Harr, this is excellent stuff. I really like your multi-dimensional "deterministic" template.
I'll share some thoughts on how I approach skill checks.
For a skill test, basically you need to decide, as a DM, one a few things.
First, the "goal". Each of your examples has multiple possible results, and thus the skill test is about how to reach the various end states.
This is the Most Important Step. If you don't have a clear idea of
why you are calling for a skill check, don't call for one. Make the action automatic (either auto success or auto fail).
In order for a skill check to be appropriate, two things must be true:
1. There must be a reasonable chance for failure and for success. Don't bother rolling if you need to roll a 21 to succeed or a 0 to fail.
2. There must be a reasonable cost for failure, or a cost for just trying. Otherwise, the characters will keep trying until they succeed. 3e had this formalized in the "Take 20" rule, which in my opinion was a phenomenally boring rule. A really easy way to add a cost for failure is to disallow retries. "You already tried your best to (pick the lock | climb the wall | decipher the runes), and you couldn't figure it out. Try something else."
Example: Climbing over a 50 ft. stone wall should require a check, since there is a cost for failure: you might fall and take damage. Climbing over a 10 ft. stone wall shouldn't require a check, because there's no cost for failure: if you fail, you can just try again. The exception is if you are under some sort of time pressure, in which case, the time you wasted failing is its own cost. (This is most relevant for skill checks in combat: spending an action is already a cost.)
Regarding point #2 and Take 20, certain skills, like Open Locks and Search, were never rolled in my 3e games. If you wanted to pick a lock, you went ahead and spent the 2 minutes to Take 20 and either got it or didn't. There was no decision to be made.
In my 4e game I added a house rule called "Taking Your Time." Basically, you spend 10x the normal time for a skill check, and get a +5 bonus. But if you fail, each retry costs a cumulative 10x time. So the first check takes a minute, the second check takes 10 minutes, the third takes 100 minutes, etc.
I also have a house rule: if you're outside of combat, you can't retry a skill check unless you Take Your Time. So if you fail to pick the lock or climb the wall or find the secret door or whatever, you now have to decide: do I want to spend another minute trying? Probably. But if that fails, do I want to spend another 10 minutes trying? Now the party has to decide whether they have time for that or whether they need to find some other way to accomplish their goal.
I also have a skill test template to contribute:
8: Buffet Style - The characters are trying to get multiple benefits, bonuses, or beneficial outcomes (or cancel multiple penalties). For each success, they can choose one benefit. This is similar to the Quantifying template, except that the benefits are discreet rather than continuous.
Example: The group is traveling through a forest, which limits their visual range considerably, slows them to 1/2 speed, and has a % chance of wandering monsters. The players can make Nature checks to navigate the woods. For each success, they can choose one of the following:
- Increase speed by 1/4 (up to 1x)
- Increase visual range (from poor to moderate to good)
- Half chance of wandering monsters (or double chance, if the group is looking for trouble)
This is how I do overland travel in my game. I actually reduce the checks even further by using a single Nature check by the best member of the party. Success allows him to choose a benefit, and every 5 points over the DC allows him to choose another benefit.
This "every 5 points above/below the DC" is a good way to get richer results with fewer actual checks. It's especially good for routine stuff like Quantifying checks when you go shopping.
It's also good when, for story reasons, it makes sense to have a single character make the check. The other members of the party make checks, but don't generate any successes. Instead, succeeding on a check grants the main character a +2 on his check. When the main character finally makes the check, for every 5 points over the DC, it counts as an extra success.
Finally, on role-play vs. roll-play, one way to do it is to treat the roll as
optional. If you can convince the Duke using just your words, great! Sometimes you say the right thing and it makes sense for the NPC to react that way and no roll is needed. But if you want to make a Bluff or Diplomacy or Intimidate check, you can, and it can sway the Duke's opinion in your favor.
This was actually formalized in the SW Saga Edition's rules for Persuasion. In that game, the NPC's actions -- his inclination to do what you want him to do -- are based upon what you are asking, and on his attitude towards you. For example, if he is Unfriendly but you are offering him a million credits or offering not to have the Empire arrest him, he'll do it! No roll needed. But if you don't have a million credits or the ability to call in Imperial forces, you have two options. You can use Deception to trick him into thinking you do have those things, or you can use Persuasion to raise his attitude towards Friendly. The interesting thing about the Persuasion option was that the DC was
not based on what you were discussing, since the relevance of your request and offer are already accounted for by the GM when deciding whether or not a person with that attitude would agree. If the Persuasion worked, his attitude would increase, making him more likely to help you for less reward/more cost.
-- 77IM
I don't know, I guess I could go into details of why we've wound up doing it like that (stuff like focusing only on the characters' victories or something), but really it's just what's most fun for us.
It already gives me ideas for some cool encounters (which is another benefit of having mini-templates that I've found: They can stimulate you to get creative with game/encounter ideas).
