I created my own skill system for AD&D some 20-25 years ago because I didn't like the NWP system. As a default, nobody got any special skills. They were things you could opt to learn (you could begin with as many as three, or just learn whenever it seems logical throughout the game), and the cost would be a penalty to all future XP earned.

For me, there were a lot of benefits to doing it this way.
-It doesn't have a fixed number of skill or proficiency slots, so you don't have to alter how you imagine your character to fit the game system.
-It's kind of like customizing your class--the more abilities you have, the longer it takes to level, so unlike most skill systems, I feel this one actually fits well with a class-based game.
-They were what I call player-optional. If a player decides this is needless complexity, they don't need to have any special skills for their character and not be at any disadvantage.
-Because the number of special skills can be zero, many characters can still be summarized in one line (especially NPCs).
-It bridged the gap between PCs who earn XP and NPCs who don't. The more you load up on skills, the less XP you earn, right on up to a -100% XP penalty, so an adventurer can transition into an expert.

I opted for a percentile mechanic. The benefits of that were
-They functioned just like Thief abilities.
-It interfaced well with the expert hireling section of the DMG, as some of them already had percentile listings. The armorer, for example, up to 50 skill could only make ring, scale, or studded armor, 76 skill was required to make chainmail, etc.
-It gave a naturalistic way of applying the numbers to the system. 50 skill as a hunter increased probability of surprising game animals by 50%, so a base 2 in 6 surprise probability becomes 3 in 6.

The skills I used were pulled from the secondary skills table plus the expert hireling professions. I was dealing in skill bundles, yes, but skills were bundled according to profession rather than similar activity. For example, D&D has come to use the skill "Sleight of Hand" which includes pocket picking. If stage magicians were a common trope in my campaign, my skill system could make finer distinctions, so pocket-picking and sleight of hand were truly different skills, but in a way that didn't require a 300+ skill list.

A few years later, I would get my hands on the beta edition of Legendary Adventure (it was spelled with a g and not a j during beta), and that was a great fit for what I was doing, so I pulled a few more skills from there. The way some of the skills interfaced with the combat system also struck me as really neat, so I began to blaspheme and include skills that could have a substantial impact on combat. In a way, this almost recreated the idea of "attack ranks" from BECMI. I would have Archery, for example, give hit bonuses with ranged weapons. An elf fighter who hit max level could then just focus on improving Archery skill up the wazoo (who cares about growing an XP penalty when you're at max level anyway?)

I think a somewhat polished skill system that could do all that would be a blessing.

In my OD&D game, I use "roll 2d6 and consult the Reaction Table" to cover any kind of attribute or "skill" checks, with modifiers for relevant class/level, background, environment, etc.  It's fast, easy, and can do simple success/fail or degrees of success.

Quote from: nightlamp;980111In my OD&D game, I use "roll 2d6 and consult the Reaction Table" to cover any kind of attribute or "skill" checks, with modifiers for relevant class/level, background, environment, etc.  It's fast, easy, and can do simple success/fail or degrees of success.

I'm Black Vulmea, and I endorse this product or service.

Welcome to the adult swim.

Quote from: nightlamp;980111In my OD&D game, I use "roll 2d6 and consult the Reaction Table" to cover any kind of attribute or "skill" checks, with modifiers for relevant class/level, background, environment, etc.  It's fast, easy, and can do simple success/fail or degrees of success.

 Clever and suitably ill-defined. I like it.

Quote from: nightlamp;980111In my OD&D game, I use "roll 2d6 and consult the Reaction Table" to cover any kind of attribute or "skill" checks, with modifiers for relevant class/level, background, environment, etc.  It's fast, easy, and can do simple success/fail or degrees of success.

That's pretty cool. How do you handle the various modifiers? It's really easy to break a 2d6 roll when you are piling on modifiers for "relevant class/level, background, environment" and so on. I know some people have tweaked the Reaction Table to a 3d6 to allow for modifiers. This might be a case where that version is useful.

Quote from: Baulderstone;980317That's pretty cool. How do you handle the various modifiers? It's really easy to break a 2d6 roll when you are piling on modifiers for "relevant class/level, background, environment" and so on.

Here's what I do for modifiers:
General attribute bonuses are relatively low (15-17 = +1, 18 = +2).  
For a relevant background or a non-combat "skill" that might be known to a specific class (a Cleric recognizing a rite or ritual of another religion, a Fighter observing an enemy camp's defenses for weak points, etc.), a +1 or +2 modifier.  
For thiefly tasks such as picking a lock or sneaking, anyone can try it but Thieves add 1/2 their level (round up.)  
Negative modifiers usually range from -1 to -4 for environment (poor light, high winds, etc.), encumbrance, and so on.  

It's easy and flexible, exactly what I want when I'm running OD&D.

I really like 5e's proficiency bonus for all proficiencies.  I dislike the exception based design that clogs much of the system but skills aren't part of the problem.

Quote from: Baulderstone;980317That's pretty cool. How do you handle the various modifiers? It's really easy to break a 2d6 roll when you are piling on modifiers for "relevant class/level, background, environment" and so on. I know some people have tweaked the Reaction Table to a 3d6 to allow for modifiers. This might be a case where that version is useful.

I've tended to leave the table alone, and include situational modifiers in what the outcome is. Partial success when playing a skilled stealth character might mean being able to sneak up to a guard on duty (full success would allow you to sneak past), where for a guy in plate armor it might only work for sneaking up on a guy off duty sitting at the campfire.

Since I last commented on this thread, I must have written about 7 different (and incompatible) methods for OSR skills...

My conclusion, I think, is that they are all useful, depending on what you want for your game.

- Do you want stats to matter more than level, or the opposite?
- Do you want PCs to start weak or somewhat competent already?
- Do you want generalization or specialization (i.e., "Perception" or "gear notiises", "finde hidden doors", etc.)?
- Do you want the perfect system for every skill, or just use one system for everything (DC being similar to AC etc)?

And so on. Any system can work. WotC D&D has the worst systems, IMO, but anything will do depending on your goals.

I fucking hate skills.

Unfortunately, most players love them.

I have never seen a D&D skill system that I've really liked and I've used all the official, all the OSR, and a dozen of house ruled ones over the years.

Palladium tacked on RQ's skill system. Maybe the best of the worst ideas in 80s.

C&C's Primes is probably the current best of the worst. It has loads of issues, but it moves fast at the table and gets the job done. I gotta respect it for that. At the end of the day, if my players want there to be skills, I demand the system to be swift and easy to adjudicate on the fly.

I've never used, wanted, or missed anything in an OSR D&D game called "skills" that wasn't prefixed by "thief". My preference for OSR is that characters have class abilities, and that a character's class is their profession, and continued advancement in that profession requires 100% of the character's "skill development" time.

In game, I prefer that characters "describe actions" and the GM "resolve actions". "Testing skills" isn't part of that vocabulary.

That said, when "resolving actions" in an OSR game I like to use whatever core mechanic the game uses for resolving action in general, rather than introducing a new system, and my preference is for that core mechanic to be strongly class- and level-based.

Quote from: Eric Diaz;980719And so on. Any system can work. WotC D&D has the worst systems, IMO, but anything will do depending on your goals.

For all the talk of a unified mechanics WotC made when it released 3E, it really was a fucking mess. TSR's most common approach in late AD&D 1E supplements, RC/BECMI, and AD&D 2E was non-weapon proficiencies. It covered all the shit that wasn't part of a class. Not everyone needed it, but it worked fine, and they were smart in tying it to attributes and not level. Level described how good you were at your class. It wasn't an entirely unified mechanic, but it was simple.

With the supposedly more unified 3E, you had NWPs split up into both skills and feats. It lead to wonky crap like Tracking being both a class ability and a feat, and you needed another skill to use it. And rather than the mostly binary NWPs, you have to dutifully put points into a skill every level or it would quickly diminish into uselessness and DCs raised every level.

In theory, you could vary where you put your points every level, which was the whole reason you got a fresh set of point to spend each time. Of course, that was just a noob trap. If you wanted a skill to stay useful, you needed to pick your skills carefully at first level and dutifully apply one point to each at each level to avoid falling backwards of the accelerating conveyor belt. As varying your skill buys from one level to the next was a trap, it made allocating skill points into stupid busy work. It's easier just to pick NWPs at first level and be done with it.

The inflating DCs in 3E meant that if you kept putting in one point per level like you were supposed to, what mattered most was your ability modifier. You might as well just buy the NWP, and make flat ability checks.

I'm probably not saying anything that we all don't know, but I have been looking at all the skill systems in D&D and its various clones recently, and WoTC really did do the absolute worst job it.

Quote from: Baulderstone;980855The inflating DCs in 3E meant that if you kept putting in one point per level like you were supposed to, what mattered most was your ability modifier.

This is true in relative terms, when comparing one (equal-level) character's ability to another. But higher DCs still theoretically represent increasingly difficult tasks and a character's ability to perform them.

Or do some people actually play 3E as if the skill DC ("easy", "moderate", "difficult", etc.) is relative to the character attempting the task? If that's the case, I see your point, though no one I know played it that way.

Quote from: Zalman;980870This is true in relative terms, when comparing one (equal-level) character's ability to another. But higher DCs still theoretically represent increasingly difficult tasks and a character's ability to perform them.

Or do some people actually play 3E as if the skill DC ("easy", "moderate", "difficult", etc.) is relative to the character attempting the task? If that's the case, I see your point, though no one I know played it that way.

IIRC this was EXPLICITLY the case in 4e, but not in 3e.

In any case, some skills will be opposed, so a character has to deal with other characters of the same level (and NPCs had classes in 3e).... I dunno, maybe it can be a problem.

Quote from: Eric Diaz;980905IIRC this was EXPLICITLY the case in 4e, but not in 3e.

In any case, some skills will be opposed, so a character has to deal with other characters of the same level (and NPCs had classes in 3e).... I dunno, maybe it can be a problem.

I guess the upshot is that if you build the character "like you're supposed to" (and I agree that 3e strongly "encourages" that line), that all the fuss over "skills" and "skill points" winds up being determined by class and level in the end anyway. To me the lesson there is that a superior system will cut right to the chase of which factors you want to stress when resolving actions.