For research's sake: What's Wrong with 3e/3.5/PRG?

[malonkey1]

As the title asks, I want to know: What, in your opinion did D&D 3.0 do wrong mechnically, what did 3.5 fail to fix from that (and what problems did it add?), and finally, what of that did Pathfinder fail to fix (or added to the pile). I must ask that we try not to turn this into an edition war, as it's obvious that all three are gonna have glaring mechanical issues, and I just want to get a handle on it. For those who were curious, I was planning to make a loosely d20-based RPG via RPGMaker VX Ace (excellent program, worth the money), and I wanted to know what needs to be fixed/ameliorated. I await your responses with baited breath.

[Rebel7284]

They all failed at any semblance of balance between classes.

[Falcon X]

This is a question that heavily relies on opinion, so I'll give my own two cents.

I'm a person that prizes good storytelling and creativity over dice rolling. I've found that the current generation of people approach gaming like they do the school system.
In the American school system, kids are taught how to beat the test, and once they can beat the test, then maybe they might learn something and thrive. Great teachers consistently give simple tests, so that the student can pass that threshold quickly.
Likewise, a gamer must first master the rules before they can dig into the roleplaying aspects of the game.
Complex systems like D&D or Shadowrun almost always wind up with players who crunch numbers and nothing else, as the systems are complex.
Give them a simple system, like Numenera, Faulkenstein, etc. and they learn how to play quickly and then start to thrive in creativity.

So, what would I do with D&D?
I would leave character creation with many beautiful options that make the player feel like they are in something complex.
But make the actual game physics incredibly simple. Like everything being a single dice roll off of one of the six stats.
That way, it's a simple feel, but it won't take but an hour to grasp all the rules.

[BowStreetRunner]

I never really played 3.0 - I just bought the books when they announced 3.5 (D'OH!!!) so didn't really bother to learn those rules.

As for 3.5 > Pathfinder, I feel that while there were certainly changes that could go either way based on preference, these were a few of the things I felt PF fixed over 3.5.

Skill points/ranks simplified.

No more situations where a multi-class character has to decide which class to take first because those skill points are x4. Also encourages dipping in class skills, since the first rank is worth more (goodbye min-maxed ranger who dumped climb and use rope entirely).
Fewer skill rolls to make (hide + move silently vs listen vs spot, ugh!) and fewer skills to keep track of (jump, tumble, balance = acrobatics, yay!).
No more max ranks and double cost for cross-class skills (the fighter can now actually take perception ranks and perform sentry duty, hooray!).
Skill monkey classes benefit particularly well from the fewer skills and reduced cross-class limitations, which still doesn't make them tier 1 but at the least it helps a bit.


Special Attack system simplified.

No more keeping track of a half-dozen different systems and getting confused between the mechanics when trying to figure out your bonus. CMB vs CMD is much easier.


Experience Point System

No more XP penalty for multi-classing. Seriously, after 3.5 put out all of the classes, prestige classes and races that they did the favored class system seemed pretty silly. If your multi-classed fighter/ranger/barbarian isn't more powerful than my single-classed wizard, then I don't see the reason to penalize you for it.
No more XP costs for spells, magic items, or pretty much anything else. I really never though XP was a good form of 'currency' for any in-game mechanics.
No more level-adjusted races. 'nuff said.

[HammeredWharf]

There's plenty wrong with them:

Simplistic handling of everything but combat.
Extreme lack of balance everywhere, from skills to classes to races to feats to spells.
Poor wording and unclear rules.

Still, the rules form a decent foundation for playing. Mostly because some of the material is rather good.

[shadow_archmagi]

As the title asks, I want to know: What, in your opinion did D&D 3.0 do wrong mechnically, what did 3.5 fail to fix from that (and what problems did it add?), and finally, what of that did Pathfinder fail to fix (or added to the pile). I must ask that we try not to turn this into an edition war, as it's obvious that all three are gonna have glaring mechanical issues, and I just want to get a handle on it. For those who were curious, I was planning to make a loosely d20-based RPG via RPGMaker VX Ace (excellent program, worth the money), and I wanted to know what needs to be fixed/ameliorated. I await your responses with baited breath.


In no particular order:

1. Characters advance via feats, stat increases, skillpoints increases, BAB, class features, hit dice, saves, and wealth, with different classes receiving different amounts of each. This creates a very "messy" system, where it's hard to eyeball the balance between, say, a fighter and a wizard, because they advance in very different ways.

2. The d20 itself isn't a very good way of modelling a lot of things. Many of the tests are binary, which is frustrating enough (Poison, for example, is either effective or completely ignored.) But even in situations like, say, knowledge tests, where progressively higher numbers are progressively more helpful, the d20 itself has a flat probability distribution. (That is to say, not only is the average result of 2d6 seven, but seven is also by far the most common result. With a d20, you're as likely to get a one as an eleven.)

3. Hitpoints are essentially binary; you are either alive or dead, with no status conditions from being anywhere inbetween. If you fall off a cliff and survive the impact, you can walk home.

4. Caveat to the above: 0 through -10 exist as an inbetween state, where one is neither healthy or dead, but damage values increase as level does, so high level characters are almost never disabled.

5. There's no real coherent definition for a feat, or estimation for how powerful one should be. Example: Skill Focus gives a trivial increase to a single skill, Invisible Spell makes all your magic invisible, and Improved Feint lets you Feint as a move action. One is a direct increase in output, one is a hard-to-quantify upgrade, and the last decreases the action cost of a normally overpriced tactical option.

6. Scaling with party size: Because characters advance in so many ways, an enemy with a lot of HP probably has a lot of HD, and thus also lots of BAB, lots of AC, etc etc. It's thus difficult to challenge large parties without pulling out monsters that one-hit-kill them in an unfun way.

7. Action Economy: There are relatively few things that grant extra actions, and extra actions are great, so if you skip out on extra actions, you are going to have a bad time. The PCs, by virtue of being roughly a half dozen people, get roughly six times as many chances to act as a single villain, forcing the GM to either have the PCs exclusively fight Ginyu-Force style evil teams or something.

8. Many Spells are far, far more flexible than mundane options. A Rogue might encounter situations where no rogue skill or class feature or feat can possibly apply, but even a Wizard who only knew the spell Fabricate would always have something to offer.

9. Spells don't require commitment, for the most part. A fighter who picks up archery feats is forever more an Archer-Fighter, and can serve only in that capacity. A wizard who picks up a bunch of fire spells can prepare ice spells tomorrow, as the need arises.

10. Spells tend to be more reliable and more powerful than mundane options. Pick Lock picks locks if you roll high enough; Knock picks locks, end of story. Jump may or may not get you across the chasm; Fly gets you across the chasm.

11. Without Tome of Battle, there aren't a lot of interesting or useful things to do in combat if you're not a spellcaster. Trip and Grapple can be useful but require significant investment and stop being useful against anything too large. Bull Rush only works if there's something worth pushing the enemy into. Feint makes you spend your turn, roll dice, and then depending on the outcome of the die roll, maybe get increased chance to hit next round.

[purpenflurb]

I can't speak on 3.0, but I do have a bit of experience with 3.5 and PF. What I like about 3.5 is, I guess sort of in contradiction to one of the earlier combats, the complexity. 3.5 gives me options. I was sitting down explaining the game to a friend, and he was trying to come up with crazy character concepts like "archer that shoots bears" and for every one of them I could basically come up with a way to make them work. I always feel like any concept I could possibly conceive of would be achievable in 3.5 with enough creativity and knowledge of the rules, which I think can make for a cool and varied landscape.

What I like about pathfinder is, conversely, it is a lot easier to pick up. I am actually not fond of their changes to skills/items by and large, but I like the classes, and in a lower player game would probably give a newer player the pathfinder base classes rather than the 3.5. They generally come built with more interesting options nad readily available alternate features, the 3.5 fighter is the epitome of a class with absolutely nothing of interest going for it.

[Friv]

My personal opinion:

3.0 made two big mistakes that echoed throughout the mechanical framework:
1) They overvalued Base Attack Bonus and HP.
2) They removed a lot of the barriers to casting spells, and then didn't reduce the power of spellcasting to compensate.

The first meant that prime combat classes like the Fighter didn't get nice toys because they already had high combat stats, and the second meant that spellcasters became more dominant than they had been before.

3.5 did not fix either of these problems. Pathfinder took steps towards fixing Problem #1 by giving more class features to the martial classes, and by making cross-classing a little easier, and by boosting the power of a lot of the base-line feats, but didn't do anything about Problem #2 because slowing spell progression drastically would have created a fan riot probably.

What 3.5 did fix was a lot of the more glaring errata-style problems; buffing up the ranger and bard a bit, toning down a few of the more absurd spells, cleaning up language. Pathfinder did the same thing.


No more XP penalty for multi-classing. Seriously, after 3.5 put out all of the classes, prestige classes and races that they did the favored class system seemed pretty silly. If your multi-classed fighter/ranger/barbarian isn't more powerful than my single-classed wizard, then I don't see the reason to penalize you for it.

The penalty's still there, it's just hidden. Instead of making you level up more slowly, Pathfinder makes multi-classing harsh by hurting you for each multi-classed level that you take that isn't favored, so your level isn't worth as much. It's a smaller penalty, though, so still a step in the right direction.

[Eldariel]

For things I feel are just straight-up wrong in all 3:
1) Full attack mechanics. The mechanic is non-sensical; having no difference between normal attack and full attack on BAB 1-5, but suddenly introducing a difference on BAB 6 is plain dumb. Melee attack mechanics specifically also tie down the archetype type that relies the most on being in the right place at the right time (due to passively possessing a threatened area around him) - melee warriors.
2) Melee combat options. Virtually, you have a bunch of combat maneuvers and a bunch of weapon styles. Practically, two-handed fighting is the only worthwhile one from a power perspective, specifically two-handed reach fighting. Two-Weapon Fighting falls into the trap of requiring additional feats for diminishing returns. Likewise, Tripping is just better than the other options. Mounted combat is basically "charge with a lance for dumb amounts of damage" and indeed, easy damage multipliers in general constrict design and are just bad. Same with independent damage bonuses that have different mechanical effects for the same practical effect (Sneak Attack vs. critical hits - also, Skirmish is just dumb).
3) Shields. Related to melee combat options, Shield only protects the Shield-user and only from physical attacks; it completely fails to help protect teammates or the wielder against magical attacks. Using a shield still gives up massive damage potential, which could be used to faster defeat enemies who attack your allies and enemies who attack you with magic or touch attacks or area of effect. In short, Shields give you some defense vs. physical warriors who attack you, but two-handed fighting/two-weapon fighting damage bonuses give comparable advantages. When fighting vs. anything but a physical warriors attacking the shieldbearer, shield is a strictly inferior option.
4) Reach. Related to the above, reach is too big of an advantage to give up. Only way to meaningfully get attacks of opportunity, it just gives you more attacks than anything else. Some kind of way to move out of turn to threaten a larger area with a short range weapon would be needed to balance it out and for short range melee characters to be able to use their primary system-given advantage, melee threat. 5' steps are a key problem in this.
5) Magic vs. mundane interaction. Force effects can only be defeated by magic except for PF, and even there it's extremely hard to interact with them physically. There are very few ways to e.g. attack a flying enemy behind a Wind Wall or to detect an Invisible opponent without See Invisibility. Touch AC doesn't scale so rays are quite hard to dodge too. Outside Tome of Battle (and even with it), the ways for mundanes to directly interact with magic can be counted with one hand's fingers. What happened to cold steel being a power opposed to magic that can cut through magical barriers & al?
6) AC scaling. Attacks scale. AC doesn't. AC gets left behind. Touch AC twice so to the point of being useless without insane amounts of work. Sucks. Also, Touch AC and Reflex are basically the same thing but measured by entirely different metrics. They should be combined somehow, probably bringing saves closer to AC á la 4e.
7) Skill system (less problematic in PF, but far from perfect). Cross-class skills are expensive, mundanes get too few skill points and have too few class skills to meaningfully compared to classes with spells. Skills are, on the other hand, too easy to boost with magic to set meaningful DCs for anything without having the system break on itself (see e.g. Diplomacy). Things like Hide also only scale numerically instead of e.g. enabling Hiding in Plain Sight with higher DC check, even though that should be a direct benefit of your Hiding skills.
8) Mundane survivability. Mundanes don't get spells to make themselves hard to kill. They need inherent bonuses for that. The system gives them precious few boons over casters. Higher HD but same Con-bonus leads to a small advantage in HP which is mostly made up by temporary HP from spells. Mundanes should get more Con to HP. Mundanes don't have spells for immunities and saving throw bonuses; they need better saves and more defensive class features. They get none of that.
9) Spell design. Some spells are just poorly written; Polymorph-line enables a way too wide variety of options (and is written superbad), Planar Bindings & Gate have some stupid mind effecting clauses from totally off-school, Simulacrums forgot to cover spell-like abilities, etc. I also feel spells are a bit too fast (one round should be the baseline, not one standard action) especially compared to weapons - spells by definition have more powerful effects so weapons need to be faster to have any degree of parity.
10) Monstrous races. Specifically, natural weapons & spell-like abilities. They're great, they basically don't scale except by stat increases, there's no way to really learn higher level versions of the spell-likes or gain more Natural Attacks/Natural Weapon Size, etc. Makes increasing HD and gaining levels as a race make the natural abilities lose relevance. Less annoying but still annoying with the standard races too; the racial differences kinda just fade as levels get higher, since they don't benefit of class levels of any sort. Level adjustment is of course also a horrid failure in design (luckily it doesn't exist in PF as such).
11) CR. Okay, yeah, I can pretend it doesn't exist but it's still just a terrible concept as likely to get a party killed as to leave them bored at the supereasy enemies they seem to constantly fight. Plus, it makes it feel like the world scales with the players, which can lead many new DMs to invoke the Oblivion Syndrome. Party power varies quite a bit independent of level. CR is, by definition, dysfunctional.
12) Christmas Tree Effect. Being expected to run around in hundreds of thousands of items (items since you aren't expected to be able to invest more than 50%/25% of it in a single item) works for specific character archetypes (think Jarlaxle from Salvatore's lolnovels), but leaves a ton of archetypes cold and at least for me, makes it feel like my (non-spellcaster) character's abilities are more about the items than his abilities.
13) Multiclassing. Multiclassing is a great concept, but with the increasing amount of experience you need for each additional level, you need increasing gains. Multiclassing "works" for mundanes in 3.0/3.5 (not in PF) since the class gains are basically null, but that's just a side-effect of the class design sucking. Casters basically never want to multiclass in a class that doesn't advance their casting (unless they already have 9th level spell access). That's a problem with the way abilities advance off-class - Tome of Battle and Ardent did this a bit better to the point, where they have level-scaling abilities instead of class level-scaling abilities while still making single classing lucrative with good class design.

The caster multiclassing in particular in 3.5 is horrendous though and needs fixing; Wizard 10/Cleric 10 isn't on the same plane of power as Cleric 20 even though a layman would think it'd basically just be a more versatile spellcaster and it has the same challenge rating and effective character level. And it's not a factor of either of the class being weak, just the multiclassing in the absence of PRCs sucking (PRCs shouldn't be used for this, btw; PRCs specialize in giving additional abilities, not providing basic competence).


That's just off the top of my head, without going into much detail with specific spells. Also, I think the oversimplification of removing armor vs. weapon considerations and the sameism of all the various weapon speeds and such is way underestimating the players and making a Dagger Fighter feel too samey with a Greatsword-wielder unless they actually have specifically tailored PRCs (and that's not always even possible if we're talking about a spellcaster).

[Grod_The_Giant]

The two major issues that persist across all the various 3.X games are linear warriors/quadratic wizards, and binary verses gradual defenses.

Linear warriors/quadratic wizards is a problem you've probably heard about before, and sums up the issues with magic-verses-mundane characters in a nutshell. A 20th level warrior does pretty much the same thing as a 1st level warrior, only with bigger numbers, while a caster gains new options every level. That's the root of the problem, not any individual spell or spells. Give warriors more and better options as they level up, and you'll solve 50% of the imbalance, right there.

Binary verses gradual defenses (http://www.minmaxboards.com/index.php?topic=1533.0)is another, oh, 30% of the problem. Basically, some offensive options (hit point damage, ability damage/drain) are gradual, meaning that they take time to build up and seriously affect you. Others (pretty much any "save or X") are binary-- they either work or don't. You either make your save, or you're screwed. That makes binary defenses much, much more dangerous than gradual defenses. And hey, guess who gets to attack binary defenses, warriors or wizards?

[Urpriest]

I think there's a deeper issue: 3e/3.5/Pathfinder are games made to be D&D, rather than with a particular goal in mind for their play experience or design philosophy. Everything else can be traced back to the fact that, like the writers for Battlestar Galactica or Lost, the designers of these games didn't actually have any plan whatsoever.

[Falcon X]

They all failed at any semblance of balance between classes.

Actually, I saw D&D learn from their mistakes late 3.5. There are lot of people out there, like me, who toss out the core rulebook classes.
If you go straight Tier 3 classes, it's a very balanced game.

Those would be: Tome of Battle, Binder, Dread Necromancer, Beguiler, Warmage, Factotum, and several others.
Just with the ones listed above, you have all the flavor of the core classes, but in a balanced way. Well, minus Druid and Cleric... But good homebrews can take care of that.

[malonkey1]

Wow, these are some really thorough notes, and pretty well-reasoned for the most part. Thanks, guys, and keep em' coming! Perhaps later I'll address ways I might try to tackle some of these issues.

[Gavinfoxx]

Wow, these are some really thorough notes, and pretty well-reasoned for the most part. Thanks, guys, and keep em' coming! Perhaps later I'll address ways I might try to tackle some of these issues.

Someone already did that, by making Legend.

[Gemini476]

Actually, I saw D&D learn from their mistakes late 3.5. There are lot of people out there, like me, who toss out the core rulebook classes.
If you go straight Tier 3 classes, it's a very balanced game.

Those would be: Tome of Battle, Binder, Dread Necromancer, Beguiler, Warmage, Factotum, and several others.
Just with the ones listed above, you have all the flavor of the core classes, but in a balanced way. Well, minus Druid and Cleric... But good homebrews can take care of that.

Wildshape Ranger is basically Druid-, and the Shugenja works fine as a Cleric replacement.

...Yeah, I'd throw in some Tier 4 classes as well. The Warlock and Dragon Fire Adept, mostly.

[Phelix-Mu]

In my opinion, aside from some of the abomination problems with core (polymorph line, demon binding, and terrible feats pretty much cover it), one of the big problems that evolved as 3.5 grew and grew was the way the game materials published synergized with earlier materials. As with many analyses, the breakdown can largely be broken down between mundane and magical options, broadly exemplified here in the categories of class abilities, extras (such as feats, skills, skill tricks, and so forth), and items.

Mundane: Mundanes, abominable in core due to lacking class abilities on par with Spellcasting, did pretty well as splats came out in terms of class abilities they could pick up. Mundane base classes in splats are generally competitive with those in core, or strictly better if we are counting ToB as mundane. PrC dipping can be anywhere from lucrative to game-breaking, but usually a mundane will have to satisfy themselves with schooling opponents due to the scope of the main focus of the game: combat. Feats are also better outside of core, as inside core they were terrible for mundanes. Items, alchemical in nature, were also nice, but generally fewer than magical items.

Magical: On top of the already gangbusters Tiers 1 and 2, we can stack all manner of broke, from alright stuff like Abjurant Champion and Paragnostic Disciple, to the brokest of broke like Incantatrix/IotSFV/PlanarShepherd/Tainted Scholar. Cherrypicking class features can bork the game almost arbitrarily quickly. Feats and such outside of core also devolved into some pretty hyperbolic options, DMM Persist, Twin/Repeat Spell, and various ways to practical metamagic cost reductions. Magical items also grew to encompass huge numbers of options, all of which generally synergize better with renewable magic (i.e., Spellcasting) than mundane schtick.

Finally, the single biggest problem, in my mind, was the breadth of the spell catalog. Even had core been balanced, by the time they finished with 3.5, the sheer volume of spells were certain to contain some borked stuff, and a whole slew of unforeseeable interactions (from a designer standpoint...us crowdsourcers benefit massively from hindsight). None of which are available to mundanes (short of UMD), and for which mundanes have no conjugate or comparable source of daily renewable power.

So, in short, there is simply a better, bigger, and more internally exploitable resource available to magic types. Mundanes are stuck swinging steel, hiding, and throwing sticks, while the casters can really set their sights on nation-building/plot-wrecking/WBL-annihilation, or what have you. The imbalance was always there (and magic should probably have decent awesome just based on genre), but it grew and grew as time went on, until it is pretty laughable from an optimization standpoint.

That said, the enduring aspect of the game that works well is that optimization can work with the huge ruleset in reverse: want a challenge? Intentionally nerf your build and/or concept. There are about a billion ways to do this and still be decent at whatever schtick you aim for. And, absent any optimization, the game still allows for huge variety and range of table dynamics.

D&D a la 3e survives despite the fact that the cancer really should have killed it years ago. Whether you feel this is a matter to be rectified or celebrated is a matter of taste, and varies hugely.

[Gavinfoxx]

It sounds like it is a bunch of houserules for 3.5e? Is that right?

Why are you trying to reinvent the wheel? There's lots of options out there

I would consider Pathfinder 3.55, Trailblazer 3.60, True20 3.65, D&D with the Frank & K Tomes 3.65, Fantasycraft 3.70, 'Mutants and Dungeons' (both versions) also 3.70, and Legend 3.75, as far as the 'number of things fixed' goes...

For Example:

Legend:
http://www.ruleofcool.com/
http://www.ruleofcool.com/get-the-game/
https://s3.amazonaws.com/det_1/Legend.pdf <-- this is the actual link to it!
http://www.ruleofcool.com/donation-thresholds/ <-- some bonus content
http://dl.dropbox.com/u/47651526/LCGb.html <-- an online character generator, a bit old though, doesn't contain everything or the current version.

Other good things to do is use mutants and masterminds 2e to write up D&D-esque characters, a la:
http://greywulf.net/2011/06/03/mutants-and-dragons-third-edition/

Also, someone is trying to make D&D stuff with mutants and masterminds *3e*:
http://www.giantitp.com/forums/showthread.php?t=279503
and
http://www.atomicthinktank.com/viewtopic.php?p=706712#p706712

Fantasycraft is found here:
http://www.crafty-games.com/node/348

Trailblazer is found here:
http://badaxegames.com/

The Frank & K tomes are here:
https://sites.google.com/site/middendorfproject/frankpdf

True20 is here:
http://true20.com/

Again, if free content is important, LOOK UP LEGEND!! It is, as far as I can tell, superior in all ways to Pathfinder.
http://www.ruleofcool.com/

If you want a D20 game that feels like a better, balanced 3.5e?

[HunterOfJello]

I think that a lot of problems in 3.5 stem from the core books presenting magic options that are far stronger than the designers should have allowed. There was likely not enough playtesting in 3e for playstyles that were significantly altered since 2e which caused a dissonance in the entire game.

Someone obviously thought that druids weren't strong enough with their spells alone and gave them a powerful animal companion and the ability to wildshape multiple times per day for longer and longer durations. Whoever wrote that stuff up obviously didn't understand what was going on in games. (Not to mention the fact that they built a class that has zero reason to ever go into a prestige class in future splat books.)

The 2e idea that "Fighters are strong" was held and brought over to 3e while it simply wasn't true any longer. The mindset of 2e being a controlling factor in the creation of 3e is quite obvious once you wrap your head around it. "Sorcerers are more dangerous than wizards because of their spontaneous casting choices and metamagic application" and "multiclassing is a powerful option that allows powerful melee classes to excel even further" are obvious examples of ideas that were created without proof and appropriate justification from playtesting.

[Lanaya]

Again, if free content is important, LOOK UP LEGEND!! It is, as far as I can tell, superior in all ways to Pathfinder.
http://www.ruleofcool.com/

I love Legend, but it's not a better D&D/PF. It's a radically different game, despite having certain themes and basic game mecanics in common with them. One of 3.5's most distinct features is the sheer wealth of stuff in the game and crazy things you can consequently do with the system. Legend on the other hand is a very stripped down game. Many iconic character concepts have been vetoed because they're too difficult to balance (summoners are the main one), the rules for anything that isn't combat are essentially nonexistent, everything is highly abstracted unlike 3.5, which at least attempts to simulate a real, breathing world rather than being a really good arena battle system.

Perhaps most importantly, the emphasis has been placed on refluffing over new options. In Legend, if you want to play a cool, calm and collected warrior type, you might grab Smiting and fluff it as learning your foe's weaknesses as you battle, Battle's Tempering for generally being a good warrior who doesn't die a lot and Discipline of the Serpent, adding properties to your weapon rather than using your unarmed strike, using the fancy combat manoeuvers to represent using your intelligence score in combat. If you were to make that same character in D&D 3.5, it's a warblade with certain manoeuvers. Or a fighter with certain feats. Or a swashbuckler. Or (heaven forbid) a DMG duellist. Rather than assemble a bunch of different appropriate abilities, refluff as you like and make it work, you pick one of the fifty million premade character concepts that suits you and just use that. Legend is a content-light system that encourages using a small amount of content in a variety of creative ways, 3.5 is a tremendously bloated system with separate classes, feats, skills, magic items and whatever else for every concept you could possibly imagine. Which is great, and I love 3.5 just as much as I do Legend, but despite seeming superficially similar they are wildly different games.