Quote from: The Traveller;548072True, but nobody is saying a massive skill level difference won't carry the day almost without exception, most rules cover that. So the highly skilled are already covered (maybe not in D&D?), but in most real life situations outside of pro sports and science experiments, randomness comes into play a lot more.

still not sure.
Back to BTs OP climbing a rope. When he was a kid he couldn't climb a 30 foot rope. When I was a kid that was simple stuff. I would beat him every time based on this evidence.
Likewise I have always been a bit crap at racquet sports. When i was a kid I could play a game of tennis against a couple of my mates who played tennis a little. They woudl always win, not occassionally always. They weren't semi-pro tennis players they were 15 year old kids with a bit of natural talent and some practice.
There are lots of sports where someone a little better will always win. Pool, tennis, squash, boxing the list is extensive.

Skills are similar. I work with a lot of coders a good coder will alwys produce better code than a crap coder. I used to make my own furniture. Learnt it from a book, was alright but then I worked with a mate who was a trained cabinetmaker and the gulf was huge. He wasn't Chipendale but he had 2 years experience and 3 years of training.

I would say in real life different skills and challneges work out different but its closer to 20% talent 70% skill and 10% luck than 1:1:1

Quote from: jibbajibba;548079I would say in real life different skills and challneges work out different but its closer to 20% talent 70% skill and 10% luck than 1:1:1

You're talking about 90% predictability there, which is clearly not the case even in pro sports. The bottom line is if luck or whatever unforeseeable factors call themselves luck were that small a factor one could predict almost everything.

Quote from: The Traveller;548080You're talking about 90% predictability there, which is clearly not the case even in pro sports. The bottom line is if luck or whatever unforeseeable factors call themselves luck were that small a factor one could predict almost everything.

Your maths is a bit hokey. In pro-sports the guys all have near max natural talent and near max skill so then luck does come into play. The skill difference between Nadal and the world number 50 is probably about 6-7 (on a 1-100 scale). Lets stat Nadal at tennis. Give him 18/20 natural ability (he's a bit short if truth be told) and 69/70 Skill. So for a game he gets 87+1d10. then Take Andy Murray 17/20 natural ability and 67/70 skill so for a game of tennis he gets 84+1d10. Nadal will win 60% Murray 30% and it goes to an extraset 1 in 10 and we reroll....... seems about right.

Quote from: The Traveller;548078Easier just to give each skill a difficulty level as I do. Anything beyond difficulty 1 or 2 cannot be learned on the fly, and even then only "gross mechanical skills" are applicable, or wit based skills like charm. Climbing is difficulty 1, quantum physics difficulty 5.

This also feeds nicely into putting skill advancements on a gradient, making it slower to advance in difficult skills, and helping to decide skill starting scores.

Well, no.
Or yes, I think that is one factor.  I just don't think that covers the issue well.  Nesting skills allow the more real life -equiv of having a character learn 'physics 101' before learning more advanced skills, while giving a character a small ability in all the subskills.

I say yes, since I use an attribute bonus and an experience modifier per skill that accomodates for the difficulty level, but the nesting is also there to allow for generalists or specialists and to bar unskilled characters from choosing upper level skills.

Quote from: jibbajibba;548079I would say in real life different skills and challneges work out different but its closer to 20% talent 70% skill and 10% luck than 1:1:1

Quote from: The Traveller;548080You're talking about 90% predictability there, which is clearly not the case even in pro sports. The bottom line is if luck or whatever unforeseeable factors call themselves luck were that small a factor one could predict almost everything.

I'm not sure what these numbers are actually saying.  What does 66% predictability mean versus 90% predictability?  

I think it would be easier to give more concrete things.  Suppose there's a task that a top expert has a 90% chance of succeeding at.  What should be the chance of success by someone who has equal talent but no skill have?  (Understanding that this is a generalization)  

Typical RPGs might say that if expert skill is 90%, then a beginner might have 30%.  (From the other design thread cited, this was one example.)  

http://www.therpgsite.com/showthread.php?t=22152

Quote from: jhkim;548110I'm not sure what these numbers are actually saying.  What does 66% predictability mean versus 90% predictability?  

I think it would be easier to give more concrete things.  Suppose there's a task that a top expert has a 90% chance of succeeding at.  What should be the chance of success by someone who has equal talent but no skill have?  (Understanding that this is a generalization)  

Typical RPGs might say that if expert skill is 90%, then a beginner might have 30%.  (From the other design thread cited, this was one example.)  

http://www.therpgsite.com/showtread.php?t=22152

see my tennis example as i think its a reasonable fit.

Re the skilled versus the unskilled ... depends . You might be realy smart but with no grounding in sub atomic physics you can't solve a string theory problem anymore than a carpenter can make a chest of draws out of a tree with no tools.

Quote from: jhkim;548110I think it would be easier to give more concrete things.  Suppose there's a task that a top expert has a 90% chance of succeeding at.  What should be the chance of success by someone who has equal talent but no skill have?  (Understanding that this is a generalization)  

Well Jibbjabba is saying any given skill roll should be (1 to 20) plus (1 to 70) plus (1d0). I'm saying I feel its more like (1 to 10) plus (1 to 10) plus (1d10).

Really its a big complicated thing with lots of caveats and gotchas, and when you boil it all down the answer is probably 42, but we're sitting at the bar blowing wind here tbh, and it comes down to your tastes in gaming.

I think half random is too much, ten percent random is too little, so a nice median fits well. The acid test if it was important, and its not really, is how well the system models what might happen in real life - in my experience an even three way split does the job.

Quote from: The Traveller;548122Well Jibbjabba is saying any given skill roll should be (1 to 20) plus (1 to 70) plus (1d0). I'm saying I feel its more like (1 to 10) plus (1 to 10) plus (1d10).

Really its a big complicated thing with lots of caveats and gotchas, and when you boil it all down the answer is probably 42, but we're sitting at the bar blowing wind here tbh, and it comes down to your tastes in gaming.

Thanks.  It's a lot clearer to talk about 1 to 10 scales of skill and attribute.  

The issue that I see is that there are people who advocate this, but then introduce various conditions to prevent the natural consequences of this.  The previous thread was full of people with all sorts of conditions to put on contests - like requiring multiple rolls for a given task, Take 10, minimum skill or a special conditions for being unskilled, and the GM allowing skilled PCs to do things without rolling.  

In practice, I think this hodgepodge of fixes can be avoided by just making the skill range larger.  For example, take car driving.  In the Call of Cthulhu game this Saturday, we had to deal with the issue that several of our PCs - being lower-class types in the 1920s - didn't know how to drive.  With the 1 to 10 scale, it's impossible to deal with the spectrum of "I'm a Pacific Island native who's barely seen a car"; "I've taken taxis and seen some things but never driven"; "I'm a typical driver who drives every day"; and "I'm an top professional race car driver".  

GMs typically deal with this with various hand-waving to deal with it - but the end result is that there should be a significant difference in the chances of a Pacific Island native and the non-driver driving; but also between the non-driver and the licensed driver; and between the licensed driver and the top expert.  

Quote from: The Traveller;548122I think half random is too much, ten percent random is too little, so a nice median fits well. The acid test if it was important, and its not really, is how well the system models what might happen in real life - in my experience an even three way split does the job.

I don't think this models real life well at all.  Let's take a typical RPG situation.  Say there's a vault sealed by a high-tech, top-of-the-line security lock.  The PC skill 9 expert safecracker has a 90% chance of opening it, but by bad turn of the dice, he fails.  What do the PCs do?  

In real life, I think the usual answer would be "get a better expert".  In the typical RPG setup, though, they have a fair chance of succeeding by just all the others trying it even though they have only default skill.  (If Skill 9 has 90%, then Skill 1 has 10% - assuming equal talent.)

I can't help but think that too much of this problem derives from the binary pass/fail nature of most games' skill mechanics.  IRL a "task" you resolve with a "skill" 1) never breaks down that neatly into tasks and skills and b) is generally a lot fuzzier around the edges.  In fact, a lot of project management focuses on breaking work down into (arbitrarily defined) discrete, 100% pass-fail tasks because that's not the way most work is actually done.

Quote from: jhkim;548245With the 1 to 10 scale, it's impossible to deal with the spectrum of "I'm a Pacific Island native who's barely seen a car"; "I've taken taxis and seen some things but never driven"; "I'm a typical driver who drives every day"; and "I'm an top professional race car driver".  

The scale isn't 1 to 10 though, at least not how I run it. Tasks are set on a difficulty scale of 1 to 30, with 30 being almost impossible.

So say an average driving check, swerving to avoid a pedestrian, might be a 15. Averge reflexes (5) plus average skill (5) plus an average roll on a d10 (5 or better) will get you there. Someone with 1 skill but 9 reflexes stands the same chance, and vice versa.

A top class driver with super reflexes wouldn't need to roll at all there (10+10), but would only have a one in ten chance of successfully completing an almost impossible task, like doing a wall of death stunt with a medium sized van along the side of a building.

Thirty points in a scale is usually plenty, in my experience, but thats the system I use, it might not copy over to every game.

Quote from: jhkim;548245In the typical RPG setup, though, they have a fair chance of succeeding by just all the others trying it even though they have only default skill.  (If Skill 9 has 90%, then Skill 1 has 10% - assuming equal talent.)

I think the layout described above would deal with that issue. Some rolls would be just too high to reach on a 1d10. You could roll open ended or exploding dice, but you've as good as or a better than even chance of tripping the alarm and somehow blowing yourself up in the process if you do that.

Quote from: daniel_reamI can't help but think that too much of this problem derives from the binary pass/fail nature of most games' skill mechanics. IRL a "task" you resolve with a "skill" 1) never breaks down that neatly into tasks and skills and b) is generally a lot fuzzier around the edges. In fact, a lot of project management focuses on breaking work down into (arbitrarily defined) discrete, 100% pass-fail tasks because that's not the way most work is actually done.

Some skills in my setup do work like that, but you want to keep it to a minimum to prevent excessive rulebook checking for the corner cases on each skill.

Quote from: jibbajibba;548058Now most of the time you aren't stretching your skill you are doing things inside your threshold. Its an argument for the all D&D PCs are adventures so then can all do this basic stuff, but I hate the idea of a world with professional adventurers and also once you do stretch yourself how do you determin the result if everyone can basically do whatever they like and where skills do intervene, like a thief having move silent of just 10% so an untrained person should have lower, or a thief having 70 climb an untrained person should have lower (and people should read the theif climb rules again, it requires hand holds etc and doesn't turn them into spiderman).

That's kind of along the lines of why I'm not a big fan of using D100 for skill checks.  D100 results are random and have no attachment or binding what-so-ever to a character's action.

Quote from: The Traveller;548078Easier just to give each skill a difficulty level as I do. Anything beyond difficulty 1 or 2 cannot be learned on the fly, and even then only "gross mechanical skills" are applicable, or wit based skills like charm. Climbing is difficulty 1, quantum physics difficulty 5.

This also feeds nicely into putting skill advancements on a gradient, making it slower to advance in difficult skills, and helping to decide skill starting scores.

I just had some crazy idea about an RPG system having rules for a character running fast or winning a running race by making those skills easy ones.  The character rolls.  And because the skill is so easy for that character, he/she easily succeeds (wins).

Quote from: Shawn Driscoll;548287I just had some crazy idea about an RPG system having rules for a character running fast or winning a running race by making those skills easy ones.  The character rolls.  And because the skill is so easy for that character, he/she easily succeeds (wins).

The difficulty of the skill isn't related to the difficulty of the task here. The only time skill difficulty gets involved is when you want to improve your skill score (training) or in chargen.

Rarely it might get involved if you need to try something out of the blue, like climbing with zero skill - possible, but hard, and even if you succeed you won't gain much. Its flat out disallowed for higher difficulty or intellectual skills, without assistance.

Quote from: jhkim;548245Let's take a typical RPG situation.  Say there's a vault sealed by a high-tech, top-of-the-line security lock.  The PC skill 9 expert safecracker has a 90% chance of opening it, but by bad turn of the dice, he fails.  What do the PCs do?  

In real life, I think the usual answer would be "get a better expert".  In the typical RPG setup, though, they have a fair chance of succeeding by just all the others trying it even though they have only default skill.  (If Skill 9 has 90%, then Skill 1 has 10% - assuming equal talent.)

That is weird how a character with almost no skill could still roll a successful result in opening such a vault.  That would lead me to believe that the character (in real life) has some sort of common sense or logical approach to things.  And if common sense or logical approach were talents (not skills) in an RPG, they could be used for opening the vault instead of rolling against a low vault skill.  A character without those talents would need to have a high vault skill to open the vault.

Maybe there is an RPG out there that has talents in its system?

Quote from: The Traveller;548271The scale isn't 1 to 10 though, at least not how I run it. Tasks are set on a difficulty scale of 1 to 30, with 30 being almost impossible.

So say an average driving check, swerving to avoid a pedestrian, might be a 15. Averge reflexes (5) plus average skill (5) plus an average roll on a d10 (5 or better) will get you there. Someone with 1 skill but 9 reflexes stands the same chance, and vice versa.

A top class driver with super reflexes wouldn't need to roll at all there (10+10), but would only have a one in ten chance of successfully completing an almost impossible task, like doing a wall of death stunt with a medium sized van along the side of a building.

OK, so let's take a simple task - like merging onto a highway without any incident.  What's the difficulty of that?  There are thousands of people who do that every day without incident, so the chance of success should be a lot closer to 100% than 90% - even for people with below average reflexes.  Let's call it difficulty 8.  

Now let's take a native Pacific Islander - a player character from my last night's Call of Cthulhu game - who has never seen a car.  What is his chance of successfully merging onto the highway?  Well, he's got above average reflexes (6), and since the skill scale is 1 to 10, I guess he has a 1 skill.  So he's got an 80% chance to drive a car and merge onto the highway successfully.  

I think that is significantly over-estimating his chances.  

Conversely, let's imagine something that is difficulty 22.  This is something that is just a little bit tricky for the best driver in the world.  At reflexes 10 and skill 10, he'll fail at this 1 out of 10 tries.  What is this like?  Well, let's keep in mind what kind of stunts trick drivers can do for a daily show - like jump a car 30 feet in reverse.  

http://www.youtube.com/watch?v=k5GS7A_frYI&feature=relmfu

Difficulty 22 is something definitely tougher than that.  

Now, let's take someone with no special driving skill but above-average reflexes: reflexes 7 and driving skill 5.  He can succeed at that stunt 10% of the time.  That also seems a stretch.