If there was one thing about 3rdE that you could change, what would it be?

[Havoc]

Four classes only, no PrC, more feats to customize them.
Some classes and PrC have 'flavour' in it (Paladins, monks) while other are more building blocks to create the concept you like. I do not like that dichotomy at all.

More streamlining:
hit points/fortitude save
AC/reflex save
BAB=skill...

I do think that these have been on the design board but have been dismissed to avoid alienating D&D grognards, though.

 

[Aaron2]

What's I'd do

The main thing I'd do is rewrite the creating character section. Remove the long laundry list of lame special abilities of the different races, have each character start with around 20+x skill points (where x is the skill points of their class). The characters can choose the starting skills and feats from a list of racial abilities. This would eliminate the rogue/fighter != fighter/rogue problem and add variety to elven/dwarven characters.
I'd stretch out the abililties that classes get early on (for ex: fighters would get heavy armor at 2nd level, monk would add Wis to AC at 2nd level) to remove the insentive to just take one level in a class. Then I'd remove the 20% xp penalty and that favored class nonsense.

Add clear ECL rules to the PH.


Aaron

 

[apsuman]

About half of this thread can be boiled down to: "Have your DM use rule zero"

As for the one thing I would change... The D20 mechanic.

The fact that there is a 5% chance that you will automatically succeed under many circumstances just seems too often.

I would suggest a mechanic that was bart of Mayfair's DC heroes game, a 2D10 mechanic.

Roll 2d10 and add the result. Double ones was automatic failure. Rolling any other double allowed you to roll again if you wanted adding the original roll in as well. You could roll multiple doubles getting some really high results. If in your roll you ever roll double ones you automatically fail.

Most results would be in the range 2-20 but there is a theoretically possiblity to roll infinitely large. The results would generally normally distribute around a mean of 11. If you can get your opposed DC more than 11 higher than than your opponent then the roll is generally in your favor more than with the straight d20.

This way critical hits could be generated by adding multiples from your attack roll. That is you roll one double you add one multiple to the damage, from normal to 2X.

g!

 

[Numion]

Most results would be in the range 2-20 but there is a theoretically possiblity to roll infinitely large.

Or would the possibility of rolling infinitely large be zero? Sorry, just a nit I had to pick.

But anyways, if you you use this 2d10 method, wouldn't the dice rolling matter even less at higher levels? Because the variance of 2d10 is much less than 1d20. Even with the current 1d20 roll the dice roll is secondary in importance to the modifiers, which can be in the 40+ range even before epic levels. If thats what you were shooting for, then fine. Just a point.

 

[Surfal]

Only one?

Sheesh, there's a bunch really. Number one on the list would probably be to rebalance the spells in the core rules, but re-doing the class system is a close second. I'm not sure if I'd go all the way down to four core classes, but I'd definitely dump the ranger, paladin, bard and monk. Probably the druid and barbarian too, so we'd have five left.

 

[mkletch]

Or would the possibility of rolling infinitely large be zero? Sorry, just a nit I had to pick.

But anyways, if you you use this 2d10 method, wouldn't the dice rolling matter even less at higher levels? Because the variance of 2d10 is much less than 1d20. Even with the current 1d20 roll the dice roll is secondary in importance to the modifiers, which can be in the 40+ range even before epic levels. If thats what you were shooting for, then fine. Just a point.

I've seen variants where people also use 4d6-4. This leads to an even tighter bell curve, and anything more than 4 dice is effectively a small variance plus a big bonus (20d6 is nearly the same as 2d4+65).

I took the opposite approach in 2E, and I'm thinking of applying it in my next campaign. Bonuses are backwards converted into dice, and handfulls of dice are backwards converted into either a few dice of large value or two dice. 20d6 is approximately the same as 6d20, though a little short on average and minimum possible value. So, 6d20+4.

The other option would be to roll 1d6x19 + 1d6. So, yes, roll two dice, one is multiplied by 19 and added to the other. Much wider variance, though not a bell curve. If that is important, roll (2d6)x9 + 2d6.

Large numbers of dice are effectively a noisy fixed value, and large fixed values, especially bonuses, are boring. If you look at the Sultans of Smack thread, it's whole point is maximizing fixed bonuses and critical multipliers. Boring; there is no need to roll anything other than an attack roll - the damage is nearly a fixed quantity.

Other than spell balance and how it relates to the magic item system, this would be my second choice. High-level combat against soft (low AC and/or low SR) opponents can be effectively resolved in a diceless manner. THAT is not D&D to me.

-Fletch!

 

[apsuman]

Or would the possibility of rolling infinitely large be zero? Sorry, just a nit I had to pick.

But anyways, if you you use this 2d10 method, wouldn't the dice rolling matter even less at higher levels? Because the variance of 2d10 is much less than 1d20. Even with the current 1d20 roll the dice roll is secondary in importance to the modifiers, which can be in the 40+ range even before epic levels. If thats what you were shooting for, then fine. Just a point.

Thanks for the question.

Right now at high levels having +40 to a skill is already making the d20 useless, imho. And, having a really high skill number makes it impossible for your average guy to beat you in opposed checks (say your listen +6 versus the rogue's hide +30). However, the possibility of rolling a double allows you to skew your results higher and gives you at least the possiblity of winning.

There are nine rolls (double 2,3,4,5,6,7,8,9, and 10) or 9% of your rolls will allow you to roll again. I consider this 9% bonus more fun for the players than a simple 5% flat out success possibility of rolling a 20.

My idea is still pretty new in my mind you you guys are getting the rought draft in this post so I will be surprized (or not) with you as I think out loud.

By comparison. Disregard doubles for now (make the math easier)

If you have------Under d20------Under 2d10
to roll---------------You win-----------You win
===============================
11-------------------50%-------------55%
12-------------------45%-------------45%
13-------------------40%-------------36%
14-------------------35%-------------28%
15-------------------30%-------------21%
16-------------------25%-------------15%
17-------------------20%-------------10%
18-------------------15%--------------6%
19-------------------10%--------------3%
20--------------------5%--------------1%

So, under 2d10 having to beat a DC by more than 12 is HARDER, having to beat it by 16 or more is much harder.

I am still thinking about this, and will add more later...

...developing...


g!

 

[RobNJ]

I would start out by disallowing idiots to play the game. You know. The kind of people who start threads designed to be a cover for, "God, 3rd Edition sucks ass!"

That said, here're the changes I would make:

Get rid of alignment. It is extremely lame. You can even have a struggle against good and evil without a lamebrained, hamfisted mechanic like alignment. Witness that no other new d20 ruleset has kept alignment. This isn't necessarily a ploy to make the morality of D&D grayer, it's just that the alignment system does a very poor job of describing motivation, is usually taken as a proscriptive rather than a descriptive mechanic, and is too clumsy to describe behavior accurately.

Bump up the levels of magic missile, haste, shield, and other broken spells that were just kept they way they are to forstall whiney purists (who are going to whine anyway).

Alternatively, junk the Vancian magic system and put in something a little less silly.

Remove barbarian, ranger and paladin (and maybe a few others) as core classes and make them prestige classes.

Include rules and descriptions on how to create core classes and core races, with the specific design elements that were in mind when these choices were made.

I think there should be extensive discussion on how to set DCs and essays that address in depth DC creep and how to make things challenging without appearing too conveniently tailored.

Most of the suggestions here seem to want to go backward. I advocate going forward. Making changes to better the game rather than undoing changes to preserve some warm and fuzzy feeling of nostalgia. Here's my rule: Don't change a single thing back. Only change new things forward.

 

[incognito]

chiming in!

Skill Points if you are coing to give characters skills like Profession, and Craft in a combat heavy game, give the players more skill points to flesh out their character.

"ShadowRun" does this by giving characters two typed of skill points - some poitns are spent on comabt type stuff, the otehr are spent on language, social skills, etc.

Playtesting It's not that the splat books (heck, even the core books) don't have a lot of good ideas, it's just a shame to see spells like Polymorph get constantly reworded, and spells like Harm be constantly debated. Then there is CR playtesting, weapon bonus pricing. etc, etc. It all really comes down to thoroughly playtesting the product.

 

[Wolvorine]

Re: Um, yeah, I already understand all that

The point remains that someone might, under certain circumstances, want to target a particular part of someone's body. The arrow through a 10th level fighter's neck fired by an orc is a bit tenuous because it would almost never happen. Most orcs would be woefully unskilled at making such an attack (I would assess a -12 penalty for that shot, providing the fighter was moving), and they'd know it. However, a 12th level halfling rogue ought to have some chance of pulling such an attack off. It would still be a difficult thing to do (especially if the fighter is wearing a great helm), but it shouldn't be impossible. For this particular case, I'd stipulate that a nat 20 doesn't automatically hit.

I tend to agree with this idea myself. Although in my own opinion these kinds of situations should be resolved with opposed checks of some sort. The 10th level Fighter should, by default of being a 10th level fighter, have a good chance to avoid it - the orc should have a tiny chance of pulling it off - and the 12 level halfling rogue a fairly good chance of pulling it off. Not by virtue of their race or class or level, but because all those things put together grant them certain modifiers to their opposed checks.
Then again, I like hit locations (with reasonable effects for those locations), too.