Best Skill Systems?

the dark eye, by far, though you'll need to get used to it.
> 8 Attributes
> Skillchecks are rolled versus any combination of 3 attributes
> Skill-Level are points for dice adjustment

so you roll 3d20 in order and try to roll below or equal to the Attribute. Whatever you are rolling over your Attributes you can make up for with Skillvalue.

Quality of Success is determined by leftover skillpoints.

d20 is not any more swingy than a d% system... or any single dice system really.

Single dice are not created equally, adding a d4 to a stat+skill gives a wildly different outcome than a d1000.

In general I dislike single die systems, especially when the die used is bigger than the skill+stat. Lower die and 3d6 type systems are much more predictable, and arguably realistic.(people tend to perform as well as they usually do, they don't swing from botch to crit like a yo-yo)

teilzeithelden.de/2013/02/23/der-stochastische-albtraum-dsa-teil-ii-die-3w20-probe/

just look at these distributions
wtf

>In general I dislike single die systems, especially when the die used is bigger than the skill+stat.

The skill+stat does not matter in itself, what matters is in comparison to the DC. Dice size really only matters for granularity.

this
> dice size -> Granularity
> dice number -> spread of results
that's why the dark eyes 3d20 in order with points to adjust rolls and determine quality is good.

When the variability from the die is greater than the variability from skill+stat, then you've got yourself a broken system.

take 10 helps

This, holy shit I cannot even begin to tell you how retarded the game got when we used skills in 5E.

Okay, listen here you stupid faggot. Bonuses don't matter in a fucking vacuum. They only matter in relation to the DCs.

If your average bonus is +5 and your DC for an average action is 15, you'll be over that DC 50% of the time with a d20. If your average bonus is +15 and the average DC is 25, it's the same. If your bonus is 105, and the DC is 115, same again.

The bonus only matters in relation to what you are rolling against.

5e's skills scale badly (imo, it's up to taste) but it really doesn't matter that the bonuses are low when so are the DCs.

But the DCs aren't that low, especially not when you're making opposed tests or the like.

The average dc is 10 (55%). If you have stat+training in the skill, that's at least a +5. You should succeed 80% of the time. With either only (+2-3), 65-70%.

In contests you got a 50-50% chance, and grappling is basically broken because players get a bonus to it much more often than monsters.

Does anyone have a build for Grapple Rogue?

My best guess is to load up on strength and grab Athletics, to grant myself advantage against those I grapple. Anything else? Just go full strength, and ignore stealth?

it matters in the ability to compare one character to other characters, not in the context of a single DC.

larger bonuses compared to dice means the dice are less the cause of success than skill level, which means the difference in potential between a skilled and unskilled character is larger.

>What rpg has the best skill /task resolution system?
My homebrew.
>what makes it the best?
Light on the rules, simple and fun to use.
>how does it work?
Roll 2d6 under ([Attribute]+[Background]) to succeed at an action, apply modifiers if necessary.

Star wars narrative dice.

It's not a matter of "my success chance is too high/low", it's a matter of wanting a larger difference between your capabilities and that of a peasant, and wanting a larger amount of specialization amongst the group.

some also want skill performance (DC achieved) to become more and more predictable as skill level increases.

I loved Star Wars Saga's skill system.

This is what I meant by >5e's skills scale badly

But not every game is going to have contested skill rolls, in which case it doesn't really matter.

You could also just play a game where the differences between character capabilities are small, it just doesn't fit 5e since you go from killing goblins to killing Asmodeus and Tiamat.

Really, I'm just taking issue with stating it as an absolute rule. It does not work in 5e, but it can work easily in less heroic games.

>it can work easily in less heroic games
If you want to heavily incentivize building jack-of-all-trades characters, sure. If you intend for people to be specialists, which would be more fitting for gritty and realistic campaigns, not really.

Then for more heroic games, you'd want a character to be substantially better than a peasant at things, and so it fails again.

Maybe there's some ultra-specific case where it works, but I'm not seeing it.

>If you intend for people to be specialists, which would be more fitting for gritty and realistic campaigns, not really.

The difference of +5-10 and 0 is still pretty big. Do keep in mind that your non-trained skills don't advance at all. Also, if you spread expertise around a bit, you got a two-step system instead of just trained/untrained.

Categorically saying that it just doesn't work because it doesn't fit the high level heroics of 5e is silly.

It's more than opposed checks.

It's also wanting a larger amount of specialization, even at the low levels, such that grok the very smart int 18 barbarian is much less able to arcana than Thule the int 18 wizard because he has training in arcana, not just the difference of a +2.

I think I'd be more satisfied with 5es skill scaling if proficiency started at a +6 rather than ending there. as is, it makes everyone way too much of a generalist, particularly for levels 1-10.

I loved SR1-3's skill system so much - it felt intuitive unless you managed to go over TN 6, the degree of success system played into the dice pool system perfectly, and it had you throwing significantly less dice than SR4 did at equal levels of play unless you were burning limited resources all on one roll.

It was nearly everything else that sucked.

I was defending the modifier range more than the exact mechanics behind it.

I'd also say that Grok with 18 int is a fucking genius, and "trained" being only 2 means that in game world, his natural talent is just worth twice as much as the training you got before level 1.

This is also not an issue because 5e is not meant to rolled for stats, so an 18 int barb already does not make sense.

I'd also posit that in other editions the wizard would be a mere 4-5 points better off with the same stats, which still isn't that bad, so this seems to be just the result of low level D&D's more tight number ranges having to accommodate both the skill and the stat bonus.

Plain doubling it works out better imo, but you should really give some sort of passive scaling because poor monsters are already bullied by grappling rogues/wizards targeting bad saves.

But the point there is twofold:

1. wanting training the be able to matter as much or more than natural talent, from level 1.

and 2.
wanting the starting difference between specialized and peasant to be at least +50%, such that their 50% tasks are now things you can't possibly fail at, your 50% tasks are such that they couldn't possibly succeed at.

>double it.
that results in much larger growth at the top end. that may/may not be desirable. I mostly want it larger at the lower levels.

Maybe giving a flat 25 "skill points" at level 1, which can be assigned to any skill, and which add to proficiency in those skills is a better approach, and put a maximum skillpoints bonus to a single skill at +5, and no more than 3 skills can have a a +4 or better. skillpoints would not apply when using a skill for an attack roll or some such vs a save or ac.

>1. wanting training the be able to matter as much or more than natural talent, from level 1.

If you play with arrays (which you should), it's going to be 3 from stat, 2 from skill. That's already pretty close, I think.

>wanting the starting difference between specialized and peasant to be at least +50%, such that their 50% tasks are now things you can't possibly fail at, your 50% tasks are such that they couldn't possibly succeed at.

That's an extremely heroic game imo, which is why I said that for a more grounded game the 0-10 bonuses on a d20 is fine, it's just the heroics you have problem with.

>that results in much larger growth at the top end. that may/may not be desirable. I mostly want it larger at the lower levels.

You'll have to rebalance either way, as proficiency is also for attacks and saving throws. If you double, you can introduce a passive scaling of half-of-prof and the math checks out.

>more specialized=heroic
a trained npc at level 1 (like a blacksmith) could be just as competent. I'm wouldn't be looking to change the difference between competent NPCs and pcs, merely between trained and untrained characters.

>have to rebalance either way
the 25 skill points at l1 solution seems to handle it fairly nicely.

I agree with these dubs.

To be honest as far as im concerned D&D would be even better played with 1d10 rather than 1d20.

I just dont think someone with 10 Str should be able to win at arm wrestling with a powerfully build 18 Str guy or amazon chick. But it only provides a 40% increase in chance (20% for the stronger dude, and another 20% because the other dude is rolling against him and is unable to roll above 20).

In fact I would even say that Skill Ranks are more important than Attributes. Trained technique in doing things better than your natural tendency to do them, this is why we have schools.

Real life is somewhat unpredictable but if you know what you're doing and you're dealing with things on a comparable level then you should be quite Reliable for doing them.

> Alternate Way of doing d20 shit

Then again I think maybe the better way would be sorta like how GURPS scrunches the attribute numbers.

Like instead of Strength 18 for d20 all of the attributes could be compressed so that 16 is the new 18 or such (75% feels like a good adjustment).

- - - Now you just ADD your score to your roll and the difficulties are increased depending on whatever is average to roll.
- - - Modifiers can be Attribute minus 10 rather than 1:2 ratio.

> but senpai The Holy EIGHTEEN is golden to d&d
> wait I thought the d20 was?
> well it sure as fuck doesnt seem all that holy since 4E

I think that like Feats your skills in a game should somehow make you capable of doing more things than just throwing a higher bonus at them.

Minimum skill ranks, or DCs that decrease by percentages when your skills get high enough.

That way with lower skills or untrained at skills the DCs are meaty as hell or it just says you cant succeed. But when you get lots of skill ranks the DCs actually decrease to "meet you halfway" so that what would've been impossible before becomes possible.

But without adding all those extra bonuses inbetween that make you so much better at other things.

> skill should mean not just being better at something but also knowing more about it too
> so that electrical engineering allows you to take apart a radio, but no matter how smart you are you're not going to be able to design a CPU processor
> and even with a fuckton of skill ranks at electrical engineering you'll need to be smart as hell to pull it off too.

Also the Aid Another crap is too easy to exploit. Helpers need to be of comparable skill/bonus (within 6 points modifier difference).

And every Doubling of help should be required for every +2.

Might need the help of an entire civilization to get up into the triple digits.

> goku spirit bombs

>50495688

>What rpg has the best skill /task resolution system?

I'm a fan of Savage worlds

>what makes it the best?

Easy and fast-flowing with a very interesting probability distribution.

>how does it work?

Roll 1d6 and 1dX (x: skill or characteristic) against a set difficulty (usually 4).

Both dices explode, take the highest dice.

Exceeding difficulty by 4 (or a multiple) give exceptional success.

If both dice end up as 1, fumble.

>>best
>shadowrun.

>>why and how?
>active skills and background skills separated.
>dice pool of attribute+skill.
>good difference between skilled and unskilled characters.
>simple resolution.
>degrees of success.
>skill rolls become less randomized as skill level increases.

Lady Blackbird is pretty interesting.

Your character is made of a few traits (general areas of skill) such as "soldier" or "acrobat" and several tags per trait (specializations) such as "swordsman" or "orc slayer" or "fights dirty".

Resolutions is based on a dice pool and getting X successes as determined by the GM. You get 1 die by default, +1 if any trait applies, and +1 per tag under the trait that applies. You also have a generic pool of bonus dice that you can apply to situations as you see fit, which regenerates by 1 die whenever you fail a check. 4+ on a d6 counts as a success (so basically 50% chance per die thrown). You can also aid another by giving them one of your bonus dice.

It's fairly deep in that your character may get to apply some or all of his tags against a particular check. If you run into something you've built your character towards, you might get a whole fist full of dice to throw, but something only semi-related might only get a few tags and then you're using half your pool instead. But that makes sense right? If you have a bunch of indirect knowledge, you're better off than a layman, but nowhere near a specialist. Counting the tags might get annoying though.

>Okay fellas, you come across a door.
>Cool, can my Rogue pick the lock?
>Sure, just give me a DEX check
>...I rolled a 15 and my DEX is an 16
>Cool, you manage to unlock the door into the next room.

How is this awful?

>What rpg has the best skill /task resolution system?
(n)WoD

>what makes it the best?
Simple and intuitive and allows you to quickly resolve rolls without having to worry about making a lot of calculations or rolling an excess number of dice just to resolve a simple action.

>how does it work?
You take Xd10 and roll, with X being an attribute's dots + a skill's dots + Equipment/Modifiers.

An 8, 9, or 10 is a success but if you roll a 10, you can roll again. If you get 5+ successes, it's an exceptional success but as long as you get as least 1 success, the action goes off without a hitch.

If your roll is 0 or less, you roll a chance die, which means that you only roll 1d10 and only succeed if you roll a 10 on the die. If you get a 1, it's an exceptional failure, which means that the result is worse than usual.

The problem with d20 is that the die itself is not only a single die but it's a die that is more than likely bigger than the modifiers attached to it.

I could go in with a dude who doesn't even have a single point invested in an untrained skill yet still have a chance of pulling off an action that's equivalent to a dude who spent months/years building up the same skill until he got a +20 to his roll.

Then you add in critical rolls to the mix and you end up with a system where a level 20 swordsman can still be hit by a 1/8 CR goblin 5% of the time while still somehow missing because he rolled a 1 and decided to stab himself in the foot.

I mean, it's great for a three stooges campaign but not for a game where you're supposed to be someone who goes from 0 to hero, fighting off dragons and liches and slimes

This is fine so long as your entire world consists of dank, unfurnished dungeons with an excessive number of locked doors. The moment you have to do pretty much anything else you're sort of limited.

Not that this is necessarily a bad thing, each system is designed for different kinds of games, you just can't really say that it competes against more skill-oriented games like CoC or Traveller.

>This is fine so long as your entire world consists of dank, unfurnished dungeons with an excessive number of locked doors. The moment you have to do pretty much anything else you're sort of limited.
The only limitation in such a system is based on what you, as a player, can come up with for your stats.

Each stat actually has a wide array of bullshit that they can do, it's just most people I find have trouble utilizing obvious solutions unless it's given to them by the book.

The sense of progression that many d20 systems goes for lies in the fact that the guy who has a +20 can reach a roll of 40, while the untrained bubkis can only reach a 20. Generally anything that needs training will need you to reach above the normal limit, otherwise you're just kind of "good at it." You're also fine to remove critical roll and fumble rules if you want your RPG to be more insistent on the difference between power levels, but to be honest it's not terribly realistic if you have absolutely no chance of stabbing someone just because they're good with a sword. Would it be hard? Of course, perhaps the chance ought to be lower than 1/20 (which is why the old rule for confirming critical with a second roll is used) but in most cases luck is something that is used to keep tabletops both fair and unpredictable.

Or just play whatever you want man, it's all up to taste after all.

Let's be honest with ourselves here, how many situations does the average PC come across where he needs to roll a 40 to succeed?

Also, if you can be taken out by a dude who is fresh out of the academy, even after sinking in years to master a skill, why even bother mastering a skill when your luck is the most important factor in every situation imaginable?

The way it stands, I could roll a 20 and produce a similar effect as the dude who has a +20 in the same skill, meaning that 5% of the time, I can perform an action that's magnitudes more than the general population just because I got lucky and rolled a 20 on the die.

If RNG has that much power then there's no reason to ever invest in anything else, especially when the resident mage is going to kick the shit out of the dude who has a +20 in a skill anyways since he prepared the right spell and BTFO'd the challenge standing before them anyways.

Lastly, RNG should only there to act as a placebo for your players.

7th Sea 1st edition was fun. Despite the system being a mess in quite a few places.

>"You are surrounded by a bunch of henchmen! What do you do?"
>Player's character was Nordic Paul Bunyan
>"I whirl my axe and attempt to hit everyone."
>Raises difficulty equal to the number of targets
>Exploding dice happen
>Hits everyone and finishes the fight
>Group's faces

I wish 2nd Edition wasn't so Fluff-heavy and crunch-superlight. I miss tossing huge amounts of d6s.

3d6, roll under your discrete skill/ability level. Just feels nice in play, straight forward and "tight."

I agree with everything in this pos...

>rng should only be a placebo for the players
absolutely not. the fact that the rng matters is what makes it a game. the rng needs to matter, but too much rng is bad.

The over/under for GURPS shilling is 5 posts. Anyone who bet the under, please pick up your winnings before close of business.

5? I didn't know you were in the business of giving away money.

I'm using a system like this in my homebrew rules. care to go into something more depth on it?

> Just finished our first play session
>kinda janky
> PCs were kinda op

>What rpg has the best skill /task resolution system?
Burning Wheel
>what makes it the best?
More natural-feeling skill progression and more fun character building
>how does it work?
Skills are gained by choosing lifepaths that correspond to the character's backstory, giving them skills based on what they actually supposedly did before the campaign. Skills are increased through successes on tests involving that skill. In addition, you can have characters practice skills to increase them or have characters teach skills to other PCs, leading to fun interactions.

It's not perfect but Burning Wheel had a really great skill system.

Because it means that every character in a class is exactly the goddamn same

Not even that guy, but you're still not explaining how that is (or why you think it's fine when applied only to picking locks).
What's your actual criticism?

it's not awful, like the guy says, but it's just good only for what it does

Imagine a comparable game with a fighter class. You can't distinguish between the skills of soldiers, generals and barbarians under this method, who would have different literacy, tactical acumen and so on. Each fighter would be as good at every skill they are assumed to have.

4 attributes - Might (Str/Con), Grace (Agi/Dex), Wits (Per/Rea) and Spirit (Cha/Will).
Backgrounds are basically jobs. So, for example, Thief 1 gives you a +1 advantage on every thing a Thief would reasonably do.

Both attributes' and backgrounds' ranks range from 1 to 6. 2 is average human. 5 is human limit. 6 is demigod level, impossible to achieve by a human without shit like augmentations or magic or whatever.

Upgrading attributes/backgrounds is done via point-buy system, points are awarded either once per session, or once per completing a major goal.

Health system is basically ticking off boxes and applying effects of those boxes (such as injury penalties etc.), until you run out of them at which point you are dead.

That's pretty must the gist of it.

But they'd have different stats, so they'd actually be different?

I mean, this is from 3d6 down the line times, there's very little chance that characters even from the same class will have the same skill distribution.

Also, the kicker was that there's still a difference between trained and untrained; untrained use of a skill is rolled under half of your stat instead of full.

The RNG shouldn't matter more than the thought that the player put into the action being performed.

If it does then I'd have to question whether I was playing a tabletop game or a P&P version of diablo.

I've considered mixing this with stuff like Lone Wolf's Kai Disciplines, in honor of Joe Dever.
Instead of classes, you can pick from a list of disciplines. The issue would be keeping disciplines in a similar tier, or restricting the number of any given disciplines of a tier a character could have.

i personally love the 40krpg and CoC d100. simple enough for anyone to understand, but still able to be used for all manner of specific things

>The way it stands, I could roll a 20 and produce a similar effect as the dude who has a +20 in the same skill, meaning that 5% of the time, I can perform an action that's magnitudes more than the general population just because I got lucky and rolled a 20 on the die.

This is false. If the guy had a +10 and you rolled a 20, you'd produce the average of his roll. A +20 literally means that someone without a bonus can't match you, if you are also rolling (since your minimum roll will be 21). With +20, your worst is still literally impossible to others.

Attributes vary too widely, especially if you're randomly generating your stats. The average character has a 50/50 chance to pull something off, while a guy with an 18 has a 90% chance. On the other hand, just using your modifiers on a d20 roll probably means that your attributes don't matter enough. But something in between...

>But they'd have different stats, so they'd actually be different?
that means the one with lower stats would be categorically worse at everything the class does eg:
soldier - might be expected to repair arms/armour
barbarian - not so much, but might be good at whittling bone

let's say whittling and armour maintenance go off the same skill. You can't make the differentiation between a barbarian and a soldier in this way so long as they are the same class.
By hard-tying skills to classes, you gain specialty preservation but you lose granularity, customisation, etc

That would only be true if you were given a preset array every time you chose that class.

Thankfully, you roll for your stats in this situation, which means that you could technically get multiple stats that have a good chance of succeeding, which also gives you both options beyond your class's abilities and roleplay opportunities that play off of your bad stats.

As opposed to the way D&D usually does it where you'll always suck harder at performing a skill than the resident wizard because you only received 2+INT skill points to use and you had to dump INT so that you could have STR and CON.

Soldiers would most likely have STR+DEX+WIS, Generals would have STR+INT+CHA, and Barbarians would have STR+DEX+CON.

There, I just distinguished each and every one of these classes for you and all it took was five seconds of thought. It's not perfect but at least it's something that can be tweaked later on.

>I don't like this system so whenever more than one person says they like it, it must be shilling
Ok.

Eh, disciplines are kind of strict and it's hard to manage them properly. I'll stick with jobs. Jobs are pretty much freeform, and the only thing I require of my players is so that they pick more specific jobs at higher ranks (e.g. Burglar or Pickpocket instead of simply a Thief).

Eh, fair enough. Maybe more of a feat-like system?
Every level you get a new feat, and every two/three levels after first you get access to a new tier of feats?
It would be easily separable from Skills that way.

It's not the fact that I can beat the dude who has a +20 in a skill that I'm untrained in, it's the fact that I can achieve the same level of competency as a dude who spent years of his time mastering a skill at least 5% of the time no matter what, to the point where the bonuses don't even matter.

There aren't many situations where having a +20 in a skill is going to be utilized to its fullest extent. Like what does rolling a result of 40 do in the grand scheme of things? Chances are you would've succeeded even if you rolled a 1 on the die but then the question remains, why roll at all?

You'd basically be the fantasy equivalent to Saitama from OPM, you spent years honing your skills to be the very best but now, none of it matters because you'll never be able to utilize that talent to its fullest potential. Yet you'll still somehow be hit by a green soldier fresh out of the academy because he rolled a 20 on his attack roll.

Just imbalanced af.

>The average character has a 50/50 chance to pull something off, while a guy with an 18 has a 90% chance.

Isn't that kinda how it should be though?

50/50 for skills that you have a good chance of performing and nearly always when you're a goddamn beast at something?

I mean, I'd certainly expect the dude with 18 DEX to succeed at his acrobatics roll much more often than the guy who only has a 10 DEX.

>Yet you'll still somehow be hit by a green soldier fresh out of the academy because he rolled a 20 on his attack roll.
That's not a good example for a realistic system. If you're unarmed, it's very nearly impossible to get out of a knife fight with someone who has any real fighting experience completely unscathed, even if you are really good at disarming and martial arts. You're better off backing down, at least temporarily.

in five more seconds of thought you realise you can't distinguish between skills that roll under the same stat
Class 1 - good with axes
Class 2 - good with swords

5 more seconds and you'll start running into further problems

No, your system generally doesn't need to make such a distinction. That's why that system is fine for that game, but not others.

Repairing Arms/Armor would be an INT check since it'd be a measurement of how well you remember the process of how to produce it.

A soldier who spent years learning the process so that he could use it in the field is going to have an easier time of doing it as a Barbarian who spends most of his combat naked (or close to it) while hefting around the jaw bone of an ass for his weapon of choice.

I mean, this is pretty basic shit that I just came up with and even if it requires some tweaking to pull off, it's still something that differentiates them as a concept.

That and since it's more focused on your stats, it means that anyone with INT who can justify it with their background/class/race/etc. would be able to perform similar actions as well, up to GM's discretion.

That's fine for a system that's built around and focused on attributes, but D&D really isn't. Sure, it's had attributes from the very beginning, but they really didn't do much (see pic), and given that they were randomly generated, this was a good thing. A guy who got lucky on his rolls would be incredibly better than a guy who got unlucky if everything you did came down to attribute checks. That guy who has a 12 as his highest score is all kinds of fucked compared to the guy who got lucky and got a 17, a 15 and a 13.

D&D is really built around letting everybody have a fighting chance to do something, at least when that something falls outside of explicit class roles.

There's definitely a line that needs to be walked, which will itself be different for different systems. But the solution's not in a mandate of how rigidly to hold to RNG, it's about how often you call for checks to begin with. A lot of GMs call for checks too much and it's an easy trap to fall into: you have a hammer, and suddenly everything looks like a nail.

If it's something you *know* you'll let them do anyways regardless of the roll, they should probably just be able to do it without the gamble. Unless you're rolling for the context of how they do it--how much time or effort something took. In that case, because the player's aren't aware of all possible outcomes of the roll, they'll likely think it was more dynamic than it was. It's not lying, discarding the roll, or a total placebo--there can and should be ramifications from these rolls, just mild ones. It's something like misdirection, setting the tone and personal context for the players and their characters without sacrificing much else for it.
If you've ever played a BioWare game more than once you should have a firm idea of what that looks like.

And if you took 5 more seconds than you'd realize that there isn't a class that's only good with one specific type of weapon.

I'm pretty sure that other than weapon specialization, the only difference would be whether or not the class can use martial weapons or just simple weapons.

I'm not sure you're reading that right. It wholly agrees that soldiers might reasonably be thought of as better armourers than barbarians.

>anyone with INT who can justify it with their background/class/race/etc. would be able to perform similar actions as well, up to GM's discretion
I'm mostly concerned about what systems do without GM intervention

There are basic knife defense technique that anyone could perform if they know what they're doing and learn to stay calm under pressure.

It's a knife, not a fucking gun, it can't hurt you if it never comes anywhere near your body and if you're good at disarming and martial arts, you should have no problem defending yourself.

That's actually pretty solid - treating mundane shit as skills, and supernatural shit as feats (not spells).
I've played around with something like this in the past, when I was designing a different system.

As long as you're rolling for stats, you'll never have a character that has a fighting chance to do anything that's outside of their niche.

I don't really care since I don't really take D&D seriously in general but if I wanted a campaign where I wanted everyone to be equal, there are already systems in place that allow for arrays or point buys so that nobody gets fucked over by RNG.

It'd take some number tweaking and rebalancing sure but it's not that difficult to pull off.

WEG's Star Wars D6 has the best skill system, I think.

Skills are each categorized under an attribute. If you don't have the skill, roll the base attribute score.

It uses d6s, so your attribute for Dexterity might be 3D while your Blasters skill might be 3D+2 or your Dodge might be 4D. But you don't use a melee weapon, so using it untrained puts you at your natural 3D ability in Dex.

It's super simple, it works well, AND the results follow a bell curve. What more can you want?

I'm honestly hesitant to believe you. A guy with a knife is hardly gonna stand there and let you take his weapon, and reaching out to grab his hand puts you closer to him, not farther.
I'm not going to really argue the point much more than hearsay because I've never personally fought a guy with a knife, but I'd rather hand over my wallet than try to pull a fast one on a dude who might kill me or at least put a slash in my arm if I resist.

> there isn't a class that's only good with one specific type of weapon
this is my point, isn't it? The skill system (as you've described it) doesn't account for differing weapon skill without adding rules beyond that skill system?
But it's not about weapons. I'm using examples. The system, as is, just in general, can't distinguish between character types with the flexibility that different skill systems can i.e. very common ones which have skill ranks. Notably, many such skill systems will not distinguish between weapon classes in their skills (beyond ranged and melee), but this does not diminish the point that they could do so intrinsically, whereas the skill system you describe would need auxiliary rules.

>As long as you're rolling for stats, you'll never have a character that has a fighting chance to do anything that's outside of their niche.
Not if you're basing everything off attribute checks and not a combination of character background and GM fiat. But I'm generally a more fair system of randomization (drawing pre-set cards in different orders, rolling and randomly distributing random arrays, etc.) in any case. Still, if we want to look at this through the lens of AD&D nonweapon proficiencies, I'm not a big fan of the guy with a 10 stat being so much worse at making pottery, cutting gems, or whatever than a guy with a 16+ stat. And not just that one proficiency, but any others they get that are related to the same stat.

Just use basic common fucking sense.

A martial class will likely know how to maintain their weaponry due to practice but a mage who wears a robe and casts spells won't have the same benefit even if they both had the same INT score.

Granted, the mage would be able to produce enchantments which could be protection from X or deal fire damage or some shit so they'd still have their own niche.

Again, spotty but not necessarily impossible to work with.

The dude with the knife is going to have to come to you if he wants to stab you, which means that until he makes a move, you have time to either get into a stance that would offer the most protection or formulate a plan to disarm him.

Also, the way you hold a knife is to basically grasp it around the handle, which means that his options will boil down to stab or slash.

If they slash at you, you gotta aim for the wrists to stop the arc. If they stab, you can push their wrist away from you so that it doesn't hit you. Lateral force is much harder to deal with because you can't really stop a dude from pushing your hand to the left or right while you're already committing to the attack.

Now, in most cases it's easier to just hand over the wallet but let's face it here, the mugger is under no obligation to let you live, so it's good to know basic self-defense techniques just in case.

Thankfully, I've never had to use mine yet.

>Just use basic common fucking sense
why buy a game book then? What about when there is no consensus about what made up classes from made up races in made up cultures should be able to do? There is rarely a consensus when talking about real-world expectations, and 'common-sense' will differ greatly based on historical or other knowledge. Alternately, you could have information in the game books you buy that tell you clearly when to expect if someone is good at something and how that goodness is mechanically reflected.

>won't have the same benefit even if they both had the same INT score
yes, that's an inherent limitation of a system that represents skills with expectations, attributes and classes. Once a precedent has been set in a group about what a mage can do with their INT, all mages can do that with their INT.

>Again, spotty but not necessarily impossible to work with.
you should note I'm not saying this is impossible for things, just not optimal for things. It is, as you say, spotty.

If you only have an average stat then you can only expect an average outcome.

And if you don't want a character with a 10 in something to become that much worse than the dude with a 16+ in a stat then you can just use an array or point buy so that you're not forced into eating a crap score because of shitty luck.

>I'm not a big fan of the guy with a 10 stat being so much worse at making pottery, cutting gems, or whatever than a guy with a 16+ stat. And not just that one proficiency, but any others they get that are related to the same stat.
To elaborate on this a bit, D&D characters are primarily adventurers, and so anybody with a pottery or gem cutting proficiency should be pretty decent at doing that, but not completely overshadowed by somebody with the same proficiency and a significantly higher relevant attribute. If the characters were primarily gem-cutters, then a more extreme stratification due to natural ability / attributes would make more sense.

>Instead of classes, you can pick from a list of disciplines...
>Every level you get a new feat...
True20 does something *kind of* like this.

It's universal and classless, with characters falling between three broad Roles: Warrior for exceptionally hard to kill combat characters, Experts for exceptionally flexible skill-monkeys, and Adepts for power-using spellcasters/technosorcerers/psychics/deckers/etc. Each has a unique core ability to double-down on their niche.

Four feats at first level, plus a new feat each level up. Adepts can trade feats for a power instead. What might otherwise be class-specific features, like Sneak Attack and Smite Evil and Favored Enemy, are now feats. Some are specific to each Role.

Weapons didn't really matter all that much in the grand scheme of play anyways.

I mean, was there really that much of a difference between a great axe and a great sword when they both deal the exact same amount of damage?

Honestly, I wish that they did more with the special weapons qualities (Ex: reach, ranged, finesse, etc.) but if I was going for straight damage, there's no reason why I shouldn't just take the best damage dealing weapon in the game.

I object to pigeonholing characters too much on a special skill that they buy into. Or on a more common task, for that matter. I'm okay with some characters being better than others at some tasks, but they shouldn't outclass them to the point of worthlessness. If I buy into a nonweapon proficiency, I should be competent at whatever it relates to. And the guy with an 18 shouldn't be earthshatteringly good at that (and any other NWP he buys into related to the same stat) to the point where he completely outshines other people. One guy should be a competent painter (or whatever) and the other a good one, but the end products shouldn't be like comparing kindergarten finger painting to the Mona Lisa.

Arbiter
>Many skills, Skill level increases with skill use, degree of success/failure matters

>why buy a game book then?
It stands to reason that the game book would give you details into the setting, the cosmology, the life of an average peasant, monster stats, etc. or at the very least an idea as to what a class knows.
>Once a precedent has been set in a group about what a mage can do with their INT, all mages can do that with their INT.
Only in cases where you're a cuck who lets others tell you what you can and cannot do even when means to do so is staring you in the face.

FFS, there's nothing stopping you from asking the resident martial to teach you how to maintain arms/armors, just as there's nothing to stop you from teaching the fighter how to apply enchantments to his weaponry and armor.

Maybe it'll cost you some EXP or whatever to learn but now, you have a system where people are actually working together to improve their character's capabilities.

Some people are born with natural talent, which is represented by the character having an 18 in their stat, whiles others are not.

It's not fair but life isn't supposed to be fair.