TheRPGSite

I'm wondering if people have a preference/opinion as to what sort of scale is best for attributes/stats.

e.g.
3-18 (D&D traditional)?
1-5 (Storyteller)?
1-10 (Twilight 2000)
1-100 (Rolemaster)
Open-ended ?
-5 to +5 ?
d4 to d12 ?
[or whatever]

I realize this is a stupid question since I haven't specified what mechanic I want to use with this (roll under, roll over, dice pool, etc). But I'm trying to figure out what scale is best in order to help decide what core mechanic is best too. (Eventually the horse will learn to push the damn cart).

Are you looking for simplicity or complexity?

Quote from: Bloody Stupid Johnson;5305103-18 (D&D traditional)?

So how strong is a storm giant on the 3-18 D&D scale? And if that value it doesn't fit within 3-18, is it really a 3-18 scale?

Or are you asking: "what is the best range for normal human attributes in a game?" because that is a slightly different question (at least IMO).

I'd rather have a scale that means something, so that I can understand what the difference between a 10 and an 18 is. Then people are added to the scale based on what they can actually do. If an Olympic weight lifter can lift 4X as much as a normal person, then his stat values should match that fact.

It would also be nice if the scale of progression covered a fairly wide range, while allowing a nice variation for human level characters.

For example, while I really like a lot of things about Mayfair's DC Heroes game, the idea that a normal person has stats of 2, and a progression where each point doubles (3 is twice as much as 2, 4 is 4X as much as 2, and 5 is 8X as much as 2) doesn't leave much of a range for normal people.

Quote from: ggroy;530538Are you looking for simplicity or complexity?

I think simplicity is a plus. In general I normally feel that extra complexity is fine, as long as there's a justification behind having the extra 'mechanical overhead'.  
In this context for example, one of the factors to consider may be whether the number can be used directly for modifying dice rolls or whether it needs a modifier to be calculated via an equation or table.

I think direct use > equation> table all other things being equal - since ability damage or the like can mean a trip back to the table to recalculate other modifiers, but there are other potential factors to consider as well.

I hope that makes sense...

Quote from: warp9;530541So how strong is a storm giant on the 3-18 D&D scale? And if that value it doesn't fit within 3-18, is it really a 3-18 scale?

Or are you asking: "what is the best range for normal human attributes in a game?" because that is a slightly different question (at least IMO).

I'd rather have a scale that means something, so that I can understand what the difference between a 10 and an 18 is. Then people are added to the scale based on what they can actually do. If an Olympic weight lifter can lift 4X as much as a normal person, then his stat values should match that fact.

It would also be nice if the scale of progression covered a fairly wide range, while allowing a nice variation for human level characters.

For example, while I really like a lot of things about Mayfair's DC Heroes game, the idea that a normal person has stats of 2, and a progression where each point doubles (3 is twice as much as 2, 4 is 4X as much as 2, and 5 is 8X as much as 2) doesn't leave much of a range for normal people.

Traditionally in D&D, the Storm Giant had a 24 (3.x its higher but I forget exactly what;  Titans went from 25 to 37).

I guess I'm asking for a generic system (not necessarily just fantasy) and it would be a plus if a system could handle superhuman attribute scores as well, while still having enough definition in the range most (presumably human) PCs will fit into.

Having the number translate into something definable as you suggest would also be good, too.

I've seen DC - I sort of agree with the granularity as far as Strength and Body go. It was actually a bit odd in that the scale was entirely different from stat to stat, though - Batman has normal human maximum Strength at 5, maximum Body of 6, and Int of about 12 (Lex Luthor's is what, 15 ?), while Hal Jordan is human and has WILL 24. The scale meant something clear-cut for Str/Body but became very vague for the other stats - its unclear to me what Batman being a thousand times as perceptive as a normal human meant, for instance.

Quote from: Bloody Stupid Johnson;530547I've seen DC - I sort of agree with the granularity as far as Strength and Body go. It was actually a bit odd in that the scale was entirely different from stat to stat, though - Batman has normal human maximum Strength at 5, maximum Body of 6, and Int of about 12 (Lex Luthor's is what, 15 ?), while Hal Jordan is human and has WILL 24. The scale meant something clear-cut for Str/Body but became very vague for the other stats - its unclear to me what Batman being a thousand times as perceptive as a normal human meant, for instance.

As his perception is worthy of Great Old Ones, Batman simply can not comprehend that some of his detective methods may not be as effective as they'd appear to his cold logic.

Example:

(http://4.bp.blogspot.com/-_dUy0V85kuc/TafUFKs9V5I/AAAAAAAAAqc/gr0mwYDFOvQ/s1600/tumblr_lj3l71Caxa1qzzoo1o1_1280.jpg)

Quote from: Bloody Stupid Johnson;530547I think simplicity is a plus. In general I normally feel that extra complexity is fine, as long as there's a justification behind having the extra 'mechanical overhead'.  
In this context for example, one of the factors to consider may be whether the number can be used directly for modifying dice rolls or whether it needs a modifier to be calculated via an equation or table.

I think direct use > equation> table all other things being equal - since ability damage or the like can mean a trip back to the table to recalculate other modifiers, but there are other potential factors to consider as well.

If such a system is going to be used in a video game, then in principle it doesn't matter how simple or complex the underlying system is.  The computer becomes a black box which does all the calculations on the fly.

For example, if one attempts to translate the system of mechanics used in a video game like World of Warcraft, most likely it would look like something which makes Rolemaster or 3.xE/d20 look "simple" in comparison.

Quote from: Rincewind1;530549As his perception is worthy of Great Old Ones, Batman simply can not comprehend that some of his detective methods may not be as effective as they'd appear to his cold logic.

Example: [snip image]

Oh I don't know, that could work :) Cutting the Gordian knot, so to speak. I'm not sure this thread isn't basically me doing that...

Sometimes the best approach to solving a problem is to run from one end of the ship to the other going "Is there a Theologician on board?!" until someone comes out just to shut you up.

Quote from: ggroy;530554For example, if one attempts to translate the system of mechanics used in a video game like World of Warcraft, most likely it would look like something which makes Rolemaster or 3.xE/d20 look "simple" in comparison.

True...  I'd be looking for something for pen-and-paper RPG use, though.

Quote from: Bloody Stupid Johnson;530555True...  I'd be looking for something for pen-and-paper RPG use, though.

The great thing about mechanics for pen-and-paper rpgs, is that it has to be simple enough such that it can written out in a reasonable sized book.

It's a completely different story when it comes to computer games, or real life in the offline world.  The "mechanics" can be as complex as the designer wants for the former, while it is literally infinitely complex for the latter.


On the other side of the coin, some systems which can be easily written down on paper, becomes laughably simple to defeat and circumvent on the computer.

A prime example of this are pen-and-paper encryption systems, such as older alphanumeric transposition and substitution ciphers.

http://en.wikipedia.org/wiki/Transposition_cipher
http://en.wikipedia.org/wiki/Substitution_cipher


EDIT:  http://en.wikipedia.org/wiki/Ciphertext-only_attack

Quote from: Bloody Stupid Johnson;530555Sometimes the best approach to solving a problem is to run from one end of the ship to the other going "Is there a Theologician on board?!" until someone comes out just to shut you up.

More generally before the internet was popular, the easiest most efficient way was to find an expert on the topic of interest.

These days, one just does a google search for a lot of stuff.  :p

Quote from: ggroy;530557The great thing about mechanics for pen-and-paper rpgs, is that it has to be simple enough such that it can written out in a reasonable sized book.

It's a completely different story when it comes to computer games, or real life in the offline world.  The "mechanics" can be as complex as the designer wants for the former, while it is literally infinitely complex for the latter.

On the other side of the coin, some systems which can be easily written down on paper, becomes laughably simple to defeat and circumvent on the computer.

Even in a computer-based game I think its still a plus for the user (of the program) to be able to understand what their attributes do, of course.

There are of course families of problems which are basically uncrackable even with massive computing power.
http://en.wikipedia.org/wiki/Transcomputational_problem

At least, short of something like Dyson-sphere sized computers and the like.
http://en.wikipedia.org/wiki/Matrioshka_brain

I do wonder if the opposite applies, and there are any problems that are easy to solve with paper that are very difficult to do with a computer. Can't really think of any. Roman message sticks, maybe ? (where they wrap the paper around a stick of the right width to decipher a code). It would be solvable with the computer, but is much easier with a stick.

Quote from: Bloody Stupid Johnson;530564I do wonder if the opposite applies, and there are any problems that are easy to solve with paper that are very difficult to do with a computer. Can't really think of any.

There are some mathematical theorems which are easy to state and prove on paper, but which can be difficult to prove using a computer.  The simplest one I can think of offhand, would be:

"Prove that there are an infinite number of prime numbers".

This can be easily proved using "proof of contradiction" in several lines.

On a computer, there's no easy way of proving anything which is infinite in number by exhaustively going through every single case.  (The computer will run out of memory or crash eventually).

Quote from: Bloody Stupid Johnson;530547I guess I'm asking for a generic system (not necessarily just fantasy) and it would be a plus if a system could handle superhuman attribute scores as well, while still having enough definition in the range most (presumably human) PCs will fit into.

There's your problem.  To my knowledge, this has been done innumerable times, but it has never been done well.  I think Marvel Supers came about the closest, really, but even with FASERIP is was more than a little wonky at the high end.

And it's not even the ability scores that cause the problem, it's the scaling of tasks.  You can certainly go with a logarithmic scale like DC Heroes, but it will take some fiddling to make the presentation simple and non-recursive.  How strong is 30?  Well, it's twice as strong as 29.  How strong is 29?  Well, it's twice as strong as 28.  And so on.  And it gets fairly ridiculous early on.  If you take just the basics, and state that a Str of 1 means you can lift 2 pounds, then a 2 is four pounds, a 3 is eight pounds, and so forth.  Essentially, 2^(Str score) in pounds.  A score of 9 would be a bit over the maximum human capacity, sitting at 512 pounds.  Add another point, and you are lifting half a ton.  Bump it up to 20, and you are lifting 500 tons.  Bump it up to 30, and it's over a million tons.  The Earth weighs 1.3E+25 pounds, or about 6.5E+21 tons.  Now, that is a pretty staggering Str score of 72.5 (or thereabouts), which is probably going to be out-of-bounds in any case.  That is still 1.125E+15 pounds, or about 5.63E+11 tons.

It's really difficult to get the scaling and the balance correct.  If you keep the PCs at just the normal human limits, however, most of the problem goes away.  ;)

Set the scale to have 2 attributes.

So Strength A3, B6, etc

The first letter is the category (you can expand this of course or create different subcategories)

A = Godlike
B = SuperHuman
c = Preternatural
D = Human
E = Weak
F = Feeble

the number indicates relative comparison

So Superman might have Strength B8 in your system and Thor might have B7 or Thor might have A6 and Magnar A8 however you cchose to run it in your 'verse.

In comparison an A always beats a B etc you only bother to compare between creatures on the same scale. A human (D) can always beat a cat (E) and a cat can always beat a mouse (F)
In effect you can then run a modern warfare game where every one has D1-9 (Or even D3 - 18 if you like) and then chuck in a tank with Strength B4.
You can run a fantasy game where an exceptional hero might have Strength C4 but he will still be weaker than a Giant (B range)
Also you can have creatures physically exceptional but mentally equal. Superman might be

Str = B8  Dex = B7 Con = B4
Int = D8  Wis = D7 Chr = C4

Or whatever......


Then you just need to work out how many Ds equal a C Lets take 10 as a good number (but you might like 8, 12, or even 16).

So in this concept 100 people (Strength D ) might have a chance in a tug of war against 1 tank (Strength B).

In a superhero game, the question is what exactly is "balance" in such a system.

Quote from: jibbajibba;530584Set the scale to have 2 attributes.
 
So Strength A3, B6, etc
 
The first letter is the category (you can expand this of course or create different subcategories)
 
A = Godlike
B = SuperHuman
c = Preternatural
D = Human
E = Weak
F = Feeble
 
the number indicates relative comparison
 
So Superman might have Strength B8 in your system and Thor might have B7 or Thor might have A6 and Magnar A8 however you cchose to run it in your 'verse.
 
In comparison an A always beats a B etc you only bother to compare between creatures on the same scale. A human (D) can always beat a cat (E) and a cat can always beat a mouse (F)
In effect you can then run a modern warfare game where every one has D1-9 (Or even D3 - 18 if you like) and then chuck in a tank with Strength B4.
You can run a fantasy game where an exceptional hero might have Strength C4 but he will still be weaker than a Giant (B range)
Also you can have creatures physically exceptional but mentally equal. Superman might be
 
Str = B8 Dex = B7 Con = B4
Int = D8 Wis = D7 Chr = C4
 
Or whatever......
 
 
Then you just need to work out how many Ds equal a C Lets take 10 as a good number (but you might like 8, 12, or even 16).
 
So in this concept 100 people (Strength D ) might have a chance in a tug of war against 1 tank (Strength B).

I don't see the point of the ranks exactly, numbers are more convenient for lots of things? - i.e.gets hit points = stat, throw tank miles = stat, whatever.
 
Here if you say have a range of 0-9 in each category, you could have a number where "tens place" of representing the scales F through A. In which case it'd really be a scale from 0 (F0) to 69 (A9). [broader ranges within each rank could also be expressed as numbers if you use e.g. hexadecimal or other base numbering]...

Quote from: Bloody Stupid Johnson;530510I'm wondering if people have a preference/opinion as to what sort of scale is best for attributes/stats.

The best scale is the one which matches the mechanics of the system. The best mechanics are the ones that match your needs in designing the system.

I know this is the obvious and expected answer, and could be seen as being so general as to be considered useless. But neither answer is vague, they are exactly correct.

For your own purposes, begin by defining exactly what it is you wish to accomplish with the game. Then build the mechanics and statistics to match those goals.

Side note: It is very difficult building a scale that can include both human-level and superhuman, without running into significant problems. The real world is just not mathematically elegant.

In the abstract, a system like Torg's is as close to the mathematical ideal as one can come (answering the objections about DC Heroes). In practice, the attempt to mathematically define a strong and intimate link between real world values and game measurements (Strength of 10 = can lift 100 kilos, Dex 10 = can run 100m in 10 seconds) inevitably requires fudge factors, caveats, and exemptions.

Even then, the scale is 5-8-13 (human min, avg, max), which doesn't appeal to simplicity or aesthetics. Plus, it's laughable. Max Str is actually 12, max Dex is actually 10, and while mass is a good stand-in for Toughness, that puts the obese human max at between 14 and 15 and regular human max at around a 9-10.

I've done the research on real world performance, and tried to fiddle with scales to match, but there are no elegant scales and no exemption-free gaming. Eventually, I had to chuck the concept entirely.

Again, it comes down to: what kind of game are you trying to make? What mechanical elements are you including, and why?

Start with those, then retroactively define Lift and Carry values to match. Shadowrun and GURPS both have different attribute scales, but their calculated values for L&C are are comparable. If the game mechanics are playable, rough equivalence to real world performance is acceptable.

(There used to be an RPG Olympics site that calculated how fast/strong characters from a dozen systems were. It was interesting, unfortunately it's vanished. But it supported my "comparable to real world performance" contention.)

Quote from: StormBringer;530582There's your problem. To my knowledge, this has been done innumerable times, but it has never been done well. I think Marvel Supers came about the closest, really, but even with FASERIP is was more than a little wonky at the high end.
 
And it's not even the ability scores that cause the problem, it's the scaling of tasks. You can certainly go with a logarithmic scale like DC Heroes, but it will take some fiddling to make the presentation simple and non-recursive. How strong is 30? Well, it's twice as strong as 29. How strong is 29? Well, it's twice as strong as 28. And so on. And it gets fairly ridiculous early on. If you take just the basics, and state that a Str of 1 means you can lift 2 pounds, then a 2 is four pounds, a 3 is eight pounds, and so forth. Essentially, 2^(Str score) in pounds. A score of 9 would be a bit over the maximum human capacity, sitting at 512 pounds. Add another point, and you are lifting half a ton. Bump it up to 20, and you are lifting 500 tons. Bump it up to 30, and it's over a million tons. The Earth weighs 1.3E+25 pounds, or about 6.5E+21 tons. Now, that is a pretty staggering Str score of 72.5 (or thereabouts), which is probably going to be out-of-bounds in any case. That is still 1.125E+15 pounds, or about 5.63E+11 tons.
 
It's really difficult to get the scaling and the balance correct. If you keep the PCs at just the normal human limits, however, most of the problem goes away. ;)

Superhumans are a bit of an edge case really. Again nice if a system handles this, but its only a minor consideration, usually (unless I was making a supers game - in which case I'd have to concede that the demands of supers mean the ideal system for them is different to everyone else..).
 
I'd agree that the logarithmic nature of DC scales it up probably too quickly for it to be super-useful outside of mega-super-heroic action ? The idea of attributes having an exponential scale, though (rather than linear like MSH ?) is workable I think, if the doublings happen less frequently; 3.5 D&D for example has a scale where +10 Strength = x4 lifting capacity.

Quote from: Bloody Stupid Johnson;530592I don't see the point of the ranks exactly, numbers are more convenient for lots of things? - i.e.gets hit points = stat, throw tank miles = stat, whatever.
 
Here if you say have a range of 0-9 in each category, you could have a number where "tens place" of representing the scales F through A. In which case it'd really be a scale from 0 (F0) to 69 (A9). [broader ranges within each rank could also be expressed as numbers if you use e.g. hexadecimal or other base numbering]...

No the idea is that there are human tasks (say rank D) these are things any human could do. You don;t need to make an Int check to open a door. A dog with E x inteligence however would need to make an Int Check to open a door.

Basically it takes a system like the Marvel one and allows you to compare between different ranks. In the Marvel system normal humans can't be effectivly compared becuase they are all the same compared to most superheroes. This system lets you still compare all humans on an effective range but it gives you alternate ranges for different categories of thing.

Take an in play example.
How far can you throw a 2 ton car?
Well throwing a 2 ton car is classed as B rank task. If you have C Rank or D rank or E rank you just can't do it.
If you have B rank Strength then you can throw a car and how far you can throw it depends on the number and you can use whatever number sysatem you like 3d6, 2d10, 0-9 whatever....

Now I am just riffing with this so I haven't even considered how it wourl work with a damage system liek hit points but my guess would be that in basic play a Rank C Strength thing just couldn;t harm a Rank B Con thing.  etc.

So Captain America (give him C3 Strength) jsut can't harm Superman no matter how hard he hits him. If he uses his sheild... maybe that means his manage moves a rank higher and now he can damage rank B con things....

Moving between scales is a BIG step, within a range everythgin can be compared......

Quote from: Bloody Stupid Johnson;530595The idea of attributes having an exponential scale, though (rather than linear like MSH ?) is workable I think,

See my post above, which is all about exactly this.

Quote from: Daddy Warpig;530593The best scale is the one which matches the mechanics of the system. The best mechanics are the ones that match your needs in designing the system.
 
I know this is the obvious and expected answer, and could be seen as being so general as to be considered useless. But neither answer is vague, they are exactly correct.
 
For your own purposes, begin by defining exactly what it is you wish to accomplish with the game. Then build the mechanics and statistics to match those goals.
 
Side note: It is very difficult building a scale that can include both human-level and superhuman, without running into significant problems. The real world is just not mathematically elegant.
 
In the abstract, a system like Torg's is as close to the mathematical ideal as one can come (answering the objections about DC Heroes). In practice, the attempt to mathematically define a strong and intimate link between real world values and game measurements (Strength of 10 = can lift 100 kilos, Dex 10 = can run 100m in 10 seconds) inevitably requires fudge factors, caveats, and exemptions.
 
Even then, the scale is 5-8-13 (human min, avg, max), which doesn't appeal to simplicity or aesthetics. Plus, it's laughable. Max Str is actually 12, max Dex is actually 10, and while mass is a good stand-in for Toughness, that puts the obese human max at between 14 and 15 and regular human max at around a 9-10.
 
I've done the research on real world performance, and tried to fiddle with scales to match, but there are no elegant scales and no exemption-free gaming. Eventually, I had to chuck the concept entirely.
 
Again, it comes down to: what kind of game are you trying to make? What mechanical elements are you including, and why?
 
Start with those, then retroactively define Lift and Carry values to match. Shadowrun and GURPS both have different attribute scales, but their calculated values for L&C are are comparable. If the game mechanics are playable, rough equivalence to real world performance is acceptable.
 
(There used to be an RPG Olympics site that calculated how fast/strong characters from a dozen systems were. It was interesting, unfortunately it's vanished. But it supported my "comparable to real world performance" contention.)

(Just to note I wrote previous post above before yours was up. Probably doesn't change anything but FYI).
I'll have to ponder your response some more.
 
I don't know that I can give up on searching for the One True Perfect System but I think your answer is interesting because the emphasis it puts on 'defining a link between real-world performance' and the scale. That's definitely something I should be considering more since I'd previously though of 'best' more in terms of mechanical convenience (minimizing table lookups and modifiers, ease of recalculating ability damage), last several posts on bench pressing things notwithstanding...
 
Possibly the need to define stats in real terms is why people tend not like stuff like e.g. Blue Rose where they dropped the normal D&D stat scale and just used the modifiers...
 
Hmm.

Quote from: Daddy Warpig;530597See my post above, which is all about exactly this.

Sorry I seem to be out of phase with the conversation!

Quote from: jibbajibba;530596No the idea is that there are human tasks (say rank D) these are things any human could do. You don;t need to make an Int check to open a door. A dog with E x inteligence however would need to make an Int Check to open a door.
Basically it takes a system like the Marvel one and allows you to compare between different ranks. In the Marvel system normal humans can't be effectivly compared becuase they are all the same compared to most superheroes. This system lets you still compare all humans on an effective range but it gives you alternate ranges for different categories of thing.

OK, gotcha. I could see it being used to extend a system without having huger numbers then, although maybe there might be problems around the scale borders. Again hmm.
 

QuoteNow I am just riffing with this so I haven't even considered how it wourl work with a damage system liek hit points but my guess would be that in basic play a Rank C Strength thing just couldn;t harm a Rank B Con thing. etc.  

That works. I guess that isn't just Con so much as actual physical toughness (some supers games do that e.g. DC Heroes has a "Body" stat where a 10 = literally steel-hard skin).

Quote from: Bloody Stupid Johnson;530595Superhumans are a bit of an edge case really. Again nice if a system handles this, but its only a minor consideration, usually (unless I was making a supers game - in which case I'd have to concede that the demands of supers mean the ideal system for them is different to everyone else..).
 
I'd agree that the logarithmic nature of DC scales it up probably too quickly for it to be super-useful outside of mega-super-heroic action ? The idea of attributes having an exponential scale, though (rather than linear like MSH ?) is workable I think, if the doublings happen less frequently; 3.5 D&D for example has a scale where +10 Strength = x4 lifting capacity.

That is what I meant with that last line.  It works fine with scores in the 0-9 range, if you want logarithmic scaling.  Since the numbers start getting pretty far apart as you get higher, apply a penalty for each percent over the minimum or something.  So a Str 8 would have a 1% penalty for every .25 pounds over 256 they attempt to lift, or something.  If you don't them having even a chance to infringe on the Str 9 character, do a half pound or a full pound for each percent.

Don't get me wrong, I am not trying to discourage you.  It is certainly possible to mix human and super-human if that is what you want, it's just very, very difficult.  You mentioned giants above; allow them a Str of 15 or 16, maybe dragons have Str 18, and anything above that not available to mortal creatures under any circumstances other than divine ascension.

Alternately, if you don't mind digging your calculator out, you could have fractional scores as well.  Like Str 3.7 or 8.2.  That would give you a lifting capacity of 12.996lbs and 294.066lbs respectively.  It may be too fine grained at the lower scores, so perhaps only use it above 7 or 8, like Exceptional Strength in AD&D 1st.  That would make Str 8.75 just about exactly the maximum weight dead lifted by a person.  You could even simplify things by using .25 increments above 7 or 8 or whatever.  Then you don't have to track 7.2 or 8.46 or something goofy.

You can set the doublings to occur less frequently, but that is starting to look like a chart.  I like charts, so I am not bringing that up as a problem, mind.  If you wanted to use a number smaller than 2, but stick with the exponent thing, you probably shouldn't go lower than 1.8; that gives a lifting capacity of 357lbs at a score of 10.  Much lower than that, and you will be looking at a pretty big range of low-end.  About 0-15 at 1.5.

The math just gets wonky as hell, but if you want to try something like that, drop me a line, I can do some number fiddling and you can see what you are looking at.

I prefer named scales where each level has a certain name assciated to it. Since people will not be able to learn a great number of them, there can't be to many levels. 5 to 6 maybe. Scale can be increased by trippling it for a low, medium and high with each designation.

Quote from: StormBringer;530753That is what I meant with that last line.  It works fine with scores in the 0-9 range, if you want logarithmic scaling.  Since the numbers start getting pretty far apart as you get higher, apply a penalty for each percent over the minimum or something.  So a Str 8 would have a 1% penalty for every .25 pounds over 256 they attempt to lift, or something.  If you don't them having even a chance to infringe on the Str 9 character, do a half pound or a full pound for each percent.

Don't get me wrong, I am not trying to discourage you.  It is certainly possible to mix human and super-human if that is what you want, it's just very, very difficult.  You mentioned giants above; allow them a Str of 15 or 16, maybe dragons have Str 18, and anything above that not available to mortal creatures under any circumstances other than divine ascension.

Alternately, if you don't mind digging your calculator out, you could have fractional scores as well.  Like Str 3.7 or 8.2.  That would give you a lifting capacity of 12.996lbs and 294.066lbs respectively.  It may be too fine grained at the lower scores, so perhaps only use it above 7 or 8, like Exceptional Strength in AD&D 1st.  That would make Str 8.75 just about exactly the maximum weight dead lifted by a person.  You could even simplify things by using .25 increments above 7 or 8 or whatever.  Then you don't have to track 7.2 or 8.46 or something goofy.

You can set the doublings to occur less frequently, but that is starting to look like a chart.  I like charts, so I am not bringing that up as a problem, mind.  If you wanted to use a number smaller than 2, but stick with the exponent thing, you probably shouldn't go lower than 1.8; that gives a lifting capacity of 357lbs at a score of 10.  Much lower than that, and you will be looking at a pretty big range of low-end.  About 0-15 at 1.5.

The math just gets wonky as hell, but if you want to try something like that, drop me a line, I can do some number fiddling and you can see what you are looking at.

Thanks Stormy. The fractions is interesting as an idea - something I hadn't seen before really (will add to design archive)- although probably not what I'm looking for exactly.
Not exactly trying to design the ultimate supers game but I could mostly avoid lots of complex calculations, I think, by working with "Size" scores of objects instead of having to work with pounds (or kilograms). Thanks, though!

Quote from: 1of3;530755I prefer named scales where each level has a certain name assciated to it. Since people will not be able to learn a great number of them, there can't be to many levels. 5 to 6 maybe. Scale can be increased by trippling it for a low, medium and high with each designation.

If you can describe it, what's the appeal of just having 'ranks' like FUDGE/FATE/MSH (/Amber), rather than numbers?

Quote from: Bloody Stupid Johnson;530795Thanks Stormy. The fractions is interesting as an idea - something I hadn't seen before really (will add to design archive)- although probably not what I'm looking for exactly.
Not exactly trying to design the ultimate supers game but I could mostly avoid lots of complex calculations, I think, by working with "Size" scores of objects instead of having to work with pounds (or kilograms). Thanks, though!

Welcome!  The size scores for various objects is a good idea, you can throw in a modifier for density as well without screwing up the calculations too much, and you can still keep track of the actual 'size'.  So, a bowling ball would be about size 2, but a bowling ball made of lead would be size 2 x density 5, for an effective 'size' of 10.  Still not bulky, but really damn heavy.

Sorry about the major derail into supers, I only intended it as an example, but I went a little overboard.  :)

Quote from: StormBringer;530845Welcome!  The size scores for various objects is a good idea, you can throw in a modifier for density as well without screwing up the calculations too much, and you can still keep track of the actual 'size'.  So, a bowling ball would be about size 2, but a bowling ball made of lead would be size 2 x density 5, for an effective 'size' of 10.  Still not bulky, but really damn heavy.

Sorry about the major derail into supers, I only intended it as an example, but I went a little overboard.  :)

Yep that'd work..!
And pfft no problem. Not that far off topic, and a thread that can't survive derailing doesn't deserve to live.

Quote from: Bloody Stupid Johnson;530851Yep that'd work..!
And pfft no problem. Not that far off topic, and a thread that can't survive derailing doesn't deserve to live.

:hatsoff:

Quote from: Bloody Stupid Johnson;530598I don't know that I can give up on searching for the One True Perfect System

I use to think this way, in regard to rpg systems and other things.

Outside of rpg games, this has not exactly been the most fruitful pursuit.

For example in the hard sciences over the last several centuries, there have been numerous individuals which have attempted to formulate a fundamental theory which can explain everything in the universe.  Einstein spent half of his career on searching for such a theory, but failed repeatedly.  The more recent heirs to this fruitless search, are the string theorists.

http://en.wikipedia.org/wiki/Unified_field_theory

Quote from: Bloody Stupid Johnson;530598Possibly the need to define stats in real terms is why people tend not like stuff like e.g. Blue Rose where they dropped the normal D&D stat scale and just used the modifiers...

4E D&D is effectively like this.  :banghead:

Quote from: Bloody Stupid Johnson;530598I don't know that I can give up on searching for the One True Perfect System

Game design is engineering. And, like all engineering, it involves choices, consequences, and compromises. And each choice you make constrains your future choices.

There is no perfect car, perfect CPU, or perfect building. Such things do not exist, because optimizing for one strength means compromising elsewhere.

The same goes for RPG's. There is no perfect system. All you can do is define your needs, and build a system that meets them.

Quote from: Bloody Stupid Johnson;530598I think your answer is interesting because the emphasis it puts on 'defining a link between real-world performance' and the scale. That's definitely something I should be considering more

Well, my argument was that you should be considering it less, because optimizing for that places severe constraints on your other choices. IMHO, its better to chose interesting and fun core mechanics, then allow them to determine the scale. Once you have the scale, then describe how it relates to  real world performance.

Quote from: Bloody Stupid Johnson;530598I'd previously though of 'best' more in terms of mechanical convenience (minimizing table lookups and modifiers, ease of recalculating ability damage)

Those seem like good places to start. JMHO.

In principle, creating a game system based strictly on the mathematics of the mechanics (ie. dice, grids, etc ...) and devoid of any resemblance or connection to the real world, is actually quite easy.

Connecting such a mathematical system to any resemblance or indirect/direct connection to the real world, is the hard part.

For example in the mundane real world offline, "balance" is actually quite rare in practice.

Quote from: Daddy Warpig;530988Game design is engineering. And, like all engineering, it involves choices, consequences, and compromises. And each choice you make constrains your future choices.
 
There is no perfect car, perfect CPU, or perfect building. Such things do not exist, because optimizing for one strength means compromising elsewhere.
 
The same goes for RPG's. There is no perfect system. All you can do is define your needs, and build a system that meets them.

Well true, although I've never seen an RPG that couldn't be improved :)
Often there are shortcuts that will at least improve performance, that are overlooked.
 
Even in engineering there are cases where experts have said something couldn't be done, but they eventually found a way (Henry Ford reportedly paid engineers to just keep thinking of ways to cast engine blocks as a single unit, rather than cylinder by cylinder, despite their protestations, and it paid off). Getting energy out of atoms was decried as "talking moonshine" by Ernest Rutherford (the "father of nuclear physics") before they actually did it.
 
The choices is the hard thing. In the past to do stuff like trying to pick which is the best mechanic, I've done stuff like listing out 100 or so specific rules that I really liked, giving each of them scores for how easily they can be applied to a given core mechanic, then adding up to get the highest number and find the 'best' core mechanic (if not the perfect one). Last time I tried it, it didn't really work though since I eventually found that I'd made a few wrong assumptions and whatnot...
 
Anyway, everyone needs a hobby. I can always convert to Cheetoism on my deathbed like the guy in The Last Samurai...

Quote from: Bloody Stupid Johnson;531078The choices is the hard thing. In the past to do stuff like trying to pick which is the best mechanic, I've done stuff like listing out 100 or so specific rules that I really liked, giving each of them scores for how easily they can be applied to a given core mechanic, then adding up to get the highest number and find the 'best' core mechanic (if not the perfect one). Last time I tried it, it didn't really work though since I eventually found that I'd made a few wrong assumptions and whatnot...

I'm sure many hardcore gamer types have done something like this several times over the years.  Even if it was just done completely mentally, and never formally written down on paper.
 

Quote from: Bloody Stupid Johnson;531078Anyway, everyone needs a hobby. I can always convert to Cheetoism on my deathbed like the guy in The Last Samurai...

I'm slowly on my way out of the tabletop rpg hobby.  (I've come to the conclusion that no game is better than playing with groups of incorrigible dysfunctional players).

Haven't found a replacement hobby yet.  (Not even watching television or crappy movies).

Quote from: ggroy;531150I'm sure many hardcore gamer types have done something like this several times over the years.  Even if it was just done completely mentally, and never formally written down on paper.
 

Ah well thanks, I was worried I was sounding like some sort of basket case...

QuoteI'm slowly on my way out of the tabletop rpg hobby.  (I've come to the conclusion that no game is better than playing with groups of incorrigible dysfunctional players).

Haven't found a replacement hobby yet.  (Not even watching television or crappy movies).

Sorry to hear it.  Good groups seem hard to find, unfortunately (My bunch aren't bad - if a bit D&D centric - having an SO who's also a gamer helps alot with this). I have no helpful advice on new hobbies, though.

Quote from: Bloody Stupid Johnson;531151Ah well thanks, I was worried I was sounding like some sort of basket case...

Back in the day, I use to buy different rpg books and box sets, largely searching for the "perfect" system.  Never was able to find such a "holy grail".

After adding in my own modifications and designing systems from scratch, I was never able to find any holy grails.

Quote from: Bloody Stupid Johnson;531151Sorry to hear it.  Good groups seem hard to find, unfortunately (My bunch aren't bad - if a bit D&D centric - having an SO who's also a gamer helps alot with this). I have no helpful advice on new hobbies, though.

The group I play one-shot evening rpg games with occasionally (ie. once every few months), are not the types I would want to play a weekly campaign with.

Ever since I got back into rpg games (shortly after 3.5E was released), I have not been able to find a steady long term gaming group.  A part of this is due to moving a few times.  In a new town or even in a different area of a large city, one essentially has to start over again.  After awhile of living in a particular location, one gets to know who most of the local players are (along with their quirks and vices).

Quote from: ggroy;530554For example, if one attempts to translate the system of mechanics used in a video game like World of Warcraft, most likely it would look like something which makes Rolemaster or 3.xE/d20 look "simple" in comparison.

I dunno.  I think I did pretty well. (http://drumsofwar.wikia.com/wiki/Drums_of_War_Wiki)

Quote from: J Arcane;531177I dunno.  I think I did pretty well. (http://drumsofwar.wikia.com/wiki/Drums_of_War_Wiki)

What I had in mind, was literally doing a 1-to-1 translation of the mechanics from the video game to a pen-and-paper rpg format.  This could be like taking the actual computer code, and translating it directly to dice mechanics.

Quote from: ggroy;531184What I had in mind, was literally doing a 1-to-1 translation of the mechanics from the video game to a pen-and-paper rpg format.  This could be like taking the actual computer code, and translating it directly to dice mechanics.

Ahh, well, yes, that would suck dickface.  I just wanted to be a smartarse.  ;)

I actually DID write a game like that once, when I was much younger, converting a CRPG I wrote into a tabletop game.  Good setting, but man, the game mechanics would never have flown in the real world.

Quote from: J Arcane;531192I actually DID write a game like that once, when I was much younger, converting a CRPG I wrote into a tabletop game.  Good setting, but man, the game mechanics would never have flown in the real world.

Which computer game?

Quote from: ggroy;531194Which computer game?

It was an unpublished game on a computer already dead when I wrote it, called "Bane of the DragonSun."

I was like 12 when I made the original game.

Quote from: J Arcane;531199It was an unpublished game on a computer already dead when I wrote it, called "Bane of the DragonSun."

I was like 12 when I made the original game.

The only computer game I ever tried to dissect back in the day, was "sword of fargoal" on a Commodore Vic20.

http://en.wikipedia.org/wiki/Sword_of_Fargoal

It turned out the Vic20 version of "sword of fargoal" was written in Basic, and the program code was readable by the "list" command.  Though it was a long enough program, that it was not easy to follow.

Quote from: ggroy;531201The only computer game I ever tried to dissect back in the day, was "sword of fargoal" on a Commodore Vic20.

http://en.wikipedia.org/wiki/Sword_of_Fargoal

It turned out the Vic20 version of "sword of fargoal" was written in Basic, and the program code was readable by the "list" command.  Though it was a long enough program, that it was not easy to follow.

Sword of Fargoal is good stuffs.  They made a Windows remake some time back, and it's fantastic.  Hard to find now though, the creator took it down so he could sell some awful iOS port instead.

There's an open source version of the game code.  Haven't tried running it.

Awhile ago I looked through the computer code of some old Diku muds from the early 1990's.

It turned out the underlying combat mechanic was THAC0, with some slight modifications.

Quote from: ggroy;531184What I had in mind, was literally doing a 1-to-1 translation of the mechanics from the video game to a pen-and-paper rpg format.  This could be like taking the actual computer code, and translating it directly to dice mechanics.

I've seen a number of RPGs where it looked like it would be playable, perhaps even good, if it was the basis for a cRPG, but as pen-and-pen just wasn't human-useable.
World of Synnibarr, Legacy (if the rpg.net thread is to be believed), or Sword's Path Glory all come to mind. There's probably more out there.

Quote from: Bloody Stupid Johnson;530796If you can describe it, what's the appeal of just having 'ranks' like FUDGE/FATE/MSH (/Amber), rather than numbers?

Actually, I don't like the ladder in Fudge/Fate all that much. Those are only adjectives and really are only hard to remember.

Good descriptors IMO tell you what kind of character or entity would have this level. For example in my little game, there is a scale:

0 - Common People
1 - Well-trained
2 - Experienced / Veteran
3 - Specialist, recognized by his/her peers
4 - Widely known master
5 - Idol: Every kid knows you
6 - Legend: Your name will be known for generations
7 - Transcendent

Clearly, this is for a rather heroic game, but there might be appropriate descriptors and referents for other styles as well. The advantage is that it's easy to determine what level a certain character should have.

Quote from: 1of3;531217Actually, I don't like the ladder in Fudge/Fate all that much. Those are only adjectives and really are only hard to remember.
 
Good descriptors IMO tell you what kind of character or entity would have this level. For example in my little game, there is a scale:
 
0 - Common People
1 - Well-trained
2 - Experienced / Veteran
3 - Specialist, recognized by his/her peers
4 - Widely known master
5 - Idol: Every kid knows you
6 - Legend: Your name will be known for generations
7 - Transcendent
 
Clearly, this is for a rather heroic game, but there might be appropriate descriptors and referents for other styles as well. The advantage is that it's easy to determine what level a certain character should have.

Took me a while puzzling over how this differed from FUDGE, exactly. After I while I'm think I get that your descriptions are more about describing how common a value should be (Rare, Very Rare- ish) rather than about what it can do (Good, Great, Superb)...which are perhaps a shade or two vaguer.
?

Also, eh, my own attempt at answering the question. I think a main point I'm struggling with is how to minimize the math while still being able to handle attribute damage and the like well, actually. High values are better for handling attribute damage I think, but at the same time having a big stat range means modifiers that force recalculating and stuff. (I don't really like roll-under either, which is also limiting my options).
 
　
3-18 (or something) - works fairly easily with roll under (e.g. d20 or 3d6). Works with d100 using [Stat x 5]. Works as [d20+bonus] if you go via a modifier chart...this doubles the quantity of numbers for stats on the character sheet, and producing points where the bonus doesn't go up (bad in point buy)...although depending on its scaling, use of the modifier can minimize how often bonuses end up being added to rolls, and mitigates how high the bonuses are. Stat follows a 'bell curve' which gives an idea of how rare high stats are. Scale is fairly intuitive for D&D players, although years of rolling with 4d6-lowest, etc. methods builds up the idea of what a "good" stat is for lots of players.
THis method is OK, although I have a hard choice picking between it and just using the modifier (-5 to +5).
 
4-14, 1-15; the scale for a couple of d20 roll under type systems (e.g. Alternity; Senzar). Setting human limit at 14 or 15 makes it difficult for PCs to break the scale and get checks that 90% likely to succeed. Seems suited to a "you are guys who are special, but not superhuman" style of play. Best with roll-under, could use with a modifier system though (modifier = [stat-10] which is at least no table.
TORG [5-13] isn't far off this,with its system using 2d10; bell curviness here changes the impact at the far ends of the scale so having 13 or so is a bigger deal than for Alternity.
 
Up to 100 (or so) scale: tried this with e.g. d10+ 10s place as modifier. Math is pretty simple, has problems with min/maxing "dead spaces" even more than than 3E/4E D&D does though.
 
Up to 30 (Paladium): allows for more superhuman characters, needs modifier chart to handle adjustments. Doesn't really do attribute checks that well (I've used roll-under on d30 as a house rule..). Same notes apply for modifiers as with 3-18.
 
d4/d6/d8/d10/d12: works only for systems where the main dice rolled on a check varies (which are themselves a bit quirky; so probably not what I'm after). Numbers generated don't really convert into "real world" measurements readily - how much someone could lift for instance, seems more arbitrary. Maybe its just me but since a high stat doesn't give an increase in minimum roll, it seems logical that anyone could try and do anything, but may screw themselves when they roll a 1. Don't really like this one.
 
1-10: works with dice pool games (though gives large dice pool), or [d10+stat], or [2d6+stat], roll under with d10, etc.
Ability damage may work OK. Tendency to use simple calculations [Str x N] means in these systems, character abilities often vary wildly from one person to the next.
Works for human-level games but something in these systems bothers me for rating "superhuman" creatures like dinosaurs or dragons??? I can't figure out why but having a stat of 30 in a 1-10 system just bothers me more than having a 50 in a 3-18 scale would.
 
-5 to +5 or so (Talislanta, True20, Ars Magica): works with e.g. d10+modifier, d20+modifier. Works OK with ability damage. The point representing a stat being useless can be adjusted (-3 to +3 for instance). Modifiers apply to basically every roll, however - unless system is designed so that a normal score is zero. Works OK with level advancement systems.
 
1-5: works with e.g d10+stat+skill, or d10+stat+skill, or dice pool. Very granular; handles ability damage badly. Midpoint is 3, but an average score being a "2" is also often used. Skills also end up being very granular, meaning "level advancement" is tricky - so games like this don't tend to use levels.

Quote from: Bloody Stupid Johnson;531248Took me a while puzzling over how this differed from FUDGE, exactly. After I while I'm think I get that your descriptions are more about describing how common a value should be (Rare, Very Rare- ish) rather than about what it can do (Good, Great, Superb)...which are perhaps a shade or two vaguer. ?

Yes. Although I wouldn't make commonality the criterium. It's more about whether I can determine the referent. For example I suggested some of the following to Malmsturm (a German game):

- Superhuman
- Supernatural
- Otherworldly
- Divine

Those do not tell, how common these characters are in a given scenario. - There might gods everywhere. - But they do give an impression about possible referents. A superhuman trait cannot be found in an ordinary human, but perhaps in another natural animal, while a supernatural trait cannot be achieved without magic etc.


EDIT/addition:

When I needed stats for space ships, I wanted them to be very simple and employ a single ressource: Energy. How much energy would each ship have?

0-2 Pod
3-5 Shuttle
6-8 Yacht
9-11 Capital Ship
12-14 Space Center

Quote from: Bloody Stupid Johnson;531208I've seen a number of RPGs where it looked like it would be playable, perhaps even good, if it was the basis for a cRPG, but as pen-and-pen just wasn't human-useable.
World of Synnibarr, Legacy (if the rpg.net thread is to be believed), or Sword's Path Glory all come to mind. There's probably more out there.

Do they look anything like the Dragon Age video game mechanics?

http://dragonage.wikia.com/wiki/Combat_mechanics
http://dragonage.wikia.com/wiki/Ability_mechanics

http://dragonage.wikia.com/wiki/Category:Game_mechanics

Quote from: 1of3;531254Yes. Although I wouldn't make commonality the criterium. It's more about whether I can determine the referent. For example I suggested some of the following to Malmsturm (a German game):
 
- Superhuman
- Supernatural
- Otherworldly
- Divine
 
Those do not tell, how common these characters are in a given scenario. - There might gods everywhere. - But they do give an impression about possible referents.

OK I think that makes sense.
Interesting in that I think an adjective system seems oriented to be more subjective or context-dependent while numbers are more objective and let you compare between populations that are unalike. So if a character was described as a "widely known master" at Running, or a Great runner, this wouldn't tell you how fast they were compared to a car or a horse, the way a number would.
(Though adjectives typically need numbers alongside them anyway, in order to work with game mechanics...and maybe numbers don't give much context on whether they're good/bad unless the reader is aware of how they're generated and other rules that interlock with them like probabilities etc. To me a "16" for a stat in D&D is more descriptive than a rank title would be, but that'd totally be a YMMV thing).

Quote from: ggroy;531303Do they look anything like the Dragon Age video game mechanics?
 
http://dragonage.wikia.com/wiki/Combat_mechanics
http://dragonage.wikia.com/wiki/Ability_mechanics
 
http://dragonage.wikia.com/wiki/Category:Game_mechanics

I'll have to peruse this a bit more to really tell you, but from a glance Sword's Path Glory looks a bit look that -excessive calculations and tables. Perhaps SPG is more complex than Dragon Age...
Some discussion here.
http://www.therpgsite.com/showthread.php?t=21194
 
I was perhaps exaggerating a bit with Synnibar - I have run it although handwaving movement and per-segment damage a bit...its damage system is fairly ornate though (armour has a finite reserve of Life Points, and divides incoming damage by from /10 (1 "tenth") to /100000000 (8 "tenths"). It has an unusual level structure for a pRPG, which goes from 1- 50 with slowly increasing percentages -- a bit like an MMO in that regard.
 
I'll try and dig up the link to the Legacy thread; this was a 70s simulational RPG about cavemen. Probably more like Civilization or a game like that, than a real RPG; you could have dozens or more of d100 or d1000 rolls (IIRC) to see if anyone in the tribe had managed to evolve.

EDIT: And here's the link for Legacy:
http://forum.rpg.net/showthread.php?359401-Legacy-the-worst-game-every-written-or-the-least-the-unintentionally-funniest! (http://forum.rpg.net/showthread.php?359401-Legacy-the-worst-game-every-written-or-the-least-the-unintentionally-funniest%21)

Thought for today on 3.x attributes, which is close enough to the topic of this thread that I thought I'd post in it rather than making a new one.

One of the problems with 3.x that irks me is the +1 every 2 stat points thing - since it makes even numbers useless. My thought was to have two modifiers which increase alternately. So for instance we have:

Strength - stamina and muscle
Dexterity - Aim and Balance
Constitution - Health and Fitness
Intelligence - Reason and Memory
Wisdom - Intuition and Willpower
Charisma - Appearance and Leadership

For instance a character with Strength 15 might have a Stamina +3 modifier, Muscle +2 modifier (each of which applies to different sets of actions..e.g. opening a door might be Muscle while Swim uses Stamina).

(look familiar? :) )

Of course, that's 3 numbers on the sheet for each stat instead of 2...still not sure its not better e.g. just to dump the stat and have a stat being the modifier as in True20...

The trick is to use stat checks for the skill system.  That way the actual number stops being a useless vestige, and starts being useful, and every number in the range actually counts for something.

Of course, it also means you need to put a sane cap on stats instead of this whole "I have 47 STR" nonsense you can get to with some 3e builds.

Quote from: J Arcane;531869The trick is to use stat checks for the skill system.  That way the actual number stops being a useless vestige, and starts being useful, and every number in the range actually counts for something.

Of course, it also means you need to put a sane cap on stats instead of this whole "I have 47 STR" nonsense you can get to with some 3e builds.

Hmm...so how would you do that? Roll under stat on d20 for skills, like in 2E?

Quote from: Bloody Stupid Johnson;531876Hmm...so how would you do that? Roll under stat on d20 for skills, like in 2E?

That's exactly what I did in Hulks and Horrors.

Quote from: Bloody Stupid Johnson;531868(look familiar? :) )

Of course, that's 3 numbers on the sheet for each stat instead of 2...still not sure its not better e.g. just to dump the stat and have a stat being the modifier as in True20...

Hard part is figuring out how much granularity one would want, to replicate a particular style of play.


More generally, stat type systems are terrible for covering stuff which are subjective.  Though within the scope of a particular game, such subjective stats in a system may be adequate enough for expediency.  (For example, INT, WIS, CHA in D&D/d20 style games).