The Randomness Question

I watched a YouTube video recently where the commentator, if that's the right word, claimed that the fun of D&D is in randomness, and not knowing whether your actions succeed. I want to state straight up that this is false, but I'm not just going to spend a whole post arguing that - I know the YouTuber in question doesn't think that the fun of D&D entirely boils down to the fun of blackjack or coin flipping. The claim was a hasty generalisation to prove that one of the key engagements of D&D is uncertainty. But it made me think hard about whether that's actually really true. What sort of uncertainty? Where does it have to happen?

(The video is linked here so you can make up your own mind if I'm speaking sense or not. Be warned, this fellow's style isn't for everyone.)

🕷

I've been running Into the Odd recently - I gathered together a bunch of my roleplaying friends, friends with whom I discovered D&D and roleplaying in its modern incarnation, and made an open table. The idea of using ItO was that learning the rules wouldn't be a problem - especially good since we have some brand new players. ItO should provide a nice counterexample to "the fun of [RPGs] is not knowing..." because it basically consists of a save mechanic and a damage mechanic. That is to say, it doesn't have any sort of skill check. If you're attempting something, anything, in a non-pressurized situation, and it's something you don't need special training or tools for, you can take your time etc., you just succeed. You only roll the dice to fight, and when you need to make a save, i.e. when something bad will befall should you fail, and you can't take all the time you need to get it right.

Climbing a wall? Sure, you succeed. Jumping a pit? Got your rope? Got some time to spare? Success. Trying to seduce the Queen? Now that could land you in jail, so maybe make some sort of roll. But seducing the barkeep in a low-stakes situation? Go right ahead. And no uncertainty anywhere.

This was a shock to me when I first read the ItO rules, and was specifically told that they contained saves, not checks. How could I ever inject uncertainty into the game? But if you think about it for a moment, it comes with some pretty hefty benefits. In fact, it basically just codifies a rule that GMs should follow anyway: When there aren't any stakes, let your players succeed.

Oh dear, two rules coined in only two posts. Lets support it a little then. If your players are rolling to try to pick the lock on a door with no time pressure, do you make them roll for it? Of course not. Everyone knows this by now; the players are just going to keep trying and trying until they succeed. There are no stakes because there's no possibility of not getting through the door.

The players are trying to find a clue in a crime scene. Do you make them roll for it? I've seen this one hotly contested. Again, the answer is probably no. If, as in most crime scene investigation situations, you have as much time as you need to pore over the scene, the players can just keep searching until they find something, or conclude there's nothing else to find.

But what if that happens? What if the players conclude, wrongly, that there are no other clues left, and leave before finding an important one? Surely that's an interesting bump on their inevitable road to success, and one worth randomising for?

Well, first, that's unlikely to come up as a result of rolling dice. Players know if they roll well or badly, so will keep trying until they get a good result. So it's most likely going to be pure player error. But the real point is that missing clues in a mystery scenario doesn't count as part of the fun. You don't have fun missing out on something you never knew was there. In fact, it stands in the way of the fun of trying to determine the significance of the clues, the real meat of the mystery scenario. All you're doing when you make players roll to find clues is gatekeeping the fun with a coin flip.

Same goes, in a very tangible way, for the door example. A door that needs a specific key is fun; you get to play a key hunt, then you get to fight/loot/save whatever's on the other side of the door. But a door you need a 15+ to open isn't fun; at best, you coast through the roll as if it was never there, and get to the looting &c. But at worst, you get neither the key hunt nor the looting. And the GM, the person in charge of presenting a game world for you to have fun in, chose to do this to you.

Why would they do that? Potentially out of some misguided thought that randomising a result is inherently fun, or part of the inherent fun of an RPG. I hope I've put that notion to bed, at least in this sense. But there is a place for random results and unpredictability in RPGs, it just needs to be implemented in the right way, in the right place.

Where is the right place? Well, everywhere else: We've said you shouldn't be rolling unless there are stakes. Well, roll when there are stakes! As the particular YouTuber I have in mind correctly says, the game would be boring if the characters succeeded at every action they attempted. But the solution to this isn't to make every trivial action a crapshoot. There is a certain sort of emergent storytelling that arises from having to adapt the plot around a failure. But notice the terms used there: "storytelling" and "plot". When there aren't stakes, you're not addressing either of those things.

There's no interest in the rogue tripping over his own feet trying climb a ledge, or the druid spooking the animal she's trying to talk to because of a botched roll. In situations like that, the players are just trying to express what makes their characters cool, and all you do with a roll is, again, jeopardise that unnecessarily. You can't build on those failures to advance the plot or present interesting complications - they are at best an inconsequential speedbump. And once you start rolling for all these little things, you take away the players' abilities to create long-term plans - they can't construct interesting plays or tactics because they don't know whether they can tie their shoes without tying their feet together, let alone execute a multi-stage scheme.

What about the cases above, like the lock that welds itself shut, or the clues that blow away after five minutes? The characters can miss those, and risk failing their mission because of it. Ergo, those must be high stakes situations! Well, yes, but if so that's bad game design. Again, all you're doing in these examples is locking off game content behind a test of luck. The only one who knows what that content is is you, so the characters won't feel the stakes.

The key thing about rolling dice is that neither the GM nor the players knows the outcome beforehand - the uncertainty is shared by everyone. But there's another sort of uncertainty that only the players can experience. That's the sort of uncertainty you have when you unveil a hand in poker, not knowing if the other guy's is stronger. He knows as soon as you declare, but you take the leap of faith; Here I stand, you say, Whether the cards are right or not, I've made my choice.

That's the sort of uncertainty the players feel when they try a solution to the puzzle, when they strike out in a direction when rations are low, when they break down the door in answer to the screams beyond - in short, in any high-stakes situation where they don't have crucial information that you do have. They don't know, even though you do, whether the chest is trapped, or the screams on the other side of the door were faked. That's the uncertainty that keeps players on the edge of their seats, not not knowing whether your lockpicking skills are up to the task, when everyone can see that fact is determined by random chance. You're not just testing the players' luck, you're testing the players. It's a form of uncertainty that respects, even hinges on the players' choices; other forms of uncertainty only undermine them.

I think we as GMs focus on the uncertainty of die rolls as the archetypal uncertainty that makes RPGs entertaining because we are overly GM-focussed. That's what our experience of uncertainty is at the table, so that must be the archetypal experience of uncertainty in RPGs. But we need to step out of our own shoes and think on the other side for a minute. What we view as delectable unpredictability may just feel capricious to those with a personal stake in it. And things that to us are a known quantity create and enable mystery, exploration and tantalizing unknowns to the other five sixths of the table.

🕷

Okay, I've rambled on for too long. And I didn't want to do another post shouting about what I think are bad GMing decisions. But I hope someone can get something out of this, because I've seen experienced GMs fall routinely into this trap of over-rolling - I think the only reason the YouTuber I mentioned doesn't think this is a problem is because he's good enough and experienced enough as a GM that it doesn't naturally come up at his table. But in the unfortunate game that it does affect, it undermines player agency and planning, and undercuts characters' particular, unique coolness. Don't voluntarily do that to your players.