Tuesday, November 5, 2019

The Secret Roll

I know there are a lot of bloggers and blog readers who favor rolling all the dice in the open. The original West Marches campaign, which I'm not faithful to 100%, also was for open rolls by the DM.

Now, I've come to believe that in combat, yes, the rolls should be open. Fair combat rolls, observed by everyone, lead to fewer complaints when things go pear-shaped.

But sometimes, rolling in secret as a DM for non-combat tasks is a good thing.

Searching for secret doors is a trade-off. There's not guaranteed to be a secret door where you're searching. And even if there is, you're not guaranteed to find it due to the roll. And each search takes a Turn, so the more searching done, the more chances of wandering monster encounters that suck up resources. In this case, if the roll is in the open and a result proves that there is no secret door (1-2 on a d6 for an Elf, 1 on d6 for anyone else, with no door found), the party knows to stop expending resources. But if the result is a mystery, they don't know if there is no door, or if the dice just weren't on their side (and chances are they weren't).

And now, they have to make a choice. Risk a wandering monster check to roll again? Or move on and potentially miss some treasure or a shortcut through the dungeon.

Now, I can understand the rationalization in the above situation that a successful roll where there is no door means the party gets definitive evidence that there is no door. So rolling in the open isn't so bad for that. But the suspense and measuring of odds of keeping that roll secret is more interesting to me.

Similarly, Thief skills are rolls that I, having learned from Mentzer's rules where he advises such, think the DM should roll in secret. Again, it adds to the suspense at the game table. And it's a situation where, as DM, if you were going to fudge the roll anyway, you might as well just tell the player straight up that conditions are such that they succeed automatically.

I mean, no one complains when a DM tells the Thief player, "Sorry, there just aren't any shadows to hide in here." Or if a door is barred rather than locked, so it can't be picked (although a clever Thief can work around a barred door too...). If the situation is such that failure is guaranteed, I don't see many players complaining. So if success is guaranteed, the DM should just tell the player that without bothering to make a roll. 

Just like players, the DM shouldn't have to roll unless the outcome is uncertain. And while certain rolls like monsters' attacks, damage, and saving throws most definitely should be rolled in the open, occasionally there are still times when it is better for the game experience for the DM to keep the roll secret from the players.

IMO, YMMV, all that jazz.


  1. In 4E and 5E D&D, I definitely roll passive Perception checks and monsters' Stealth checks secretly. I often do not even announce anything, otherwise players will know something is lurking nearby.

  2. Why not, when you put a secret door in a dungeon you're going to run, you roll the dice right then to learn if they discover the door? Surely it can't make a difference when you roll the dice. And if it's found, you can leave the door on your map and simply announce to the party that yes, there's a door, without needing to roll a die.

    And meanwhile, you can simply remove any secret doors from your dungeon map the players won't find. And if one of these is necessary to reach important places in the dungeon, why don't you just change the door from a secret door to one that's not?

    1. Why not, when you hide treasure in a dungeon you're going to run, you roll the dice right then to learn if they find it or not? And if they can find it, why not skip the search for loose flagstones or concealed caches behind walls and just place the treasure out in the open, so they find it automatically?

      And meanwhile, you can simply remove any treasure they won't find.

    2. Exactly. Demonstrating how silly these rolls really are, as minimal obstructions that don't matter in the larger scheme of things. We should concentrate on larger objects and purposes for die rolls, instead of lowering ourselves to hiding doors and treasure when really we want the players to go through and be enriched by their experience.

    3. It does make a difference when you roll the dice. And how many times players choose to roll, risking having more random encounters.

      IF there were only one roll allowed, and there were no consequence for choosing to take the time to perform the activity, then sure, why not, roll everything in advance. But really, rolling everything in advance is the same as not rolling at all (which is probably your point).

      I see value in playing D&D as a game of imperfect/incomplete information.

    4. With respect, you claim the value, but you don't define it. "A value" is a subjective, emotional argument that may stem from nostalgia or superstition as much as it might from real purpose or effect. Therefore I ask, not to insult, but to stimulate thought: demonstrate that value as something that does not arise from mere habit of playing the game that way.

    5. Alexis, I don't take this as an insult, so don't worry. I understand what you're saying, but I'll need to dedicate a whole post, or maybe a series of them, to answering as to the value of keeping some rolls secret.

      But I will say, "players being enriched by their experience" is equally subjective and arbitrary. And yes, you've spent years trying to explain what you mean by that on your blog. But any sort of value judgment in the end comes down to an arbitrary emotional response.

      I'll do my best to explain the reasons for my arbitrary emotional response to secret die rolls, and I fully expect to fail to convince you of their value. But it should be an interesting discussion nonetheless.

    6. "Oh, by the way, Marge, I was being sarcastic."

      I do not believe in eliminating rolls for secret doors or for finding treasure. Or eliminating doors or treasure that we know in advance the players will not make. Aside from the assumption that there's only one chance to make a roll (where do the rules say that?), it eliminates player agency. You are assuming that players must search, or that, failing a search, they can't find some other solution, like casting divination spells, coming back with an elf, or just asking someone who might know where the secret passage or treasure is hidden.

      Rolling in advance and eliminating all secrets winds up making adventures more linear. I don't want linear adventures. I want players to try different tactics and decide how much effort they want to put into a given action.

    7. With all due respect, an adventure game that isn't open and reactive to all possible player action can be bought on sale on GOG for about $10.

      To explain:
      You put a secret door in a dungeon. The party may or they may not search it. They may or may not return, with a device that can detect it by magic. They may or may not die earlier in the dungeon, with a later part entering behind them, this time bringing a skill-set for detection not possessed by the first party. This is not even considering possibilities like a monster fleeing to the secret room to escape the party, possibly leaving a trail behind it; a character accidentally using passwall to stumble into it; or - and this issue is the more relevant - simply designing the secret door such that the method to find it, rather than simply checking a dice, is based on actually describing the search process, as one might do in a role-playing game with the aim of immersion.

      There can be no meaningful choices made in a game, by the players in that game, if the DM is arbitrarily cutting off choices before the game even begins. Perhaps something is missed in play? That happens. The most that you, as the person drafting the map, has lost out on, is the minute or two of time it took to put together the room; an act which, in theory, should be enjoyable in and of itself. On the other hand, your players can engage in a game that responds to their actions.

      Incomplete or imperfect information is the basis behind creating a responsive world, rather than a static one. This is of value. Hidden rolls compliment this by ensuring player and character share in a lack of information. This, too, is of value. The alternative is functional, but sacrifices immersion at the table for... what gain? That everyone at the table knows you didn't cheat? Self-control and trust with your players accomplish this just as well.

      If you want a proper numerical definition to the value, I'd also be interested in seeing player groups run through campaigns and giving off 20-point Iron Chef assessments on taste, presentation, and enrichment; but the only people with the funding to run that study are Wizards, and they seem more concerned with selling products than indulging our curiosity.

  3. I've written a response on my blog (spoiler: I agreed with you, then took it a step farther.)

    1. Yes, I read it. I just haven't had time to respond.

  4. Catching up on this now.
    Two thoughts:
    1 In Nethack, for characters with search skill (ie rangers or ring of searching) the roll is automatic as they pass a secret trap or door. Everyone else has to press the S button and spend a turn in which monsters may come, but incurs no encounter roll per se. In the game's procedurally generated dungeons, this fits perfectly, as I've had unskilled characters starve trying to find a door to proceed.
    2 In the Gumshoe system, Investigative Skills result in auto success for those with a high enough skill. This too makes sense in the game's focus on advancing the story by furnishing clues.
    Both make sense in their context.