Poll: What is your preferred skill check mechanism?

[Desolate Dancer]

So I have been thinking about this a lot lately, since we have directly opposing (and a somewhat in-between) views about skill checks in rpgs as per the following:

1. There's always a marginal chance for success and failure regardless of skill level/character level.
Which means that you can always succeed or fail occasionally, regardless of your skill level and character level. This can be great for beginners where sometimes they can have a lucky shot at an attempt, but can also be annoying at higher levels where you fumble at something (e.g. spring a trap) which can have a devastating effect. A good example for this is d&d of course, where your roll a d20 and a 1 is a guaranteed failure whereas a 20 is a guaranteed success, no matter what. As such, the chances are fixed: 5% for both eventuality.

2. Chance of failure decreases (success increases) by skill level, but both outcome is ever-present.
Although you can still fail and succeed occasionally, regardless of your skill level or character level, at least you can be sure that upgrading your skill level will lead you to less fumbles and more easy successes. This is an in-between solution between the above and the next one.

3. Sometimes success or failure is guaranteed, depending on task difficulty relative to skill level.
This is something, that is very attractive to some because of the safety it provides: the higher your skill level, the more tasks you'll ace. A simple lock is no trouble for a thief with a high open locks skill, and they will always manage to unlock those mechanism that are way under their current skill level. This means that your guaranteed minimum roll/score increases, the higher your skill level is. It also means however, that certain locks will be out of your reach if they require a higher roll/score than your maximum possible score.

To sum it up, both extremes have their own advantages. The 1. option can make 'lucky shots' even if you are way below in level and talent to succeed at a task. The 3. is good to make sure that you'll not waste time on fumbles and needless irl actions e.g. in a computer game like BG2 you'd have to click multiple times just to succeed in Open Locks on a chest, but why would you waste time on this in a single-player game, if you could open it right away - one might ask?

In comments you might also present your favorite method/dice combination/mechanism on how would you make such skills progress, how many levels you prefer for your skills (and what's the max level cap for chars) and how do you handle yet-untrained skills (i.e. something that is a class skill, but you have not invested any points in it yet). Also, how many different possible outcomes do you prefer in your system? I just mentioned success and failure as the two basic ones, but we can talk about critical failures, critical successes, success BUT and failure BUT situations as well.

 

[Chaotic_Heretic]

Sacrifice a chicken to the RNG gods and pray for success. If a goat is sacrificed, pray for critical success.

 

[Desolate Dancer]

Sacrifice a chicken to the RNG gods and pray for success. If a goat is sacrificed, pray for critical success.

Neat. I updated my post to include the other possible outcomes as well, since failure and success is just the two basic eventualities.

 

[Bester]

In case of lockpick, there shouldn't be any chance of failure if have the required skill. Otherwise you have to save game before each attempt, reload if you critically failed.
You can leave a 1/1000 chance of success if you lack the needed skill by a few points. Just for lulz.

^ This is how I'd do maybe all non-combat skills. Combat is a different beast.

 

[Desolate Dancer]

Otherwise you have to save game before each attempt, reload if you critically failed.

So I guess you hate critical failures then. As for save scumming, we can't really do anything about it. Even the best and most perfectly made game will fall victim of it, unless you make it so that each skill attempt, dialogue choice etc... will use a super-variable that will overwrite save game files... which of course would produce its own ire, and own circumvention (e.g. Diablo II where you could alt+F4 in order to avoid saving something you didn't want to). But I do agree on the notion that we should try and skim everything that is prone to invoke save scumming behavior in players.

 

[Bester]

(e.g. Diablo II where you could alt+F4 in order to avoid saving something you didn't want to).

Easy fix: whenever your character's state changes, save it immediately like in Dark Souls.

 

[deuxhero]

2 for things that can be retried with meaningful time cost (like actions in combat where failing a check costs your action but you can try with your next action), 3 for everything else.

 

[Desolate Dancer]

Easy fix: whenever your character's state changes, save it immediately like in Dark Souls.

True, but DS is more like the exceptions that strengthens the rule, the entire game was sold on its veritable difficulty (at least that's how I heard about it from normies). And yes, it was surprising how people went nuts over such an unconventional game at the time, not just hardcore players.

2 for things that can be retried with meaningful time cost (like actions in combat where failing a check costs your action but you can try with your next action), 3 for everything else.

Good point, we should consider the duration of such attempts. If it was annoying in BG that we had to click multiple times to disarm a trap, NWN took this even further where you both had to wait for the attempt to succeed/fail and then retry it if it was the latter.

 

[mondblut]

The one that checks for attack skill.