I am curious what are other people are doing for hidden tests. In most cases, where the outright difficulty may be hidden, I typically roll the challenge dice, but give the player the misfortune modifiers that stem from environmental/internal factors.

In cases where the act is quite simple, but the specific thing to obtain is difficult (i.e. searching a room for a hidden trapdoor), I typically set it at Simple (0D), layer in challenge appropriate to environment, and then determine success by the outright number of successes. As a rule of thumb, the number of necessary successes is equal to twice the challenge. So if someone hidden in a room is Hard (3D), instead of announcing that to the player (and clueing them into the presence of something difficult to find), I see whether they roll 6 successes.

Is that ad hoc difficulty about right, do you think?

Not quite. An average Challenge die will give you 0.75 failures (it has 2 sides with 1 Fail, and 2 sides with 2 fails). So, on average 3D should be 2.25 failures.

For hidden rolls, or rolls where they shouldn't know for certain if they've succeeded, I'd suggest the player rolls Characteristic, Stance and Expertise dice, and you roll any Fortune, Misfortune or Challenge dice.

That way, they will have *some* idea of the person they're interrogating has spilled the beans, but they won't know for sure if they got the truth, the whole truth or nothing but the truth.

I don't do it often but if needed I would simply roll the challenge dice hidden or preroll them (e.g., know that in this case player whatever you roll is compared against 2 failures and a bane).

HedgeWizard said:

i like that alot!

for me it always depends on the check. if i want to see if they are ambushed and recognize it early enough i would roll all dice hidden, so they might know that something is going on but they won't know what.

in case of a trap or a hidden door i let them roll the characteristic/stance/fortune/expertise dice and i roll the misfourtne/challenge dice. to give the player attempting the check a real chance i sometimes set the difficulty one level lower if their appropriate skill is not so well trained. but i only do that for a group that has no player who is better trained than the attempting.
so, if a dwarf slayer would make a nature lore check so locate an animal to hunt, and there is a hunter in the party, the difficulty will stay the same!

I do agree that when something is opposed (say interrogation) then the GM rolls the challenge dice behind the screen.

What I like about counting the successes as opposed to an outright difficulty/vs model is:

1. it isn't hard to do some of these things outright. Sweeping a room for instance isn't inherently difficult. It's a question of how well you do it.
2. You can have scaling reveals - you get 3 successes you find x. You get 6 successes you find x + y, etc.

The problem can be outlined as so:

If there are two things in the room to find on a search, a Hard (3D) door and an Easy (1D) piece of incriminating evidence, what do you roll against? How do you handle that roll?

In my suggested model, I'm thinking the test would be East (1D) (it's very simple to search a room). The PC is rushing and tossing things so they convert 1 dice to reckless, and gain 1 misfortune for being too haphazard. They have two PCs helping = +2 Fortune. They need 2 successes to find the piece of evidence, and 6 successes to find the door. But I agree the success threshold is a bit off, but also seems to be close to right.