out what the foreshadowing means ahead of time.
Contrast with role-playing games. Prophecies and visions serve a similar purpose, but the audience and the participants are usually the same people, so visions have a very different weight: they're clues. Clues don't have to be planned ahead of time — astute players in an investigative scenario are going to interrogate environmental details that the game master didn't explicitly plan, but can surmise based on what they know of the larger shape of things — but they do need to be deliberate and included with purpose. More importantly, the game is always about what the players do at the table. So, as with any clue, the game can't come to a screeching halt if the players aren't interested or don't understand the prophecy they receive.
What does all this mean? In my experience, prophecies, oracles, and visions all work pretty well, but the game master needs to be deliberate about their placement, and the game has to be able to continue running if the vision is ignored. I'll give you a handful of tips and examples:
- In the above video, Colville indicates that he included a vision almost as an afterthought: no preamble, no warning that the players were going to receive it, and most tellingly, he probably planned it only a session or two in advance. (Having watched some of his liveplay stuff, he seems to turn the plot on a dime as cool ideas occur to him. Protip: use your cool ideas, but make sure they're well-integrated with your existing game.) Even if a prophecy is about something minor, being able to tell the future is A Big Deal™, and should be both well-telegraphed and thoroughly considered in advance. How will this impact things? What happens if the players interfere? What happens if they ignore it? (Always assume your clues are going to get ignored. What happens next? Usually, ignoring a vision means that things escalate.)
- As long as we're talking about giving players oracular abilities, there is an alternate method to giving a player a vision. I did it sometimes in my Dungeon World game, and it is one of the recommended methods for some oracular powers in Unknown Armies: give the character an in-game bonus (a reroll or whatever), and then when they use it, ask the player what they saw in their vision earlier in the day. While it is a much more narrative way to solve the problem than giving the players a puzzle to solve every time you give them a vision, it gives them a little more agency when it comes to how they use their oracular abilities. Of course, as with trying to use oracles in the first place, that won't work for every game, either.
- In my Los Angeles-area Unknown Armies game, things were ramping up, and the player characters heard there was an Oracle in Las Vegas. They talk to the Oracle and receive the following poem. Unsurprisingly, the players didn't understand most of it, but it did reinforce that Something Big is happening, and that the player characters needed to interfere or else something bad is going to happen in Los Angeles. (This was further reinforced by the fact that many members of the occult underground left the city, evidently having divined that some sort of big trouble was on its way.) The core purpose of the prophecy was to let them know they were on a timer; any other clues they could derive from the prophecy were just icing on the cake.
- My Sunday night D&D game keeps encountering weird signs: a picture they can't perceive properly, a mysterious symbol they've found unobtrusively on at least one person in each town they've entered, various people asking them, "Are you forgetting something?" A merchant once claimed he met the characters before, but when they were insistent they had never encountered him before, he said he must be confused. (He's from the Underdark, so maybe he doesn't have a great memory for the faces of surfacers.) The cleric has received a couple of strange dreams featuring the player characters, his goddess, and a dwarven woman they've never met before. The players still don't know what to make of it, which is perfectly fine: it lets them know that something is happening in the background that they don't fully understand, and each weird clue or creepy vision just informs them that a clock is running down to an unknown revelation.
- My arctic Ravenloft game has two prophecies: a tarokka reading telling the characters where to find key elements to oppose Khan Yemur (a deliberate homage to Ravenloft's tarokka reading for treasure placement), and a prophecy. (Both the tarokka reading and prophecy may be found here.) They've been dutifully seeking elements from the tarokka reading, but they have a little less information on the prophecy itself. Again, that's perfectly fine: the prophecy is vague, and its main purpose is to let them know that they are important and that their destinies have been "claimed" by the Dark Powers of Ravenloft.
In every case, I fully recognize that the players aren't going to guess all the elements on the first try. Instead, these are sources of tension: each one lets the players know that their actions are important, that there's probably some sort of clock running in the background that they should be considering, and that there is some sort of confrontation coming that will relate to each prophecy. The more clues they uncover ahead of time, the more likely they are to have some warning for the coming event, but they don't need to perform any particular actions to enact the prophecy.
That's probably the key: don't treat prophecies as something you have to force the players to do, instead putting them there as additional sources of clues so the players can figure out what's about to happen. And if the players miss the prophecy date or misinterpret the vision and head in the opposite direction, so much the better! Whatever terrible thing they foresaw and failed to stop now gets to happen in their absence, and they can deal with it in the aftermath. Or not; maybe they decide to flee instead. If the continent is doomed, why is it always their responsibility?
Remember: the game is whatever the players decide to do at the table. Everything else is just fluff.