Witch Finder is a game set in a dark version of Medieval Europe at the time of the witch hunts. Slightly darker, a bit more gritty and with real magic. Not glowing lights and controlled expression magic but witches selling their souls to the Devil in dark pacts kind of magic. Characters take the role of witch hunters, who are only debatably less dark than those they seek.

Synopsis aside, let's talk resolution mechanics.

Characters have dice from three sources: Fear, Force and Finesse - don't worry about where these come from. You'll need d6 of three different colours - one for each source. Typically there'll be between about 8 and 13 dice in total, divided between the three sources.

Resolution goes through four stages

Declaration

You declare what you want to achieve (not how you want to do it); the GM decides the difficulty number.

Rolling

You roll all the dice from the three pools and match up dice with the same number showing (regardless of colour). Then you find the highest number of matches (with the highest number showing if there's more than one of the same matches) and compare that to the difficulty set by the GM. If it's less than the target number, you have failed, if it's equal you have taken one step forward and one step back from the goal, and if it's higher you've succeeded.

Then there are six qualifiers; some positive, some negative. These are words such as 'style', 'division', 'injury', 'betrayal', 'follow-on', 'acquisition'. I'm thinking there will be four fixed words, and then two more (one positive, one negative) that depend on the character being played. Each qualifier is assigned to a number between one and six (so each number has a qualifier) - these will probably be fixed for ease of handling.

All matches other than the highest one are then assigned to their number (i.e. if you rolled three 4s they'd go to the qualifier assigned to four).

Discussion

The GM and the players discuss how to interpret the results into a final narration. This is where the qualifiers and colour of the dice come in. For each match see which colour of dice dominates (Fear, Force or Finesse or possible a combination thereof in the case of a tie) this determines how the outcome is reached. The qualifiers add complications and consequences to the result. Players are invited to put forward their own ideas and suggestions and negotiate with the GM over what happens so that a satisfying outcome can be achieved.

Narration

Finally, the GM narrates what exactly actually happens. They are encouraged to base this on consensus from the Discussion phase as much as possible.

An example

The Inquisitor Torveda (played by Ant) finds his path blocked by an angry crowd; Ant declares he wishes to get past them and via some mechanical means it is determined he will roll 5 fear dice, 3 force and 2 finesse. The GM decides this is difficulty 3.

Ant rolls and gets four 4s (three fear and one finesse), two 5s (both fear) and two 2s (both force), a 6 and a 1. The 6 and the 1 are discarded, the four 4s go to determine that he succeeds through fear, two 5s go next to 'acquisition' and the two 2s next to 'consequence'.

A suggests that Torveda uses his terrifying presence to simply cow the crowd into submission, the GM likes that idea and riffs off it: "yeah, but a lot of them don't like being treated like that, so in consequence they'll try to undermine you".
A thinks for a moment "Ok, but as they're fleeing I see one of them drop something and pick it up".
GM "Stealing from the peasant folk, huh, a bit of a stoop for you? How about you command some of them to follow you and gain some followers?"
A: "Alright"

The GM then narrates how the scene happens using the ideas described above and fleshes out the outcomes.

Comments? Suggestions? Similar games? Requests for clarification? All are welcomed.

Interesting. Sort of like Otherkind (thank you Vincent Baker, you gaming god) only set in the Inquisitorial era. I like how dice results get assigned to character based "qualifiers". This resembles Otherkind, but is unique in that the results are based entirely on the individual character's qualifiers.

Actually, I'm already doing my own spin on the Inquisitor concept (see my thread: http://www.indie-rpgs.com/forum/index.php?topic=16456.0 ).

Heh. You find the witches. I'll burn them.

Kirk

Okay, quite frankly, my first response is, "Whoa, that's a whole lot of handling time."  But then when I read Godlike, which uses similar dice, I thought that was insane, and apparently it plays relatively smoothly.  Have you tested this dice mechanic with real people around a table yet, Jack, to make sure it's not overwrought?

Secondly, the player/GM dynamic seems rather skewed to me.  The players and GM 'negotiate' but in the end, the GM decides what really happened, and is only encouraged to base it off of the discussion.  With so much information falling out of the dice, is it possible to assign some of that articulation of what the dice mean to the players rather than the GM?  Perhaps the 'lesser' matches can facilitate player input that cannot be vetoed by the GM but cannot veto the GM's success/fail articulation?

Lastly, how is it determined where the lesser matches go to in terms of the qualifiers?  Do I read correctly that (for instance) 'style' is attached to 2, so if you had a pair of twos in there, you'd get a little stylish color in your results?

Reading over the game again, the “traditional” (D&D) player-GM power imbalance is quite present here (I don’t know if this is just how you wrote the example or how the game is structured). I’d make it that the player gets to narrate the results.

Lets see if I understand the mechanic:

You roll a number of dice from Fear, Surprise and Ruthless Efficiency (I mean Fear, Force and Finesse), each with their own colour.
Certain numbers are attached to certain “qualifiers”
If the highest number of a number rolled is higher than the difficulty, you succeed.
The results of the rolls are assigned to their qualifiers (If you roll two threes, and acquisition has all threes, then Acquisition is equal to two)
According to the dice and where they came from, and augmented by the qualifiers, you narrate the results.

Read Otherkind at www.lumpley.com (I mentioned it before). It has a similar mechanic to this, but is a bit more streamlined. This seems a bit convoluted. Not bad, just a lot of factors that you have to take into account. What I’d do is I’d say to only use the qualifier that has the highest occurrence of numbers. Chances are that there will be dice from more than one trait in each qualifier, so that cuts down the amount of things you have to deal with in your narration.

Say you roll 4 threes. Threes are all to be assigned to Acquisition. One of the die is Fear, two are from Force, one is from Finesse. Acquisition has the highest concentration of dice (all the others have two or one), so you achieve your goal (we will assume that you succeed) by the use of Force with a little bit of Fear and Finesse, augmented by Acquisition.

What do you think? Or have I completely missed the point... (it happens. A lot)

Kirk

Kirk, I've not played Otherkind but I have played Pretender which is derived from it and uses a similar system. As you can see, my concept is heavily inspired by ideas from Pretender.

Joshua, I've not tested it out around a table yet but I'm not too concerned about handling time. While it does take longer than most traditional systems it does that because it is giving more result. I also intend reolution to be large scale; you're not going to be rolling more than once or twice in an entire scene very often. Also, matching up dice is a very quick operation, so like Pretender the big part of it is fitting the given elements into an outcome; I don't have a problem with the creative side taking time.

Lastly, how is it determined where the lesser matches go to in terms of the qualifiers?  Do I read correctly that (for instance) 'style' is attached to 2, so if you had a pair of twos in there, you'd get a little stylish color in your results?

Hi Jack,

Looks interesting, I like the idea of player created qualifiers, although I'm unsure of having both positive and negative qualifiers, maybe it just seems unsymetrical.

Would the player have to take thier highest number of successes as the primary 'success' count? I could see a situation whereby I'd want to have less success but avoid a ton of dice in 'consequence'.

If you had qualifiers that were a mix of good and bad, maybe allow the players and GM to take turns assigning the qualifiers? So with 4 successes, and 2 pairs, the player might pick first and set the first pair in 'aquisition' and the GM going second to put the second pair into 'betrayal'.

I've got something vaguely similar I'll show you on Wednesday.

Maybe I don't understand the mechanics correctly, but I have the impression that having more dice does not grant you a greater chance of success. I haven't done a complete analysis by enumerating all possibilities, but I would argue as follows:

a) The chance that the largest number of matching dice will show a 1 are the same as for all other numbers from 2 to 6. This is just a symmetry argument, i.e. all dice are equally likely to show a 1 than any other number.

b) You only choose the better number in case that the maximum match size is not unique, i.e. two or more rolls appear equally often.

c) The more dice you roll, the more likely it is that the maximum match size is unique. This is simply a consequence of having more and more possible outcomes.

So for a large number of dice, your probability to get 1, 2, .., 6 is approximately 1/6 each, same as for rolling a single die. To get the best chance of success, you want to roll the minimum of 8 dice.

Of course, this is only relevant if the size of your die pool is supposed to affect the chance of success.

Hi Nsruf (do you have a real name we could call you by?),

I can see how you read the mechanics as that, but that's not how it's meant. The success/failure is read off the number of dice in the match, not the number matched (so three 6s is a score of 3, not 6 and rolling 1112233445 is a one-step-forward, one-step-back success against a difficulty of 3).

Jack wrote:
Hi Nsruf (do you have a real name we could call you by?),

I can see how you read the mechanics as that, but that's not how it's meant. The success/failure is read off the number of dice in the match, not the number matched (so three 6s is a score of 3, not 6 and rolling 1112233445 is a one-step-forward, one-step-back success against a difficulty of 3).

Jack, in terms of similar games -- or at least similar mechanics -- my game, Verge, uses pools of 6-siders and counting of the mode like yours does. Verge uses opposed rolls instead of your game's roll vs. static target number.  My rules also get a lot of mileage out of rerolls. Essentially, you set aside the most common result (the mode, more or less) and check off traits to reroll the rest of the dice. The rerolls can improve your result if the rerolled dice come up with the same face value as the dice you set aside.

Hi Jack,

Did I somehow miss how equal numbers of finesse/force/fear dice in a match are handled? Because given the relatively small dice pool, you will probably run into that quite often. Also, you should want to do some serious calculating of probabilities. I'm not really very good at that, but I think that getting a match of 4 with 10d6 is already quite tough.

One more problem I see: You have a good chance that a player might keep hitting "one step forward one step back" results, which can get tiring, especially if you get more and more qualifiers.

- Frank

Hi Frank,

It's not very well explained above but the idea is that where there are ties between two elements both should be brought into the narration while if all are involved, none of them should dominate. You're right about the odds, the odds curve is rather peculiar for matches (which is why in Chanter, I used an opposed dice roll mechanic). However, I intend for the TNs to be low, even difficulty 3 would represent an unusually difficult challenge so the odds should work out OK, although I haven't actually calculated them for the higher dice numbers.

I take your point about 1-forward, 1-back results.

Hi All,

After some more playing and consideration I'm thinking of limiting the number of qualifiers to two, one to be picked by the player, one by the GM (with instructions to be nasty about it) - where there's only one qualifier, the GM picks on success and the player on failure on 1 forward-1 back. This should limit the number of elements to be involved and add more player choice into the process.

Hey Jack,

I'm a big fan of GOG and Chanter, so I'm interested in this game, too.

As a mechanic-in-a-vacuum, this seems servicable to me. Without a gameplay premise and reward system, though, I can't judge whether it's "good" or not, in terms of achieving your goals for the game. So far, all we know is that any conflict will be resolved through a mix of Fear, Force, and Finesse. As far as I can tell, the distribution of those three elements is very randomized, with no adjustment by players or GM. Also, resolution doesn't involve decision-making at the system level. The player announces intent and then rolls and sorts the dice the same way every time. For a system with (apparently) significant handling time, this means the player might have a long wait between decision points during the game.

I think target numbers are going to be tricky. 4 is very hard and 2 is an auto-success, since you always have six or more dice. Which leaves only 3 as a number you might actually use. Is a single target number (for almost all conflicts) okay for what you want to do?