Working on a new game

Post by Scooter the Mighty
It seems to me like there are two varieties of skills, one in which you
pretty much always have to roll to see if you're successfull (like
hitting someone with a mace) and the type where it's usually enough to
just have the skill.

We came to the same conclusion.

Post by Scooter the Mighty
Because there is a lot of gray area (what if the carpenter is trying to
repair a boat while shipwrecked on an island with limited tools?), I
thought I would give each skill both a percentage chance of success,
and a "rating." A master carpenter would be considered to create
consistantly higher quality work than a beginner, but neither would
have do a lot of rolling of dice.

I don't like this approach, because I find that if you give the
skill a percent chance just like other skills and then never roll
against it, the player may feel cheated. This can also interact
badly with your advancement mechanics--if the player has to pay to
increase skills, he's unlikely to increase this kind, but then
the numbers end up being meaningless when you do need them.

Instead, I'd just give this type of skill a rating, and in the rare
cases where you need a roll, base it on a stat or another skill
and then interpret it in terms of the rating. A carpentry success
for a journeyman would be "it holds together" and for a master
"it's excellently crafted."

Post by Scooter the Mighty
What sort of percentage chance of success should a skill that you don't
normally have to roll against have? What should it represent? You're
chance to succeed in some sort of unspecified "difficult" situation?
Given that most difficult situations are going to come with a modifier,
what should the concept of "baseline" be?

This is my beef with percent-success mechanics in general--it forces
you to think about "average" task, which is hard to generalize
across skills in a good way. I do better thinking in different
terms: "Here is a bonus to your roll. The difficulty of the roll
is set by the difficulty of the task." That is, don't have the
skill establish the target number, have the task establish it.

Mathematically I think these are the same, but psychologically
there's a big difference between "The GM states the task difficulty
and the player uses skill to modify her roll" and "The player
knows that she has a 87% chance of success and the GM modifies
that to indicate difficulty." In my experience, the latter
discourages modification--it seems appropriate only in exceptional
circumstances--and leads to sticky problems like yours.

I'd also really argue against having the baseline for a boolean
skill be different from that for a normal skill--it's just
confusing. I'd either make them the same, or get rid of success
chance for boolean skills at all.

Mary Kuhner ***@eskimo.com

Scooter the Mighty

2006-07-27 20:24:01 UTC

We came to the same conclusion.

I don't like this approach, because I find that if you give the
skill a percent chance just like other skills and then never roll
against it, the player may feel cheated. This can also interact
badly with your advancement mechanics--if the player has to pay to
increase skills, he's unlikely to increase this kind, but then
the numbers end up being meaningless when you do need them.
Instead, I'd just give this type of skill a rating, and in the rare
cases where you need a roll, base it on a stat or another skill
and then interpret it in terms of the rating. A carpentry success
for a journeyman would be "it holds together" and for a master
"it's excellently crafted."

Hmm...I tend to agree in that I'd really rather just go with a rating,
but I'm not sure that basing the roll on another skill or craft would
work. For one thing, there wouldn't always be a good skill to base it
on, or if there was the character might not have it and be pissed that
his Master carpenter couldn't do woodworking in tough circumstances.
Basing it on a stat would imply that anyone could do it. Although I
suppose I could base it on a stat and say that if you don't have the
skill you can't do it.

This is my beef with percent-success mechanics in general--it forces
you to think about "average" task, which is hard to generalize
across skills in a good way. I do better thinking in different
terms: "Here is a bonus to your roll. The difficulty of the roll
is set by the difficulty of the task." That is, don't have the
skill establish the target number, have the task establish it.
Mathematically I think these are the same, but psychologically
there's a big difference between "The GM states the task difficulty
and the player uses skill to modify her roll" and "The player
knows that she has a 87% chance of success and the GM modifies
that to indicate difficulty." In my experience, the latter
discourages modification--it seems appropriate only in exceptional
circumstances--and leads to sticky problems like yours.

I like that idea.

Post by Mary K. Kuhner
I'd also really argue against having the baseline for a boolean
skill be different from that for a normal skill--it's just
confusing. I'd either make them the same, or get rid of success
chance for boolean skills at all.

Let me play around with that.

Erol K. Bayburt

2006-07-29 16:53:44 UTC

On Thu, 27 Jul 2006 11:52:36 +0000 (UTC),

We came to the same conclusion.

I like the way d20/3e D&D uses "take 10" for this sort of situtation.
The upside is that it allows all skills to be converted into skills of
the second sort. The downside, however, is that when taking 10 isn't
good enough, the chance of success drops from 100% chance of a bare
success to a <50% chance of success.

Post by Mary K. Kuhner
Instead, I'd just give this type of skill a rating, and in the rare
cases where you need a roll, base it on a stat or another skill
and then interpret it in terms of the rating. A carpentry success
for a journeyman would be "it holds together" and for a master
"it's excellently crafted."

It helps, IME, if degree-of-success is built deeply into the system,
at least for certain skills. In d20-speak, if the quality of the item
is based on the DC of the craft skill check needed to make it, and if
that quality then affects how well the item works.

In the case of the shipwrecked carpenter, the carpenter's skill + the
difficulty of the working conditions determines the quality of the
boats repairs. And the quality of those repairs then determines how
well the boat stands up in Bad Conditions.

I'd say that the "average" (+0 modifier) task for a given skill has to
be part of the definition of the skill. Also it's helpful for
modifiers to be as easy to apply as possible and fairly chunky.

For example, I have all situtational modifiers set in units of 10
percentage points in my "Percentile HERO" houserules for the HERO
system. +10% is close enough to +1 in HERO's original 3d6 system that
they can be equated, and multiples of 10% are much easier to work with
in the heat of play than multiples of, say, 8% or 5%.

Post by Mary K. Kuhner
Mathematically I think these are the same, but psychologically
there's a big difference between "The GM states the task difficulty
and the player uses skill to modify her roll" and "The player
knows that she has a 87% chance of success and the GM modifies
that to indicate difficulty." In my experience, the latter
discourages modification--it seems appropriate only in exceptional
circumstances--and leads to sticky problems like yours.

Another advantage of "The GM states the task difficulty and the player
uses skill to modify her roll" is that it tends be easier to calculate
degree-of-success results that way. E.g. the GM can state the
difficulties for both success and for superior-success, and the player
can roll to see which, if either, he gets.

Or for our shipwrecked carpenter, the GM might say "The boat will gain
+1 structure point from your repairs for every 10% by which you beat
difficulty X."

--
Erol K. Bayburt
***@aol.com

Peter Knutsen (usenet)

2006-07-30 09:29:21 UTC

Post by Erol K. Bayburt
I like the way d20/3e D&D uses "take 10" for this sort of situtation.
The upside is that it allows all skills to be converted into skills of
the second sort. The downside, however, is that when taking 10 isn't
good enough, the chance of success drops from 100% chance of a bare
success to a <50% chance of success.

As I mentioned in another post a few minutes ago, this issue gradually
disappears as you increase the number of dice rolled. The entire "take
10/20" mechanic is indicative of D&D3's designers being aware of how
problematic their flat distributon roll mechanic is.

It helps, IME, if degree-of-success is built deeply into the system,

I agree about degree-of-success mechanics in general. As formal as
possible. "GM interpretation" mechanics in which the GM is supposed to
let the result be more favourable the better the roll is, gives the GM
far more power than he needs in order to do his job.

One problem, however, is that degree-of-success mechanics often end up
involving numerical ranges which must then be *memorized*. That's a
burden upon both the GM and the players.

My designs, both Sagatafl and the Action Movie RPG, avoid this problem
because all the "ranges" are exactly 1 unit wide. 1 Success is inferior
to 2 Successes which is inferior to 3 Successes.

Such an approach is much better than requiring the GM and the players to
memorize value ranges such as "If the roll is made exactly, it is a
Partial success. If the roll is made by 1 to 3, the roll is a Normal
Success. If the roll is made by 4 to 7, it is a Great Success. If the
roll is made by 8 or 9, it is a Superb Success. If..."

Unfortunately, it works only with multiple-dice mechanics. You could try
it in a 1d6 or 2d4-based system, but I wouldn't recommend it. It would
be ugly.

Post by Erol K. Bayburt
at least for certain skills. In d20-speak, if the quality of the item
is based on the DC of the craft skill check needed to make it, and if
that quality then affects how well the item works.

When it comes to skills to create things in particular, I'm not sure I
want it to be possible for characters to accidentally create something
of a higher quality than what they were aiming for. There are some
processes, such as swordsmithing, where it may actually be realistic
(when metallurgy is little understood, there's apparently a quite random
element in the process), but they are the exception.

Post by Erol K. Bayburt
In the case of the shipwrecked carpenter, the carpenter's skill + the
difficulty of the working conditions determines the quality of the
boats repairs. And the quality of those repairs then determines how
well the boat stands up in Bad Conditions.

[...]

Post by Mary K. Kuhner
This is my beef with percent-success mechanics in general--it forces
you to think about "average" task, which is hard to generalize
across skills in a good way. I do better thinking in different
terms: "Here is a bonus to your roll. The difficulty of the roll
is set by the difficulty of the task." That is, don't have the
skill establish the target number, have the task establish it.

I'd say that the "average" (+0 modifier) task for a given skill has to
be part of the definition of the skill. Also it's helpful for
modifiers to be as easy to apply as possible and fairly chunky.
For example, I have all situtational modifiers set in units of 10
percentage points in my "Percentile HERO" houserules for the HERO
system. +10% is close enough to +1 in HERO's original 3d6 system that
they can be equated, and multiples of 10% are much easier to work with
in the heat of play than multiples of, say, 8% or 5%.

In the Action Movie RPG, the standard modifiers are +3, none, and -3.
The GM is not in any way told to not use other modifiers when his "sense
of realism" tells him to, but those are the standard ones, the ones
labeled "in case of doubt, use these".

In Sagatafl, the standard Roll Difficulties are 6, 8 and 10. In both
systems, they cover actions that are almost routine, that are difficult
(standard adventuring conditions roll), or that are very difficult.

Another advantage of "The GM states the task difficulty and the player
uses skill to modify her roll" is that it tends be easier to calculate

Doing so also prevents the GM from applying secret modifiers which the
players are never told about, as can easily be the case when the players
are expected to simply relay to the GM the degree to which they made the
roll, e.g. "I rolled 3 under my skill level", which the GM then
translates into the actual world event.

Post by Erol K. Bayburt
degree-of-success results that way. E.g. the GM can state the
difficulties for both success and for superior-success, and the player
can roll to see which, if either, he gets.

That's quite a bit of information required of the GM, two numbers, but
it is not excessive. Just don't make it 3 or 4 numbers. That'd be too much.

Post by Erol K. Bayburt
Or for our shipwrecked carpenter, the GM might say "The boat will gain
+1 structure point from your repairs for every 10% by which you beat
difficulty X."

--
Peter Knutsen
sagatafl.org

Scooter the Mighty

2006-07-30 18:46:25 UTC

Post by Peter Knutsen (usenet)

It helps, IME, if degree-of-success is built deeply into the system,

I agree about degree-of-success mechanics in general. As formal as
possible. "GM interpretation" mechanics in which the GM is supposed to
let the result be more favourable the better the roll is, gives the GM
far more power than he needs in order to do his job.
One problem, however, is that degree-of-success mechanics often end up
involving numerical ranges which must then be *memorized*. That's a
burden upon both the GM and the players.
My designs, both Sagatafl and the Action Movie RPG, avoid this problem
because all the "ranges" are exactly 1 unit wide. 1 Success is inferior
to 2 Successes which is inferior to 3 Successes.
Such an approach is much better than requiring the GM and the players to
memorize value ranges such as "If the roll is made exactly, it is a
Partial success. If the roll is made by 1 to 3, the roll is a Normal
Success. If the roll is made by 4 to 7, it is a Great Success. If the
roll is made by 8 or 9, it is a Superb Success. If..."

I'm actually thinking about getting away from the dice rolls as much as
possible. Unless people are contesting against each other or there is
a harsh time limit so that only one attempt is possible, I don't see
that rolls are necessary. I'd like to assume that Master Carpenters do
Master level work. If they screw up, they'll just try again anyway, so
why bother rolling a lot of dice?

Simon Smith

2006-07-27 20:50:02 UTC

Post by Scooter the Mighty
I'm starting to work on a new frp game, and wouldn't mind hearing a few
opinions. My motivation is that I find myself liking the idea of d20
(a game in which anyone can make supplements) but I don't really like
the system all that much, so I thought I'd make my own system and post
it on the net. I don't really think it'll seriously challenge d20,
I'll be lucky if a few people read it and pass it around. However, I'm
having fun with it.
Anyway, the system is going to be a "build your character from points"
type of system, a bit like GURPS. I'm thinking about how I'm going to
work skills.
It seems to me like there are two varieties of skills, one in which you
pretty much always have to roll to see if you're successfull (like
hitting someone with a mace) and the type where it's usually enough to
just have the skill. For example, I usually wouldn't bother to make a
carpenter roll to make a wooden box, because a) it's something
carpenters can usually just do, and b) if they mess up they can just
try again, so why waste time on the boring minutia of whether they
forgot to measure twice and cut once the first time?
Because there is a lot of gray area (what if the carpenter is trying to
repair a boat while shipwrecked on an island with limited tools?), I
thought I would give each skill both a percentage chance of success,
and a "rating." A master carpenter would be considered to create
consistantly higher quality work than a beginner, but neither would
have do a lot of rolling of dice.
Does this just sound stupid?
What sort of percentage chance of success should a skill that you don't
normally have to roll against have? What should it represent? You're
chance to succeed in some sort of unspecified "difficult" situation?
Given that most difficult situations are going to come with a modifier,
what should the concept of "baseline" be?

I would suggest treating 'don't need to roll' tasks as routine tasks, and
giving them a 99.9% or even 100% success rate.

A useful benchmark is to pick, say, three distinct levels of skill - lemon,
professional, master, assign the skill levels you want them to have, and
then structure the system such that a pro succeeds at a standard task, oh,
90% of the time, a master never fails, and a lemon has a 50-50 chance. If
you grade the difficulty levels throughout the system according to how a
trained professional would view them, it makes it a lot easier to choose the
right numbers to get the feel you want. You can also define the
probabilities of 'lemon beats pro', 'lemon beats master', 'pro beats master'
(and the reverse cases) and ensure that the odds the system generates are
the odds you want. In a system with a heroic tone, lemons have a better
chance of being lucky; in a 'realistic' system, their chances would probably
be lower (but still non-zero, I would submit).

Two other tips for better representing the differences between a journeyman
and a master:

1. Allow the master to choose whether he wishes to surpass the journeyman's
best possible work, /or/ produce work of acceptable quality faster. In other
words he can sacrifice quality for speed in those cases where the work
merely needs to be 'good enough' and finishing quickly is more useful.

2. Try to ensure that characters become less erratic with increasing skill.
To borrow the DnD term with which you are familiar, allow characters the
option of 'taking 5', 'taking 10', 'taking 15', 'taking 20', and so on with
increasing skill. That's just a suggestion - obviously there are lots of
dice mechanics you could use to produce this sort of effect.

As far as using percentile dice are concerned, I have to confess I'm less
keen on them these days. They can make characters very erratic. A total
novice (say, 10% skill) can beat a master (say, 90% skill) 1% of the time,
and the master has a mere 90% success rate at his own skill rolls, which is
pretty poor in my view.

The cause of the problem is obvious - if you have a skill range spanning
about 100 points, and that's compared against a D100, then the D100 - i.e.
random chance - will contribute about half of a character's chances of
success or failure. For an unskilled character with only 10% skill, random
dice luck pretty much swamps their feeble level of skill. But if you made
skill, say, five times as important as dice luck, that might mean a
character with 60% skill could never hope to beat a character with 81%
skill. Getting the relative contributions of luck and skill right is
difficult, and whatever design decisions you make in this area will
profoundly influence the tone of your game.

Good luck, and do keep us posted :-)

--
Simon Smith

When emailing me, please use my preferred email address, which is on my web
site at http://www.simon-smith.org

Scooter the Mighty

2006-07-28 19:39:06 UTC

Post by Simon Smith
As far as using percentile dice are concerned, I have to confess I'm less
keen on them these days. They can make characters very erratic. A total
novice (say, 10% skill) can beat a master (say, 90% skill) 1% of the time,
and the master has a mere 90% success rate at his own skill rolls, which is
pretty poor in my view.

I guess that's why I'm leaning towards having a category of skills in
which you rarely if ever roll. A diva opera singer is just assumed to
be able to sing well, and if you're skill isn't that high then you just
can't compete.

Post by Simon Smith
The cause of the problem is obvious - if you have a skill range spanning
about 100 points, and that's compared against a D100, then the D100 - i.e.
random chance - will contribute about half of a character's chances of
success or failure. For an unskilled character with only 10% skill, random
dice luck pretty much swamps their feeble level of skill. But if you made
skill, say, five times as important as dice luck, that might mean a
character with 60% skill could never hope to beat a character with 81%
skill. Getting the relative contributions of luck and skill right is
difficult, and whatever design decisions you make in this area will
profoundly influence the tone of your game.

Hmm...Well, that does bear thinking about, although I'm not sure I'm
following your math. How does random dice luck swamp a 10% skilled
character's skill?

The main area in which characters contest with each other is combat,
and right now I'm relying on the fact that mutiple rolls will usually
be required to smooth things out a bit. I've also got 3 kinds of
defense which can be used simultaneously (Dodge, which subtracts from
your opponent's chance to hit, Parry, which gives the target a chance
to block the attack even if it hits, and Armor, which reduces damage),
so it wouldn't be a simple, 60% VS 81%.

Post by Simon Smith
Good luck, and do keep us posted :-)

Thanks!

Post by Simon Smith
Simon Smith
When emailing me, please use my preferred email address, which is on my web
site at http://www.simon-smith.org

Simon Smith

2006-07-28 22:40:22 UTC

In message <***@h48g2000cwc.googlegroups.com>
"Scooter the Mighty" <***@hotmail.com> wrote:

<snip>

The cause of the [dice luck] problem is obvious - if you have a skill range spanning
about 100 points, and that's compared against a D100, then the D100 - i.e.
random chance - will contribute about half of a character's chances of
success or failure. For an unskilled character with only 10% skill, random
dice luck pretty much swamps their feeble level of skill. But if you made
skill, say, five times as important as dice luck, that might mean a
character with 60% skill could never hope to beat a character with 81%
skill. Getting the relative contributions of luck and skill right is
difficult, and whatever design decisions you make in this area will
profoundly influence the tone of your game.

Hmm...Well, that does bear thinking about, although I'm not sure I'm
following your math. How does random dice luck swamp a 10% skilled
character's skill?

Well, I have to admit it partly depends on the dice mechanics you use. But
one generally assumes (whether a set of rules explicitly say so or not) that
a skill roll that's within a few points of success is close, and a skill
roll that's far adrift is a more serious failure. Now, consider a mechanic
like: roll the dice, add the character's skill, and compare to a target
number.

So in the case of the novice we might roll a D100, add 10% skill, and
compare against a target number of 100. This obviously has a 10% success
chance, so it's mathematically identical to rolling a D100 and requiring a
roll of less than or equal to the character's skill. But it also makes it
plain that when the character succeeds, ten points out of the 100 needed for
success were supplied by the character's skill, the other 90 points by the
D100. So in other words, dice luck is nine times more important than skill.
Now imagine another character, only half as skilled as our 10%er. If these
two compete, the contest between them is almost pure luck, even though one
character is twice as skilled as the other.

In some games, it woud be perfectly OK for a contest between two such
bunglers to come down to luck. In other systems, being twice as skilled as
someone else is always supposed to be enough to confer a decisive advantage.
If you want to ensure that characters are never this erratic in your system,
you have make even the novices start out with perhaps a 50-50 chance of
success.

But if a novice has 50% skill, doesn't that mean a master needs at least
95% skill? And how good is a pro? Or do you take a leaf out of RuneQuest's
book, allow skills to exceed 100%, and allow for bonuses and penalties to
apply to bring skills more-or-less back into line with the D100 skill dice?
If you do that, you have to allow for possibility of the final result being
less than one or more than 100. And that in turn means you probably need a
rule that 01 is always a success, 00 always a failure.

The other way is to rule that a 100% skill is perfect, and that means the
experience system has to be non-linear. Characters' skills may increase
rapidly from 40-50%, but that they start to slow down by 95-99%. And if you
do that, it will take an awful lot of effort to get from 97-98% skill, and
yet the in-game difference between 97% and 98% may be barely perceptible.

Post by Scooter the Mighty
The main area in which characters contest with each other is combat,
and right now I'm relying on the fact that mutiple rolls will usually
be required to smooth things out a bit. I've also got 3 kinds of
defense which can be used simultaneously (Dodge, which subtracts from
your opponent's chance to hit, Parry, which gives the target a chance
to block the attack even if it hits, and Armor, which reduces damage),
so it wouldn't be a simple, 60% VS 81%.

That sounds reasonable. However, do watch out for any mechanic that involves
subtracting one character's skill from another's, because you can't assume
that the subtraction will always work out in the expected direction. If you
had a master Dodger (90%) fighting an almost-master swordsman (89% to hit),
the almost-master has the same chances of success as Ug the troll who has
a 5% skill. i.e. None. Is that fair? Do you consider that a desirable trait
in your system?

Another thing to watch for: when playtesting you might determine that a
typical warrior has 75% attack skill, 50% defence, and therefore two typical
warriors fighting will have a 25% chance of hitting each other. Which
might seem perfectly OK. But then along come a couple of defensive warriors
with 60% skill and 80% defence, and they can't hit each other at all!

So my advice is to pay very close attention to the number ranges generated
by interacting skills like this, and ensure that characters only become
invincible when they should be invincible. Your mechanic sounds similar to
how RuneQuest II did it - you had a Defence skill, typically 5-20% for most
characters, which subtracted from opponents' attack skills and rose
relatively slowly. But a character who somehow got a 90% Defence skill would
have strained the game system, rather. Under RuneQuest III, Defence became
Dodge, which you could use instead of - rather than in addition to -
parrying, and if you succeeded at a Dodge roll you completely avoided an
incoming blow. But the skill was easier to improve. This change did mean
that a character could only use one attacking skill and one damage-avoiding
skill in a combat round. Runequest II had one attacking skill, but two
damage-avoiding skills that could be combined against it, and characters
being able to contrive a situation under which two skills can be pitted in
combination against one is a common cause of game balance issues.

--
Simon Smith

When emailing me, please use my preferred email address,
which is on my web site at http://www.simon-smith.org

Scooter the Mighty

2006-07-29 19:00:04 UTC

Post by Simon Smith
<snip>

Hmm...Well, that does bear thinking about, although I'm not sure I'm
following your math. How does random dice luck swamp a 10% skilled
character's skill?

Well, but I think that in such a contest the 10%er is going to win
twice as often as the 5%, so it probably does work in that sense. If
we're defining "twice as talented" as "having twice the chance of
success" this system works just fine.

I think the problem comes in defining talent as a percentage chance of
success. If you take two pool players, and one is twice as good at
shoot balls in the pocket and twice as good at setting up the next
shot, he or she is going to win a lot more than twice as often. But
simulating each step of a skill rapidly becomes tedious. I'm going to
try to think of a way to do an end run around that, probably in part by
using a rating system instead of a percentage system for a lot of
skills.

<snip>

Post by Simon Smith
But if a novice has 50% skill, doesn't that mean a master needs at least
95% skill? And how good is a pro? Or do you take a leaf out of RuneQuest's
book, allow skills to exceed 100%, and allow for bonuses and penalties to
apply to bring skills more-or-less back into line with the D100 skill dice?
If you do that, you have to allow for possibility of the final result being
less than one or more than 100. And that in turn means you probably need a
rule that 01 is always a success, 00 always a failure.

I am planning to allow percentages in excess of 100%. For combat and
maybe other skills as well there will be an assortment of beneficial
techniques that you can use in exchange for taking a penalty to your
chance of hitting, so it will be useful to have a very high skill.

<snip>

Post by Simon Smith

Actually, part of the reason I'm planning on using a 3 part system of
defense is to avoid this kind of thing. I've mostly be playing in the
GURPS system and IMHO it's too easy to get characters with defenses in
the ~90% range. My plan is to use Dodge to help justify not giving
super duper parry scores.

Right now, a person of average stats (that's average for a human, not
for an adventurer, who will have generally higher stats) gets a dodge
of 17%, and a person with maximum stats gets a dodge of 35%. Shields
and perhaps some armor and skills will add to that, but I plan for a
dodge of 50% to be very difficult to achieve. Which I think is pretty
reasonable, in real life a person who could just stand there and jump
out of the way of a spear thrust 50% of the time without either running
away or using their hands to block would be pretty exceptional.

Post by Simon Smith
Another thing to watch for: when playtesting you might determine that a
typical warrior has 75% attack skill, 50% defence, and therefore two typical
warriors fighting will have a 25% chance of hitting each other. Which
might seem perfectly OK. But then along come a couple of defensive warriors
with 60% skill and 80% defence, and they can't hit each other at all!

I agree. I'm planning to not allow 80% dodge. It may be possible for
parries to get that high in the super talented, but a person will only
get a limited number of them per round and there will be moves
available to lower your opponents parry score (albiet at a cost to your
own chances of hitting)

Simon Smith

2006-07-29 20:16:20 UTC

Post by Simon Smith
<snip>

I've snipped most of the workarounds you aready have planned, but there are

Post by Scooter the Mighty
Well, but I think that in such a contest the 10%er is going to win
twice as often as the 5%, so it probably does work in that sense. If
we're defining "twice as talented" as "having twice the chance of
success" this system works just fine.

Yeah, /in the long run/ - i.e. thousands of samples, the numbers work out fine.
In the short term - a dozen rolls - things may feel rather different. Most RPGs
tend not to generate enough samples for long-term trends to become apparent.
It's the short-term randomness that's most noticeable.

<snip>

Post by Scooter the Mighty
Right now, a person of average stats (that's average for a human, not
for an adventurer, who will have generally higher stats) gets a dodge
of 17%, and a person with maximum stats gets a dodge of 35%. Shields
and perhaps some armor and skills will add to that, but I plan for a
dodge of 50% to be very difficult to achieve. Which I think is pretty
reasonable, in real life a person who could just stand there and jump
out of the way of a spear thrust 50% of the time without either running
away or using their hands to block would be pretty exceptional.

If you evade one thrust, then step inside the range of the spear so that
your body blocks the haft and attempt to throttle the spearman, have you
used dodge, parry, movement, or what?

If in subsequent rounds, the spearman is too stupid to drop his spear and
continues trying to attack with it, how would defence, dodges and parries
work?

Given a tighly-packed phalanx of spearmen, can the guys in the middle use
dodge or parry at all?

Hm. You say shields and armour /add/ to your Dodge skill? That doesn't sound
right. At least rename the skill 'Defence' or something, or you'll get some
raised eyebrows from your players, I'd have thought.

BTW, How well does the system handle unusual weapons? I assume it handles
swords and the like OK, but in writing my own weapons rules I have found
the following weapons sometimes troublesome, and often very useful to
consider:

Weapons with very long hafts such as pikes and lances; cavalry charges, and
other weapons used while mounted; whips; impromptu weapons such as bicyle
chains, torn-up trees (if you're a giant), broken bottles, a rifle used as
a club, a pistol used to pistol-whip somebody, a chair or chairleg, an
unconscious opponent (if you're a big strong character); mismatched weapons
between the two sides - e.g. knife against claymore; off-hand weapons -
especially if a critter has more than the usual two hands; weapons with
special characteristics such as lightsabers (cannot be parried, except by
another light saber), flame lances (cannot parry, cannot be parried),
magswords (cannot be parried by lightsabers), characters using missile
weapons in melee - say a broken-off chair leg to parry with, a pistol for
attack, and so on.

<snip>

--
Simon Smith

When emailing me, please use my preferred email address, which is on my web
site at http://www.simon-smith.org

Scooter the Mighty

2006-07-29 23:55:32 UTC

Post by Simon Smith

Post by Simon Smith
<snip>

I've snipped most of the workarounds you aready have planned, but there are

If you evade one thrust, then step inside the range of the spear so that
your body blocks the haft and attempt to throttle the spearman, have you
used dodge, parry, movement, or what?

Evading a thrust would be a dodge, unless you used your hand to push
the spear point out of the way as part of the evasion in which case it
would be both a dodge and a parry. Stepping inside the range of the
spear would be taking advantage of the fact that spears have a minimum
strike distance. It wouldn't be foiling a specific attack so much as
positioning yourself to prevent your opponent from making an attack.

Post by Simon Smith
If in subsequent rounds, the spearman is too stupid to drop his spear and
continues trying to attack with it, how would defence, dodges and parries
work?

Well, if he tries to attack you in a way that is physically impossible,
you wouldn't need to defend at all, you could just laugh. If he wanted
to use his spear haft as a quarter staff and bash you with it, you
would have to defend against it.

In my system, you can always dodge if you are free to move and know the
attack is coming. You get a limited number of parries per round,
usually 1 unless you have some special ability that allows you more or
you're using both a weapon and a shield.

Post by Simon Smith
Given a tighly-packed phalanx of spearmen, can the guys in the middle use
dodge or parry at all?

That would be GM's call. I'd probably allow it at a reduced rate. The
point of a phalanx of spearmen is to keep everyone out of attacking
range, if you met with a phalanx of spearmen with longer spears or a
group of archers you'd be better thinking of something else to do.

Post by Simon Smith
Hm. You say shields and armour /add/ to your Dodge skill? That doesn't sound
right. At least rename the skill 'Defence' or something, or you'll get some
raised eyebrows from your players, I'd have thought.

The idea is something like GURPS 3rd Edition "Passive Defense," the
assumption being that being armor coated means some weapons slide off
even if you don't do a great job dodging. Shields help you dodge by
limiting the angles your opponents can attack from. Heavy armor
wouldn't help you dodge, because the weight slows you down. The
difference between my system and GURPS is that in GURPS Passive Defense
always protects you, even if you're lying unconcious. In my system you
have to be able to move.

Post by Simon Smith
BTW, How well does the system handle unusual weapons?

I don't know yet, I haven't gotten that far. Right now I'm working on
the magic system, and I only have the weapon's list done in a
rudimentary fashion.

Post by Simon Smith
I assume it handles
swords and the like OK, but in writing my own weapons rules I have found
the following weapons sometimes troublesome, and often very useful to
Weapons with very long hafts such as pikes and lances; cavalry charges, and
other weapons used while mounted; whips; impromptu weapons such as bicyle
chains, torn-up trees (if you're a giant), broken bottles, a rifle used as
a club, a pistol used to pistol-whip somebody, a chair or chairleg, an
unconscious opponent (if you're a big strong character); mismatched weapons
between the two sides - e.g. knife against claymore; off-hand weapons -
especially if a critter has more than the usual two hands; weapons with
special characteristics such as lightsabers (cannot be parried, except by
another light saber), flame lances (cannot parry, cannot be parried),
magswords (cannot be parried by lightsabers), characters using missile
weapons in melee - say a broken-off chair leg to parry with, a pistol for
attack, and so on.

Good thoughts. I don't have rules to cover all of those things yet,
although some of them seem pretty straightforward (such as lightsabers:
all you get to do is dodge or run away unless you have a lightsaber or
force shield). I will definitely have rules for trying to get inside
your opponent's reach. Impromptu weapons will usually require a GM
ruling, I'll probably have a skill like "brawling" to cover their use.

Peter Knutsen (usenet)

2006-07-30 09:14:18 UTC

Scooter the Mighty wrote:
[...]

Post by Scooter the Mighty
The main area in which characters contest with each other is combat,

Not necessarily.

Post by Scooter the Mighty
and right now I'm relying on the fact that mutiple rolls will usually
be required to smooth things out a bit. I've also got 3 kinds of
defense which can be used simultaneously (Dodge, which subtracts from
your opponent's chance to hit, Parry, which gives the target a chance
to block the attack even if it hits, and Armor, which reduces damage),
so it wouldn't be a simple, 60% VS 81%.

--
Peter Knutsen
sagatafl.org

Peter Knutsen (usenet)

2006-07-30 08:43:51 UTC

Simon Smith wrote:
[...]

Post by Simon Smith
Two other tips for better representing the differences between a journeyman
1. Allow the master to choose whether he wishes to surpass the journeyman's
best possible work, /or/ produce work of acceptable quality faster. In other
words he can sacrifice quality for speed in those cases where the work
merely needs to be 'good enough' and finishing quickly is more useful.

I think it is vitally important, in order to achieve an improved
player/GM power balance, for an RPG rules system to have a "work faster"
mechanic for regular skills. Not craft skills which is what you are
discussing, but standard adventuring skills such as First Aid and
Lockpicking.

Post by Simon Smith
2. Try to ensure that characters become less erratic with increasing skill.
To borrow the DnD term with which you are familiar, allow characters the
option of 'taking 5', 'taking 10', 'taking 15', 'taking 20', and so on with
increasing skill. That's just a suggestion - obviously there are lots of
dice mechanics you could use to produce this sort of effect.

[...]

A good rule of thumb is that the fewer dice you roll (a single die is
the worst, as seen in d20) the more erratic your performance is.

Also, the fewer times you actually get to try to use a given skill, the
more random an impression will his comrades get of your character's
prowess. In standard D&D, for instance, attack (skill) rolls are made
very frequently, and therefore the difference in combat prowess between
two characters (say a Fighter with the Weapon Focus feat, and a Cleric)
will become obvious to everyone. They'll be keenly aware of who of the
two is best with weapons. On the other hand, rolls to pick locks occur
rarely, and therefore if over the course of half a dozen sessions only
four such rolls have been made, the party's locksmith may very well have
given the other members a very wrong impression about how skilled he is.
If he rolled badly, he will seen as incompetent even though he isn't,
and if he rolled very well, he'll be seen as very skilled even if he is
only mediocre. However, using a less erratic roll mechanic (rolling many
dice instead of only one) serves to mitiate this problem.

--
Peter Knutsen
sagatafl.org

Erol K. Bayburt

2006-07-30 22:37:43 UTC

On Sun, 30 Jul 2006 10:43:51 +0200, "Peter Knutsen (usenet)"

Post by Peter Knutsen (usenet)
A good rule of thumb is that the fewer dice you roll (a single die is
the worst, as seen in d20) the more erratic your performance is.

However:

o A flat distribution has its own advantages, and a multiple-die roll
its own disadvantages.

o What a single die has is a larger variance for a given range. This
isn't always a bad thing.

o The erratic performance from a flat distribution creates more player
and GM tolerance for workarounds (like "take 10"). This can be a good
thing, since while multiple dice make performance less erratic, it can
still leave it too erratic for some situtations and gaming
preferences.

I suspect that it depends on one's taste for auto-failures, critical
hits, fumbles, botches, etc.

For example, standard HERO uses 3d6 for skills etc, and a "15 or less"
gives a ~95% chance of success. In my "Percentile HERO" house rules,
the equivalent of a 15 skill gives a 100% chance of success. I
consider this, the chopping off of the 3d6 "tail," to be a feature.
But I recognize that many gamers (especially GMs) consider it to be a
bug.

--
Erol K. Bayburt
***@aol.com

Scooter the Mighty

2006-07-31 22:52:18 UTC

Post by Erol K. Bayburt
On Sun, 30 Jul 2006 10:43:51 +0200, "Peter Knutsen (usenet)"

Post by Peter Knutsen (usenet)
A good rule of thumb is that the fewer dice you roll (a single die is
the worst, as seen in d20) the more erratic your performance is.

o A flat distribution has its own advantages, and a multiple-die roll
its own disadvantages.
o What a single die has is a larger variance for a given range. This
isn't always a bad thing.
o The erratic performance from a flat distribution creates more player
and GM tolerance for workarounds (like "take 10"). This can be a good
thing, since while multiple dice make performance less erratic, it can
still leave it too erratic for some situtations and gaming
preferences.
I suspect that it depends on one's taste for auto-failures, critical
hits, fumbles, botches, etc.
For example, standard HERO uses 3d6 for skills etc, and a "15 or less"
gives a ~95% chance of success. In my "Percentile HERO" house rules,
the equivalent of a 15 skill gives a 100% chance of success. I
consider this, the chopping off of the 3d6 "tail," to be a feature.
But I recognize that many gamers (especially GMs) consider it to be a
bug.

Besides which, if you want to you can replicate bell curves with
percentile dice. You have to round 3's and 18's up to 1%, and round
other numbers off, but otherwise there's no problem.

1=3
2=4
3-5=5
6-9=6
10-16=7
17-26=8
27-38=9

and so on. Or you can build any curve you'd like into your charts.

On the other hand, you can't replicate a non-bell curve with 3d6.

Simon Smith

2006-08-01 22:31:47 UTC

Post by Erol K. Bayburt
On Sun, 30 Jul 2006 10:43:51 +0200, "Peter Knutsen (usenet)"

Post by Peter Knutsen (usenet)
A good rule of thumb is that the fewer dice you roll (a single die is
the worst, as seen in d20) the more erratic your performance is.

Besides which, if you want to you can replicate bell curves with
percentile dice. You have to round 3's and 18's up to 1%, and round
other numbers off, but otherwise there's no problem.
1=3
2=4
3-5=5
6-9=6
10-16=7
17-26=8
27-38=9
and so on. Or you can build any curve you'd like into your charts.
On the other hand, you can't replicate a non-bell curve with 3d6.

Similarly, a D1000 (D10 for hundreds, tens amd units, 000 = 1000) lets you
approximate the distribution of 4D6. (24 on 4D6 is 1/1,296 so it's not a
perfect match), but personally I only start to prefer multi-die systems to
percentile systems once they can represent extreme probabilities of around
0.1%. At this point their advantage is that the average rolls - say 10-18 on
4D6 - span only a few points, but the extremes of the curve - 4 and 24,
represent one-in-a-thousand flukes. A D1000 does give you even finer
granularity, but the possible rolls span 1-1000 rather than 4-24.

Quibble: Actually, you can use the exact same principle that we commonly use
with D10s to get a non-bell curve out of D6s (or indeed dice of any shape),
although it's impractical in actual use unless all your players are wizards
at mental arithmetic: treat 6 as zero on each die, multiply the first die by
36, the second die by 6, the third die by one. 000 = 216. You can even
produce D40s, D60s, D80s, D120s or D200s by combining D4, D6, D8, D12, D20
with the old standby D10, and with other dice combinations you can produce
most number ranges you might be interested in. In fact, the
D40/D60/D80/D120/D200 trick is about the only die-rolling mechanic I have
yet to see used in an RPG thus far.

--
Simon Smith

When emailing me, please use my preferred email address,
which is on my web site at http://www.simon-smith.org

Peter Knutsen (usenet)

2006-07-30 08:37:58 UTC