2007-10-13 Skills

I’ve been thinking about skills, yesterday.

I find cross-class skills frustrating as a player. They thwart my ambition in creating curious and interesting characters. I also found they complicated skill-point checking. Away!
Maxing out your skills is quick and easy, but inflexible. Use a different abstraction to get rid of the inflexibility. In other words: The number of skills you know is fixed, and the effective skill rank is equivalent to a maxed out skill, but every four levels you gain another skill.
Variable skill points are a mess. Maybe Rogues need lots of skill points because they are not smart and still need to learn a lot of skills. Maybe Wizards are smart and get few skill points so that it balances out. The solution is to change the application of the various skills. Give everybody the same number of skills (independent of intelligence and class), and make typical rogue skills very broad in application. Keep the skills for other classes rather specific (physical and knowledge skills in particular).
Having characters that can Climb but not Swim etc. are fun, thus there is no drawback in not unifying these skills.
Balance, Jump, and Tumble → Acrobatics. These skills have seen little use in my games.
There should be enough skills to require rogues to specialize and enough Knowledge skills to require wizards to specialize.
Gather Information, Search, Spot and Listen → Perception. This is the skill to use for rogues that want to specialize in scouting.
Hide and Move Silently → Sneak. This is the skill to use for rogues that want to specialize in breaking in and infiltrating.
Sleight of Hand, Forgery, and Disguise → Trickery. This is the skill to use for rogues that want to specialize in spying and stealing.
Open Lock, Disable Device, Use Magic Device → Traps. This is also the skill to use if you want to build traps.
Get rid of Use Rope and Escape Artist. We rarely use them and we find that it doesn’t enrich the game. Use Trickery vs. Perception if you want to do an escape scene.
Craft, Profession, Spellcraft, and Decipher Script → Knowledge. In my games most Knowledge skills have been used again and again. And I’ve noticed a problem with my fighter who wants to be a map-maker. Is it really necessary to have two separate skills, Craft (map) and Knowledge (geography)? I don’t think so.
Intelligence remains important because the Knowledge skills remain important, and they use Intelligence as a modifier.
Appraise, Concentration, Heal and Survival sound interesting in theory but turned out to be boring for the game. Use Perception if you have the Track feat. Use an appropriate Knowledge skill for healing, if you need it – Knowledge (medicine), for example.
Bluff, Handle Animal, Sense Motive, and Diplomacy → Persuasion. Use opposed persuasion rolls to resolve conflict peacefully.

What follows is the list of skills I want to use. Notice that I changed Swim from Str to Con, and that Forgery no longer requires Int since Trickery is just Dex. Pet peeves... 😄

Climb (Str)
Ride (Dex)
Swim (Con)
Acrobatics (Dex)
Perception (Wis)
Sneak (Dex)
Trickery (Dex)
Traps (Dex)
Persuasion (Cha)
Knowledge (Int) with subtypes arcana, religion, planes, history, nature, dungeoneering, geography, medicine, and many more for various specific cities, countries, races, or whatever takes you players’ fancy

As you can see, Knowledge subtypes is what most people will be gaining at the end. And I don’t mind.

Each character starts the game with two skills. Every four levels you gain a skill.

At level eight, a typical fighter would have Climb, Ride, Swim, and Acrobatics. A typical level eight rogue would have Perception, Sneak, Trickery, and Traps. A typical level eight cleric might have Persuasion, Knowledge (religion), Knowledge (planes), Knowledge (history). A typical level eight wizward might have Knowledge (arcana), Knowledge (planes), Knowledge (history), Knowledge (alchemy).

Making a more unusual character will be no problem, however: A level eight fighter having Climb, Ride, Perception, and Knowledge (cartography) is easy to do.

It’s not as simple as the M20 system which my players did not like. But it’s simpler than the default D&D 3.5 system. I’ll have to give it a try one day.

M20

For simplicity’s sake, I’ll do away with the +3 for maxed out skills. Thus, having a skill trained basically allows you to do a d20 + level + ability modifier check.

I’m not going to introduce any bonus for untrained skills. Just use a d20 + ability modifier for your roll...

I should do a little cross-checking with Justin Alexander’s essay on skills in D&D, D&D: Calibrating Your Expectations.

D&D: Calibrating Your Expectations

*Update*: Actually Justin Alexander wrote an essay arguing against a simplification: 4th Edition — Thoughts on Skills. He seems to have a point...

4th Edition — Thoughts on Skills

Comments

(Please contact me if you want to remove your comment.)

⁂

Interesting thoughts about skills. I’ve generally given out more skill points in my own campaigns, and eliminated cross-class skills entirely, but I can see the advantages of eliminating the skill point system entirely.

However, having the large disparity between trained (d20 + level + ability mod) and untrained (d20 + ability mod) means that, once you reach mid-to-high levels, a character is either at the pinnacle of ability for a skill (for his level) or effectively worthless. In the current system, you could be very competent in a few skills, have average ability in a few more, so the drop-off isn’t as marked. The new Star Wars Saga Edition rules supposedly give you d20 + half level + ability mod in all untrained skills (however, some skills still can only be used trained, or have usages that only trained individuals can access), +5 if you are trained, +5 if you have Skill Focus. Maybe under your system, you could introduce the idea of specialization (either through feats, class abilities, or when you gain a new trained skill, also add a specialization), and give a base d20 + half level + mod for untrained skills. You do lose some verisimilitude (a 20th level bookish wizard is a better climber than a strapping 5th level ranger).

Perhaps a middle ground is to introduce a middle ground: trained, dilettante, and untrained skills. Trained skills gain your level + ability mod; dilettante skills gain 1/2 level + mod; untrained rely on just your mod.

The only other problem I see is with rogues (and to a lesser extent bards and rangers). Int is not usually their highest stat, but it usually ranks 2 or 3, meaning they have at least 50% more skill points than wizards (who only get 6/level with an 18 Int) and twice as many as other classes. This is, in many ways, a key feature of rogues, and would take a hit in your proposed system.

But I think you are definitely heading in a great direction.

– Adrian 2007-10-16 09:38 UTC

---

Yeah, I’ve been looking at the Star Wars Saga skill system, and I’m not sure I like how you’re getting better and better with your untrained skills. (You’re welcome to take a look if you’re interested.) Unfortunately I can’t make an informed decision because I have so little experience with high level games. “At the pinnacle of ability for a skill […] or effectively worthless” – that doesn’t sound too bad to me. I think I like it!

As for the “skill monkey” taking a hit: My hope is that the broad application of the rogue skills makes up for the perceived reduction of skill points. Perception, Sneak, Trickery, Traps, Persuasion. Basically by picking four out of these five for your level eight rogue you will be an awesome rogue. I’m claiming this is equivalent to using the d20 rules. You’d maximize 11 skills (assuming Int 16 or 17) and pick them from the following list of 17 skills: Bluff, Sense Motive, Use Rope, Escape Artist, Tumble, Gather Information, Search, Spot, Listen, Hide, Move Silently, Sleight of Hand, Forgery, Disguise, Open Lock, Disable Device, Use Magic Device... You’d cover less ground in d20. So I’m not sure rogues will in fact loose. Obviously if other classes decide to pick rogue skills they’ll make just as great spies and thieves. As far as I’m concerned, however, that’s ok. They have to sacrifice their own typical physical or knowledge skills.

At level 20 a wizard will have 3 + 5 (one at 4, 8, 12, 16, 20) = eight skills maximized (”trained”) for example. He’s free to pick Climb, Ride, Swim, Perception, Sneak, Trickery, Persuasion, and Knowledge (nature) if he wants to play a pictish shaman warrior. And he’ll be able to climb and ride like a ranger. Just don’t ask him about the books he has read. 🙂

I’m not sure how another of my changes will turn out: Using the d20 rules, the number of skills you can maximize never changes. My system would allow characters to learn more and more tricks. I’m assuming people will learn new Knowledge skills because these have seen so much use (aside from Listen & Spot). Obviously some playtesting is required.

– Alex Schroeder 2007-10-16 14:26 UTC

Alex Schroeder

---

I agree that getting better at untrained skills just because you level doesn’t really jive with my sensibilities either (though this is a heroic game; regardless of actual physical ability or skill, don’t heroes always seem to outshine the normal people?). The problem with introducing new skills every four levels is that a character goes from being useless in a skill to suddenly being a master (for that level). In some cases, not a big deal; in other cases, it means you might have to suffer through several levels before your character can even consider doing things. In the core rules, your example rogue is about as good as he can be (for his level) in 11 skills throughout his career. In many ways I think your system is a lot better and more elegant, but it creates a very stepwise progression in character capability. I will say that your proposed system can really allow players and DMs to find creative uses for skills, since each skill is so broadly defined, and I think that is a great benefit (because it inserts DM judgment back into the equation, which I think I like).

At high levels, in my experience, the problem becomes one of specialization. To challenge the character with max ranks in X skill (X typically being things like Listen, Spot, Search, Hide, and/or Move Silently, but sometimes also things like Balance, Climb, Jump, and Swim), the DM has to craft challenging situations and set DCs so high as to put it effectively out of reach for the remainder of the party. Now, the DM can also decide, “This player has invested a lot of resources in being awesome at skill X. So let that character auto-succeed, and set the DC to give the rest of the party a fighting chance.” But that offends my DM sensibilities --TANSTAAFL and all that. That is why I’d like to see a slightly less binary system than untrained/trained skills (and, for that matter, the difference between good and bad saving throw progressions, but that is another day).

– Adrian 2007-10-16 15:51 UTC