RPG mechanics you are done with. And why.

[mean_liar]

It depends on how you want that to manifest in game. For example, on a bell curve system a +X doesn't translate into a constant percent increase of success.

Take a 2d6 system, and say that it after accounting for all the skill interactions it takes a '7' or more to hit someone in combat in a featureless plane. That's a 58% chance of success. A +1 to your roll, say for striking a distracted opponent, means a 72% chance of success. Say that another +1 is for the opponent being on unstable ground: now it's an 83% chance of success.

So the first +1 added +14% and the second +1 added +11%. Now consider a lesser-skilled warrior would have needed (say) a base roll of 9 to hit on a featureless plane (28% chance of success), meaning that the first +1 would have added +14% and the second +1 added +17% to the success chance.

A desire to use uneven weighting means that you don't just use percentages. There are other minor differences in that increasing Dice Pools (say, from Skill increases) still gives you marginal chances of failure (all those dice could still come up 1s) as opposed to a straight +X percentage system (which would have to just have a roll of 00% an insta-fail to approximate this).

I'm not really sure of the advantage of one over the other, other than to say that maybe sometimes you want marginal rates of return on +/- modifiers for extreme cases: the hyper-skilled get less off of a +1 than the moderately- or lower-skilled, and likewise the lower-skilled are less affected absolutely by -1 than the higher-skilled (unless they're off the RNG).

[Murtak]

Exactly. Quite often a bell curve is exactly what we are looking for - we want to see a distinct difference between a level 6 and a level 7 swordsman, but we want the level 1 swordsman to still be a theoretical threat. That is impossible with linear percentage modifiers.

[violence in the media]

Ok, we want a bell curve. But is there a difference between saying you need a 7+ on 2d6 and you need 58 or less on 1d100? Couldn't you implement a variable bonus on a d100 that depended on where your success rate started? Say flanking provided a +20% bonus to hit if your to hit was 35% or below, +15% for 36-55%, and +10% above that. Sure, you'd get weird edge cases where a 33% accuracy opponent is marginally more dangerous on a flank than a 36% accuracy opponent, but the evaluation of action success rates would be much clearer to the GM and players.

Using your 2d6 example in the context of wargames, I've seen people who look at 7+ and act as if it's practically an auto-hit. Sure, they're wrong. But I think it would be clearer to them if they were dicing against 58%.

Using another example from a 7th Sea game I played in years ago, the GM would toss out a TN of 35 on a 4k4 roll without realizing whether or not that was a proper difficulty. And this difficulty assignment would be partially influenced by the person who rolled a 73 (exploding dice BS) on a 6k3 roll for a similar task. If someone pointed out that he just assigned a success rate of >1% he'd be surprised and change it, as he genuinely didn't intend to make it that tough most of the time.

[Josh_Kablack]

Variable dicepools with variable TNs is not merely linear vs bell curve probability.

It's various exponential curves crisscrossing each other in ways reminiscent of systems of differential equations.

In a linear probability system, there are a number of discreet points equal to the size of the die used (20 in d20) and a +1 means the same for everyone (except maybe at the extreme ends where a 1 on the die auto misses or a 20 auto-hits). So rational players will stack as many as they can on to their most common tactic and then try to use that tactic as often as possible

In a bell-curve probability system, there are a number of discreet points equal to the possible sums of all the dice (15 in HERO or other 3d6 systems) and a +1 means more the closer you are to the center of the curve. So rational players will try to grab enough +1s to be somewhat above average but then move on to grabbing +1s for other schticks to also be above average.

In a variable-dicepool, variable-TN system, you have a number of discreet points equal to the product of the potential dice times the number of potential TNs (possibly multiplied by the number of successes needed). This gives a massive matrix of potential probabilities, which all but precludes prefiguring a character's odds of success before the time of task resolution. So a rational player will whip out a TI-94 in the middle of combat and plot the inflection points of overlapping exponential curves to determine if it's better to increase their dicepool or reduce their TN. Actually that's a lie, a truly rational player will give up and either go play something else or accept their GM's version of magical tea party.

[mean_liar]

I happen to like straight d% systems, btw. I wouldn't agree that they're necessarily worse than a bell curve system.

[Juton]

A rational player will just print out a big chart before the game.

[Murtak]

The issue with a variable-dicepool, variable-TN system is not players having to weight dicepool vs TN penalties. A sane system would not let you choose. The issue is, that there is no (easy) way to generate a smooth difficulty curve, which in turn leads to hard to calculate probabilities.

Imagine for a second that increasing the TN on a d6 by 1 would always cut your chance of success for a single die roll in half. The dicepool is determined by character ability, the difficulty is determined by circumstances (in a hurry, difficulty of task, distracted, etc). No character need ever recalculate his dicepool. No one ever has a dicepool of zero. Everyone can succeed at every task (but more skilled characters still have the advantage). Basically you get all the benefits of a fixed TN, variable dicepool system and you get rid of the ugly corner cases. Of course no one has figured out a way to get a smooth difficulty curve, and so at best you can trade your ugly low dicepool scenarios for utter lunacy at all levels of play - which is lunacy of course. But the basic idea is fine. It just doesn't work without computers.

Edit: Also it produces a much flatter probability curve than a fixed TN variable threshold system, which is fine for many games, but definitely not for Shadowrun or WoD.

That said here is another thing I hope I will never see in a RPG again: bullshit bookkeeping
That is, bookkeeping for no good reason. Tiny bonuses are bullshit. Tiny conditional bonuses are bullshit squared. Tiny conditional temporary bonuses are bullshit cubed, and should be fed to the responsible designer until he pukes. DnD, I am looking at you. You can't tell which character has dodge, and which has not, even after analyzing 10 consecutive fights. Why do I have to keep track of it? The same goes for weapon focus, and at least I get that bonus for every attack I make. Single-digit temporary hitpoint bonuses at level 10? Attributes that give a whopping half-a-modifier per point? Why am I doing this again? Why do we stack 4 +1 bonuses and keep track of them separately just finally see a meaningful difference between Conan and Super-Conan? Can't we just have a single +4 and keep track of that?

[Lago PARANOIA]

One thing I don't care for in bell-curve systems are the huge amounts of extra dice that you often have to roll. Which should be a gripe in of itself. If a task requires more than 10 dice to be rolled (and even that's really pushing it) then that's just out of control. I listen to after-action reports of Exalted and WoD and I'm just completely boggled at why or even how people can game like that.

Linear probability systems have a lot of problems to them, especially when using them in games like D&D where the power curve is exponential, but I don't think anyone should undercut the advantage of just rolling one die per task.

[NineInchNall]

dammit
post got eaten

Granular Combat!
When fighting 30 kobolds I don't want to have to handle attacks and saves for and against all those bastards.

Seriously. It's tedious. It's boring. Gimme the Weapons Of The Gods way of handling mook squads any day.

[RandomCasualty2]

Fixed Character Roles: 4E really shows us why it's a bad idea to have one character whose always the "damage guy", another guy who just tanks and another guy who is supposed to be a healer. Not necessarily because any of these are bad, but because this sorta thing should change based on what you're fighting. Having a game where you do the same thing all the time is boring.

Universal Powers: Another criticism of 4E, this is the fact that you have powers that work all the time and happen to be the best move you can make all the time, which just turns you into a spammer and removes any strategy.

Individual XP/ Different leveled PCs/ XP costs: The concept that your party is of different levels and some PCs get more powerful characters than others is bullshit. Similarly paying with things with your XP is a bad idea.

LA: Being lower level in a game where level is supposed to determine power. Bad idea.

'I don't have to think' abilities: These are any abilities that basically lets the player avoid thinking about a problem and just toss a skill or spell at it. In many games this includes the diplomacy skill, as well as psionic powers like hypercognition.

[Crissa]

Healing and buffing do have tactics.

But at that point, you're not playing the same game.

In Guild Wars and Warcraft and many healing-or-damage games, all you're playing is whack-a-mole while other people are playing a keyboard based guitar-hero. When the damage guy screws up, his difficulty goes up linearly; but suddenly the healer guy's difficulty has gone up exponentially - from the actions of someone else!

It's a dumb way to pay a game together. Especially when challenges are beat by Damage and never by Support. Support can't go out and do things without Damage, but the reverse is often true.

-Crissa

[Lokathor]

Individual XP/ Different leveled PCs/ XP costs: The concept that your party is of different levels and some PCs get more powerful characters than others is bullshit. Similarly paying with things with your XP is a bad idea.

This. Hardcore.

[Blicero]

Individual XP/ Different leveled PCs/ XP costs: The concept that your party is of different levels and some PCs get more powerful characters than others is bullshit. Similarly paying with things with your XP is a bad idea.

In my games, I do quest-based XP that all PC's get equally, but I also give out (much) smaller role-playing awards. I mean, like 150 XP max for 3E, generally more like 50-100. That way, it's still an incentive for not just hacknslashin' through everything, but generally the worst that happens is that one PC might be a level higher for one quest segment, max.

But things like XP costs and losing a level need to die in a fiery fire of flames.

[CatharzGodfoot]

I've pretty much come to the conclusion that XP have no place in a modern RPG, period.

[Maxus]

I've pretty much come to the conclusion that XP have no place in a modern RPG, period.

+1

[Manxome]

Healing and buffing do have tactics.

But at that point, you're not playing the same game.

If by that you mean that the exact stuff you do within the game is different, that is the whole fucking point. I'd be quite offended if playing a Warrior, Monk, and Mesmer were all exactly the same, wouldn't you?

In Guild Wars and Warcraft and many healing-or-damage games, all you're playing is whack-a-mole while other people are playing a keyboard based guitar-hero. When the damage guy screws up, his difficulty goes up linearly; but suddenly the healer guy's difficulty has gone up exponentially - from the actions of someone else!

It's a dumb way to pay a game together.

I have no clue what this means or how you reached this conclusion. I wasn't even aware that difficulty had been quantified, let alone an equation for it derived, and without those things, to say that it increases "linearly" or "exponentially" is unmitigated bullshit. But regardless, it would be a pretty sad excuse for a cooperative game if your tasks were in no way impacted by any of your allies' actions.

If you mean that it quickly becomes hard to keep someone alive when they don't adequately defend themselves, that's probably true, but to conclude that this makes the healer's job harder faster would appear to require the additional leap of saying that if Bob stops protecting himself and ends up dying, that the healer failed in his task and Bob did not--which is such an amazingly stupid assumption that I very much hope I'm barking up the wrong tree.

Different party members don't even typically have different goals in this sort of situation. They all have the shared goal of "win combat" and share in victory or defeat. To say that one person's job is harder than someone else's actually means that the division of labor that has been created by the players is unfair, not anything insightful about the game's mechanics.

Especially when challenges are beat by Damage and never by Support. Support can't go out and do things without Damage, but the reverse is often true.

That's because if you can win just by doing X, then X is definitionally not "support."

Regardless, there's lots of well-established ways to address that, it's questionable whether it's even a problem under most conditions, and I fail to see how it is in any way related to any other points raised in your post.

[Jacob_Orlove]

I've pretty much come to the conclusion that XP have no place in a modern RPG, period.

External rewards have no place in a modern RPG. Players should never "grind" for anything ever.

[Lokathor]

XP that goes into levels, probably not needed. BP or something like it as an alternative to a level based system, it works pretty well if that's what the group wants.

[Akula]

Individual XP/ Different leveled PCs/ XP costs: The concept that your party is of different levels and some PCs get more powerful characters than others is bullshit. Similarly paying with things with your XP is a bad idea.

So behind this.
Story time! One of the guys in my gaming group wanted to run a game, vanilla 3.5, and have the players role 2d20. If you got a 20 on a die your PC became a prodigy and got an extra level (presumably if you got two 20s the PC would be a double prodigy). I instantly hated the game.

[Thymos]

I also agree that we should just do away with xp entirely. It's no better than simply having the DM hand out levels, and in many ways can be worse.

[MGuy]

XP is different from the DM just handing out levels in that you can associate hard rules to it easier than you can give rules to level up when you feel like it. The option of "level the PCs when you feel its the right time" is always going to be there but how would you write solid rules for that? I'd find it much easier to say "When creating an adventure judge the experience given by the average number of encounters PCs are likely to come across and allow that to be the reward." From that I can give example xp values for various encounters (creatures, traps, maybe even situations). If I wanted to write a game with the former (no xp) then all I'd be giving is advice instead of rules.

I do agree that experience should be a group reward and there shouldn't be mechanics in the game that let you fiddle around with your xp.

[Judging__Eagle]

JE, your first one isn't a mechanic.

It's the result of several mechanical problems that have been pointed out.

This is related to the DD and Healbot = fail character concepts.

As well as the: Stat + Class pair-ups, Race + Class pair-ups, issues.

Having classes in 4e, where the "fighter" is only going to be "every fighter you've ever seen in a vidjya game" is a mechanic that I'm done with.

The human fighter needs to be as interesting, and varied, as the Yakfolk Genielord or the 14-year old Aasimar Conduit with "love, love, power". If your game can't allow anything besides "sword and board" for fighters, or "single type of fighting", then that game can GTFO.

If the game doesn't allow for players to make unconventional characters, that fit the setting, then the game is weak.

[CatharzGodfoot]

In a game like D&D, advancement should come at an appropriately dramatic moment, and (as per 2e rules) at a break in the gaming session. That's actually really easy to write rules for, and might be helpful to new DMs that haven quite gotten a handle on the OOC logistics yet:

You should plan for the heroes to advance at a suitably dramatic moment (the defeat of a major enemy and the completion of a mission are the most common). You should also plan to end you gaming session at this point.

The players will need time to advance their characters, so setting aside a bit at the end of the session for this is a good idea. The players will also be more likely to remember what their characters are up to: 'We had just defeated the shadow dragon warlock' is easier to remember than 'We were searching the orcs' 8'x5' pantry'. Think of Stephen Colbert's interviewing style.

In a BP system, you can do the same damn' thing. The only difference is that the quanta of advancement are smaller, and sometimes you have the choice of not advancing at all to get greater power later (which is bullshit).

[Thymos]

You could also just give a default amount of sessions to level up after.

Personally I just think the DM should level players up to match what needs to happen for his adventure.

[NineInchNall]

Personally I just think the DM should level players up to match what needs to happen for his adventure.

This sums up my feelings on the matter.