What does the d20 system do well?

[Grinner]

I have an abusive relationship with D&D. As much as I want to hate it, I just can't help but go back to it, hoping it will be better this time. The system, with all of the OGL material and alternate rules available, is just so ripe for modification and remixing.

I'm also of the opinion that there is no "universal" RPG, as much as Steve Jackson may protest. Each game can only accomplish a single kind of gameplay experience. Alternate rules and hacks can nudge it towards different experiences, but in the end, each system is optimized for a certain kind of game.

With these two things in mind, what kind of game do you think I can hope to stitch together from the dissected remains of the D20 SRD? It seems to not stray too far from its wargaming roots, but there must be something else I can do with it...But what?

[Yora]

Character creation is great. Actually doing stuff not so much.

The whole way BAB and saves adavance, how all ability scores have the same modifiers, how you get feats, and increase your skills are all great.
But actually using skills is often tidious (especially stuff you can do in combat), or things liks grappling, fighting on a grid, bull rushing, fighting on slippery floor, and all these things are made way too complicated.

[Novawurmson]

Character creation is great.

Also, character class creation; it's fairly simple to make a new class, modify an old class, or even create a new sub-system.

[Roguenewb]

The D20 system as a whole, and 3.5 in particular, is a good Modelling Engine. Almost any effect you can think of can be represented, and most characters from most IPs can be represented, at some level. I can model any character from a house cat to Yahweh. I've played in parties with a Kwai Chang Caine expy (Kung Fu), an Indiana Jones, Professor Farnsworth and Tony Stark (warforged warlock/self components, kind of expy). We fought against variations of Lex Luthor, Satan, PETA, and Inigo Montoya.

I've played other systems, but I've never seen a system that can model that much diversity. This comes with a downside though. There are soooooo many rules, so many interactions and wonkiness that if everyone isn't doing their part, and doing it well, it gets so bogged down and confused that the game goes no where. In the example above we used ToB for Caine, the Inspiration system for Indiana, ECS for artificer for farnsworth, and invocations for Stark. Not to mention dips and splashes and bits from all over the place. The character planning pages we worked on were probably a total of like 30 pages long to get these characters from level 6 to 15. That's insane. Any combat where the artificer forgets how infusions interact with AMF means we sit for ten minutes while its researched. When does inspiration refresh? Are invocations dispellable? Which of these manuevers count as supernatural again? It's mind boggling.

All told though, D20/3.5 allows for a huge diversity of modelling and really rewards players with high system mastery. Once the DM and the players have established an understanding and trust, each of you just turns to the other as the expert and you can smash around faster.

tl;dr D20 good for modelling all sorts of characters and effects, diversity causes complexity creep.

[Thomar_of_Uointer]

I think one of the biggest strengths of the D20 system is the class tables. You can look at that table and say, "hey, in two levels I get Evasion!" It makes it fun to play a low-powered character because you can plainly see your character advancement.

[137beth]

Also, character class creation; it's fairly simple to make a new class, modify an old class, or even create a new sub-system.

This is my vote. 99% of the supposed problems people complain about with the system are actually problems with class design.
There is one thing that bothers me which can't be fixed just by replacing classes with other classes: the skill point system. What I've done is replace it with a system in which each skill rank costs more points than the previous one. I also made it so that the minimum total skill ranks you can have is the same as in 3.5, so people who max as many skills as possible won't notice the change, and people who spread out their skill points get a lot more.

Ultimately, the real strength of 3.X is its flexibility. It is really easy to design new classes, modify old ones, introduce new feats, items, spells, races, or monsters.

[Thomar_of_Uointer]

There is one thing that bothers me which can't be fixed just by replacing classes with other classes: the skill point system. What I've done is replace it with a system in which each skill rank costs more points than the previous one. I also made it so that the minimum total skill ranks you can have is the same as in 3.5, so people who max as many skills as possible won't notice the change, and people who spread out their skill points get a lot more.

What do you think of the Pathfinder version of the skill system?

http://www.d20pfsrd.com/skills#TOC-Acquiring-Skills

[137beth]

What do you think of the Pathfinder version of the skill system?

http://www.d20pfsrd.com/skills#TOC-Acquiring-Skills

It's a little better, but I have the same issues with PF skills that I do with 3.5 skills--dipping one rank in a skill costs the same as improving your best skill, so it encourages focusing most of your skill points into a few skills.
Here's what I do:
Increasing a skill from 0 to n costs an amount in the following table:
{{1, 1}, {2, 3}, {3, 6}, {4, 10}, {5, 15}, {6, 21}, {7, 27}, {8,
34}, {9, 42}, {10, 51}, {11, 60}, {12, 69}, {13, 80}, {14, 91}, {15,
102}, {16, 114}, {17, 126}, {18, 139}, {19, 153}, {20, 167}}

So to increase a skill from 9 ranks to 10 ranks, it costs 9 skill points. Each level, you get enough skill points to increase S skills from (your old level) to (your new level), where S is the number of skill points you would get each level under the Pathfinder skill system. Max ranks and class skills work the same as in Pathfinder. Unused skill points can be saved from one level to the next (while it may appear to involve more calculations, note that this system requires no more or less work to keep track of except when leveling up. )
The point is that if you focus all your skill points among as few skills as possible, you get the same result as you would in PF. On the other hand, if you put only 2/3 your maximum ranks in every skill, you can train twice as many skills as you could in PF. It encourages dipping "knack" skills, and skillmonky classes like the rogue can get a decent modifier in all or almost all skills.
Once again, the strong point of 3.X/PF is that it is highly flexible--I was able to implement my modified skill system with no clunkyness at all: it fits perfectly with the rest of the system.

[Thomar_of_Uointer]

It's a little better, but I have the same issues with PF skills that I do with 3.5 skills--dipping one rank in a skill costs the same as improving your best skill, so it encourages focusing most of your skill points into a few skills.
Here's what I do:
Increasing a skill from 0 to n costs an amount in the following table:
{{1, 1}, {2, 3}, {3, 6}, {4, 10}, {5, 15}, {6, 21}, {7, 27}, {8,
34}, {9, 42}, {10, 51}, {11, 60}, {12, 69}, {13, 80}, {14, 91}, {15,
102}, {16, 114}, {17, 126}, {18, 139}, {19, 153}, {20, 167}}

So to increase a skill from 9 ranks to 10 ranks, it costs 9 skill points. Each level, you get enough skill points to increase S skills from (your old level) to (your new level), where S is the number of skill points you would get each level under the Pathfinder skill system. Max ranks and class skills work the same as in Pathfinder. Unused skill points can be saved from one level to the next (while it may appear to involve more calculations, note that this system requires no more or less work to keep track of except when leveling up. )
The point is that if you focus all your skill points among as few skills as possible, you get the same result as you would in PF. On the other hand, if you put only 2/3 your maximum ranks in every skill, you can train twice as many skills as you could in PF. It encourages dipping "knack" skills, and skillmonky classes like the rogue can get a decent modifier in all or almost all skills.
Once again, the strong point of 3.X/PF is that it is highly flexible--I was able to implement my modified skill system with no clunkyness at all: it fits perfectly with the rest of the system.

I think Pathfinder class skill do encourage splashing. Your first skill rank gives you a +4 bonus, which is significant if you ever need to make that skill check.

D&D Next is apparently using more of a flat DC curve for everything (for example, a level 20 fighter only has a +5 to hit). Sadly, the skill system hasn't gotten much focus (and for good reason, because 3.5's main design flaw was class and spell design). I like that kind of system because it's easier to homebrew for.

Anyways, back on topic. The core D&D rules also do a great job of using archetypes for classes. D20 Modern fell flat on its face in this regard. If you can't describe a class's major abilities using only the full-body character art on the page, it's not worth writing a class for it. The philosophy is, "your abilities come from what you are", rather than other systems which have you pick abilities appropriate to the character you have in mind. It saves a lot of time in character creation, and does a good job setting the tone for the setting.

[Grinner]

tl;dr D20 good for modelling all sorts of characters and effects, diversity causes complexity creep.

Can't say I agree on the first point, since I've found that the numerous anime and superhero systems can model characters much better. In fact, that's exactly what they're expressly designed to do.

The second point is certainly true, what with all the different abilities a multiclass character receives, particularly a multiclass spellcaster.


Anyways, back on topic. The core D&D rules also do a great job of using archetypes for classes. D20 Modern fell flat on its face in this regard. If you can't describe a class's major abilities using only the full-body character art on the page, it's not worth writing a class for it. The philosophy is, "your abilities come from what you are", rather than other systems which have you pick abilities appropriate to the character you have in mind. It saves a lot of time in character creation, and does a good job setting the tone for the setting.

This is both what I love and what I mildly dislike about the system. You can do homebrew easily by

creating a concept,
selecting a few choice abilities from extant classes and monsters too,
distributing them among three to twenty levels,
and slapping on a hit die and several progressions.


And then you realize there's absolutely no reason to.

[Thomar_of_Uointer]

And then you realize there's absolutely no reason to.

If you're homebrewing and all of the players are skilled homebrewers, yes. But for new players that kind of structure is an incredibly helpful tool to learning the game.