Combat Realism vs. Playability

12 views
Skip to first unread message

heckr...@yahoo.com

unread,
Jun 16, 2008, 6:54:58 PM6/16/08
to
I would say that the two aspects of a roguelike, Combat Realism and
Playability, are not directly dependent upon each other but often work
against each other. That is, you could theoretically have a very
realistic game that is very playable, but many of the ideas here in
RGRD tend to support one at the cost of the other. Having realistic
combat reduces playability. Having un-realistic combat reduces the
suspension of disbelief, and otherwise makes the game less fun. The
"solution", in vague and broad terms, is to "Strive for
Balance!" (fence sitting FTW).

Let's focus on the positive for a moment though. Statistical
representation of combat. Dice rolling. It adds randomness to a
fight, which is both fun and realistic. There's all sorts of ways to
screw this up, but as a concept it fits perfectly with both realism
and playability.

Hit points. We've talked about this is a recent post. Hit points are a
game mechanic that allow people to be injured without dying. This is,
if I may go out on a limb, an unrealistic side to games. The
alternatives are somewhat few and far between (yes yes, dwarf
fortress). What would be realistic? Actually, yeah, dwarf fortress is
probably the best stab we have at being realistic. But I'd have stress
that every wound impairs you in some way. The balance? Well, how about
a game that reduces your stats by a factor of your hit points over
your max hit points? Area-specific damage is a leap forward in both
realism AND playability, but, well, it's hard and I'm lazy.

Outputting the effects to the user. Should our combat formula be open
or closed to the player? or, Do we get to see the die rolls? In the
open method the player sees all the math that goes on to determine how
his actions affect the world around him. This support "Power Gaming"
and kind of reduces the game to a spreadsheet. The alternative, with
closed formulas where the player is fed what output you want him to
see. Although, with time* he will deduce the basics of the
calculations and "get a feel" for the game. Jeff and Paul pointed out
that any complicated calculations are meaningless since they're hidden
to the player. This allows for a simple combat system, which is nice
from a development perspective.

I think the open format would be better, although most games are
closed. I'm not sure, I just like to know what is going on. Then
again, I would probably enjoy a game implemented in excel.

Dying means permadeath, and it's a realistic aspect of roguelikes that
I fondly embrace. I think it makes it fun, but I know plenty of people
who would argue that point. But how many of them peruse RGRD?

"Going Epic". In any situation where a human goes up against the
stereotypical dragon and survives, nay, can EXPECT to survive, realism
has been thrown right out the window. But becoming powerful is fun,
and it belongs in the game. I was delighted the first time I got to
the endgame in nethack and realized that my own character was the
embodiment of war. So in this, I have to agree that realism can take
it in the pants.

Limiting abilities is an important factor of realism that shouldn't be
broken. For such games that HAVE ability scores (STR, DEX, INT, ETC)
most of them have some sort of maximum that you can reach and go no
further. This is realism in play. There are limits to how much a man
can lift and how fast he can run.

Magic. Now, by definition, I believe that magic is unrealistic. It's
rather pointless to get into any specific ruleset, so I'm just going
to have a blanket argument that selectivly breaking the rules of
reality is what makes magic systems fun.

Now, I've never played GURPS, and a slew of Roguelikes, so I might be
missing out.
Each and every sentence in this post could be considered flamebait.
Everyone, please try to be nice.
-Heckruler

Paul Donnelly

unread,
Jun 16, 2008, 11:58:33 PM6/16/08
to
heckr...@yahoo.com writes:

> I would say that the two aspects of a roguelike, Combat Realism and
> Playability, are not directly dependent upon each other but often work
> against each other. That is, you could theoretically have a very
> realistic game that is very playable, but many of the ideas here in
> RGRD tend to support one at the cost of the other. Having realistic
> combat reduces playability. Having un-realistic combat reduces the
> suspension of disbelief, and otherwise makes the game less fun. The
> "solution", in vague and broad terms, is to "Strive for
> Balance!" (fence sitting FTW).

I agree. There's no point in going realistic if you wreck your game in
the process. But I think some people are too pessimistic about realism
and just assume that it must be an opaque mess where you usually end up
instakilled anyway.

> Let's focus on the positive for a moment though. Statistical
> representation of combat. Dice rolling. It adds randomness to a
> fight, which is both fun and realistic. There's all sorts of ways to
> screw this up, but as a concept it fits perfectly with both realism
> and playability.
>
> Hit points. We've talked about this is a recent post. Hit points are a
> game mechanic that allow people to be injured without dying. This is,
> if I may go out on a limb, an unrealistic side to games. The
> alternatives are somewhat few and far between (yes yes, dwarf
> fortress). What would be realistic? Actually, yeah, dwarf fortress is
> probably the best stab we have at being realistic. But I'd have stress
> that every wound impairs you in some way. The balance? Well, how about
> a game that reduces your stats by a factor of your hit points over
> your max hit points? Area-specific damage is a leap forward in both
> realism AND playability, but, well, it's hard and I'm lazy.

I think that attention to scale is the key to making a realistic yet
playable combat system. To wit, limiting the number of "wound
effects". Like others have said, Dwarf Fortress can get away with
leaving dwarves permanently maimed and having a complicated wound effect
system because the player doesn't have to bother with it unless they
happen to look at a given dwarf's character sheet. Understanding the
combat abilities of one dwarf is not so important as it would be in a
normal RL because losing one dwarf doesn't lose you the game (except in
adventure mode).

In a RL, I think it's better to simplify. There's no need to precisely
model the effects of wounds beyond the simplest level, because
(according to traditional gameplay) you either get healed or dead pretty
quickly. There's no need to seperate current wounds (cuts and
contusions) from old, healed wounds (deviated septum), because a game
just isn't long enough for those wounds to heal. Wounds can simply make
your character less effective until some death threshold is reached.

Any of the above can be invalid, of course, based on your game
concept. Maybe your games go on for years and your character actually
does have time to heal. But I think that in most cases, making wounds
nothing more than a performance hit with a death condition when they are
severe is a good strategy for bringing realism back into the realm of
desirable complexity. There can also be a logical progression of wounds:
the player is likely dodge until their legs are damaged, likely to block
until their arms are damaged, and not likely to sustain critical blows
to the head until both the arms and legs are hurt. That would be nearly
as linear as hit points.

> Outputting the effects to the user. Should our combat formula be open
> or closed to the player? or, Do we get to see the die rolls? In the
> open method the player sees all the math that goes on to determine how
> his actions affect the world around him. This support "Power Gaming"
> and kind of reduces the game to a spreadsheet. The alternative, with
> closed formulas where the player is fed what output you want him to
> see. Although, with time* he will deduce the basics of the
> calculations and "get a feel" for the game. Jeff and Paul pointed out
> that any complicated calculations are meaningless since they're hidden
> to the player. This allows for a simple combat system, which is nice
> from a development perspective.
>
> I think the open format would be better, although most games are
> closed. I'm not sure, I just like to know what is going on. Then
> again, I would probably enjoy a game implemented in excel.
>
> Dying means permadeath, and it's a realistic aspect of roguelikes that
> I fondly embrace. I think it makes it fun, but I know plenty of people
> who would argue that point. But how many of them peruse RGRD?

Probably not many of them do. I, of course, agree with you. I think the
loss of "Game Over" has taken a lot of the life out of games. If the
player has to tread through the same initial levels after dying, you'd
better make sure your game is fun to play. If they can just reload: not
so much.

> "Going Epic". In any situation where a human goes up against the
> stereotypical dragon and survives, nay, can EXPECT to survive, realism
> has been thrown right out the window. But becoming powerful is fun,
> and it belongs in the game. I was delighted the first time I got to
> the endgame in nethack and realized that my own character was the
> embodiment of war. So in this, I have to agree that realism can take
> it in the pants.

It's the difference between "physics engine" realism and simply being
mundane. Maybe in a space game you accelerate at a realistic rate, but
your engines have unreal fuel efficiency. It's physically accurate, but
a space ship like that would not be possible to build. The RL equivalent
of such a space ship would be the hero who can beat a dragon in a biting
match. Using hit points for combat is like playing Asteroids: no attempt
at all has been made at physical accuracy, to the extent that it's not
even possible to say where it goes wrong. It's just not even in this
world. Using a more realistic system is like the space ship at the start
of the paragraph: a world like ours is modeled, but selected real-life
constraints are ignored. In a RL, it's: this is how it would go if a
person really could get that strong or that tough. Wanting realism in
some regards doesn't mean you give up the right to say, "What if?"

> Limiting abilities is an important factor of realism that shouldn't be
> broken. For such games that HAVE ability scores (STR, DEX, INT, ETC)
> most of them have some sort of maximum that you can reach and go no
> further. This is realism in play. There are limits to how much a man
> can lift and how fast he can run.

I see those limits as more of an attempt to reduce the limits of the
game to ones the designer can actually fill. Without bounds, a character
is going to eventually be strong enough to simply punch her way through
the floor to the next dlevel, or at least be totally immune to the
creatures in the dungeon, and the game won't have much point anymore. I
guess you could view limits as necessary for realism, but the strongest
reason I can think of to cap players below godhood is to make the game
designer's job more reasonable, and to keep the play-style of the game
consistent from low to high levels without the need for an infinite
number of monster types.

> Magic. Now, by definition, I believe that magic is unrealistic. It's
> rather pointless to get into any specific ruleset, so I'm just going
> to have a blanket argument that selectivly breaking the rules of
> reality is what makes magic systems fun.

"What if?" Right?

zai...@zaimoni.com

unread,
Jun 17, 2008, 12:40:11 AM6/17/08
to
On Jun 16, 5:54 pm, heckrule...@yahoo.com wrote:

> Magic. Now, by definition, I believe that magic is unrealistic.

That, I'm not so sure about (especially the science-like magic typical
in RPGs and roguelikes). Especially after considering how our current
technology would look to someone from the 12th century. [Monk, you
shall not keep the works of Archimedes out of *our* grasp. The
Uneraser has lesser uses, such as recovering second-rate Epicurean
texts from Herculaneum.]

To the extent that a science-like magic system is as internally
consistent as a projection of future science (say, Larry Niven's Known
Space series), it can be as realistic as said projection of future
science.

Krice

unread,
Jun 17, 2008, 1:18:17 AM6/17/08
to
heckrule...@yahoo.com kirjoitti:

> Each and every sentence in this post could be considered flamebait.

It was just a long post without content. I don't know why people
assume that "realistic" (it never is) combat is always less
playable. It could be more playable and fun if made that way.

Inuga...@gmail.com

unread,
Jun 17, 2008, 1:28:58 AM6/17/08
to
I don't think the problem lies in Realism vs Playability, but in
Realism vs Games.

Realism and playability are things that need to be balanced. Sure, you
can model a game after statistical data, but that wouldn't be fun
having an @ dying every time it gets run over by a speeding SUV, in
fact it could suck.

The problem, as stated, is Realism vs Games, or more specific, Realism
vs Abstract Environments. We're taking some realistic or quasi-
realistic (dragons, magic) facts and sticking them into an abstract
game environment and transforming them into numbers. Somewhere along
the way, these things are distorted, depending on how abstract the
environment is. The less abstract (and more realistic) an evironment
is, the more complicated its programming and vice versa. Roguelikes
are EXTREMELY abstract. Roguelike movement is one step for one tile.
FPS games demonstrate less abstract in movement, and would result in
really large decimal places.

The other problem is how do we interpret the abstract number to be
realistic? Different people argue different points on what is
"realistic". It's like comparing abstract measurement systems: 2.54
centimeters or 1 inch? You wouldn't really know which one is correct
unless you tested them in reality, yet they are the same length. Now
we have a complicated combat models. Let's compare Crawl, Angband,
Gearhead, and Dwarf Fortress. I can argue each one has a realistic
combat system. The problem is, they are based in abstract numbers or
involve an RNG, defeating the purpose of realism.

Sure, you can hide the numbers or impose stat restrictions on the
player, but that's just giving the player a false sense of realism. In
reality, people can actually exceed that 50 STR cap of yours, and
there will be people that break existing barriers. Also we have to
consider the smart, strong, AND fast people. They do exist. They also
exist at young ages and old ages alike.

As to magic and fantasy elements (mechs, monsters), I call them quasi-
realistic. They aren't realistic, but based on facts taken from the
real world, we could fashion up some sort of explaination as to how
they work. Monsters would be equivalent to whatever creatures in
reality that could be frankensteined into its likeness. Magic would do
what it says on the box if it is physical and the effects of the
physical part could be equated to the counterparts in reality
(lightning spells could be compared to lightning itself and/or
defibrilators). However, how utility spells, curses, et cetera and the
existance of magic itself would have to be handed over to deus ex
machina.

Now consider the following (quasi-)realistic situation: A warrior
might be smart and fast enough run under the dragon to get to the
weakest part in the scales and have enough strength to stab through
its flesh into its heart, yet is killed by the falling dragon, not
because the dragon fell on top of him (crushing alot of bones), but
because he was too weak and the dragon too heavy, he partly drowns in
his own blood due to a broken rib puncturing his lungs, and thus dies
of asphixiation due to drowning in his own blood and lack of air
caused by the dragon's flesh covering his face in four minutes twenty-
seven seconds. Improbable? Yes. Impossible? No, because it happened.
Yeah, the situation is realistic (or quasi-realistic, because I think
the dragon weighs as much as an elephant, and has a hide as tough as
the sheet metal on your car door) because there is nothing that can
not be explained, yet you cannot describe that situation in numbers
that easily, but you can write it up in words. If you were to program
each of the individual elements, it'd take you a LONG time, or would
be impossible since you'd have to write up rules as to the physical
reality of the situation, which would be better time spent on writing
up amusing/interesting ways to kill the player.

Ido Yehieli

unread,
Jun 17, 2008, 6:17:32 AM6/17/08
to
On Jun 17, 12:54 am, heckrule...@yahoo.com wrote:
> Magic. Now, by definition, I believe that magic is unrealistic.

I believe the word "realistic" is inappropriate in this context. The
game world should be self-consistent, not necessarily similar to
ours.

E.g. I have no problem with warp drives in star trek, as long as they
are used in a consistent matter.

-Ido.

David Damerell

unread,
Jun 17, 2008, 6:37:40 AM6/17/08
to
Quoting <heckr...@yahoo.com>:
>against each other. That is, you could theoretically have a very
>realistic game that is very playable,

Let's see it in roguelikes before we accept that assertion. The only place
I think realism for its own sake works is simulation (flight, driving,
train driving, whatever) and devotees of those games tend to want graphics
- rightly, for once, I feel, because if I'm devoting myself to the real
problems of driving a 4-6-2 Pacific I _do_ want it to look like I'm
driving a 4-6-2 Pacific.

>combat reduces playability. Having un-realistic combat reduces the
>suspension of disbelief, and otherwise makes the game less fun.

_Completely_ unjustified assertion. There are a great many fun roguelikes
out there; they all have unrealistic combat. Realistic combat would mean
the inevitable small chance of being flat-out killed - how would that be
fun?

>The "solution", in vague and broad terms, is to "Strive for
>Balance!" (fence sitting FTW).

The solution is to note that there isn't a problem.

>representation of combat. Dice rolling. It adds randomness to a
>fight, which is both fun and realistic.

Jeff Lait has done some fun games with deterministic combat; and how
exactly is the normal roguelike situation, where when the average monster
rushes up to unwounded-you and takes a swing and has _no chance at all_ of
incapacitating you, realistic?

>Hit points. We've talked about this is a recent post. Hit points are a
>game mechanic that allow people to be injured without dying. This is,
>if I may go out on a limb, an unrealistic side to games.

Stating the blindingly obvious is going out on a limb?

>probably the best stab we have at being realistic. But I'd have stress
>that every wound impairs you in some way. The balance? Well, how about
>a game that reduces your stats by a factor of your hit points over
>your max hit points?

So once you start to lose you keep on losing. Just _great_ for the "single
hero against the hordes" so typical in roguelikes.

>Area-specific damage is a leap forward in both
>realism AND playability

Unjustified assertion re playability.

>Dying means permadeath, and it's a realistic aspect of roguelikes that
>I fondly embrace.

Is it realistic? Free hint; after really dying you won't get to start a
new character and try again.

>Limiting abilities is an important factor of realism that shouldn't be
>broken. For such games that HAVE ability scores (STR, DEX, INT, ETC)
>most of them have some sort of maximum that you can reach and go no
>further. This is realism in play.

Ordinarily the limits are exactly the same for every member of a given
species. Normally the physical limits are independent of gender.
Sometimes (NetHack) you can exercise up from wimp to superman as a
side project in the time of a dungeon delving. Very often the strength
limit gives literally superhuman carrying capacity. All ever so realistic?
--
David Damerell <dame...@chiark.greenend.org.uk> Distortion Field!
Today is Second Stilday, June - a weekend.

Jakub Debski

unread,
Jun 17, 2008, 7:00:30 AM6/17/08
to
heckr...@yahoo.com formulated the question :

> I would say that the two aspects of a roguelike, Combat Realism and
> Playability, are not directly dependent upon each other but often work
> against each other.

You mean *realism* in turn-based game where two letters stand side by
side? :)

regards,
Jakub


I Own The Letter O

unread,
Jun 17, 2008, 10:20:38 AM6/17/08
to
On 17 Jun, 12:00, Jakub Debski <debski.ja...@wp.pl> wrote:
> heckrule...@yahoo.com formulated the question :

If I've spent several hours (and several characters) getting my
Gelatinuous Cube Space Ranger (realism plus!!!) to level 30 with a
glut of good kit to boot but then lose him to a level 1 kobold with a
spear who stabbed him from behind and managed to pierce his nucleus
resulting in an instant kill I'd through my laptop at this beast
called realism. I've got an AC of 100, damage resistance/absorption
through the roof and yet a little kobo manages to stab me, me!

Realism - A sharp pointy thing hitting you from behind, even if you
are Arnie in his prime with Stephen Hawkings Int, Dali Lama's Wis and
Brad Pitt's Cha, has a good chance of piercing something vital.

Roguelike - Ha ha ha, my incredibly high AC/Dodge/Whatever means the
puny kobo misses, even though he snuck up on me, my 360 vision allowed
me to dodge. Now I shall go and wash in the river of lava with my fire
resistance before swallowing a gallon of poison to sate my hunger with
my poison resistance. Then I will fill my ears full of wax before
luring a banshee to an underground cave full of living trees to get
the treasure that they somehow carry with them!!!

Jeff Lait

unread,
Jun 17, 2008, 11:01:41 AM6/17/08
to

Any sufficiently advanced magic is indistinguishable from technology.

I have no interest in "science" magic for this very reason. The more
you make magic a orthogonal, internally consistent, system, the less
it is magic. Magic is supposed to have mystery and surprise.

My rule of thumb is that every magical spell must require new code or
it is not sufficiently different from existing spells to deserve to
exist.
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)

heckr...@yahoo.com

unread,
Jun 17, 2008, 11:45:11 AM6/17/08
to
Waiiiit a minute, I thought it was:
Any sufficiently advanced technology is indistinguishable from magic.

The main distinction between magic and future tech is their
believability. People have an easier time accepting that which that
which they can foresee. So a space elevator inflicts less disbelief
then a witch turning people into toads.

That's for laymen though, I fully support such fantasy clichés. I've
seen enough witches transforming people that I believe it to be
internally consistent with the genre. Sure it's mysterious or
whatever, but it's old hat. My father on the other hand simply won't
accept anything like that. We rented a Japanese fantasy-action flick
and at some point he asked what the heck was going on. I tried to
explain that the guy was summoning a demon (well, an Oni, but one step
at a time).
"it's magic"
"Like Harry Potter"
"Um... sort of?"
And then he just sat back with a look I won't forget. He was just
dismissing the whole movie. He no longer cared. He could no longer
suspend his disbelief.

-HeckRuler

I Own The Letter O

unread,
Jun 17, 2008, 11:51:03 AM6/17/08
to
On 17 Jun, 16:45, heckrule...@yahoo.com wrote:

> Waiiiit a minute, I thought it was:
> Any sufficiently advanced technology is indistinguishable from magic.

Depends on who you are quoting, many authors have stated the inverse
of the famous quote.

heckr...@yahoo.com

unread,
Jun 17, 2008, 12:00:51 PM6/17/08
to

>Is it realistic? Free hint; after really dying you won't get to start a
>new character and try again.

So you just assume I'm not Hindu?

To each his own my friend.
-HeckRuler

David Damerell

unread,
Jun 17, 2008, 12:23:22 PM6/17/08
to
Quoting <heckr...@yahoo.com>:
>>Is it realistic? Free hint; after really dying you won't get to start a
>>new character and try again.
>So you just assume I'm not Hindu?

There are no fairies at the bottom of the garden. You can believe all you
like that there are, but there aren't.

But in any case Hindus do not suppose that people come back with a memory
of what killed them last time and a determination not to get nobbled by it
again.

David Damerell

unread,
Jun 17, 2008, 12:27:04 PM6/17/08
to
Quoting I Own The Letter O <lord...@yahoo.co.uk>:
>my poison resistance. Then I will fill my ears full of wax before
>luring a banshee to an underground cave full of living trees to get
>the treasure that they somehow carry with them!!!

That's the spirit. Sod realism.

zai...@zaimoni.com

unread,
Jun 17, 2008, 12:44:00 PM6/17/08
to
On Jun 17, 5:37 am, David Damerell <damer...@chiark.greenend.org.uk>
wrote:

> Quoting <heckrule...@yahoo.com>:
>
> >against each other. That is, you could theoretically have a very
> >realistic game that is very playable,
>
> Let's see it in roguelikes before we accept that assertion. The only place
> I think realism for its own sake works is simulation (flight, driving,
> train driving, whatever) and devotees of those games tend to want graphics
> - rightly, for once, I feel, because if I'm devoting myself to the real
> problems of driving a 4-6-2 Pacific I _do_ want it to look like I'm
> driving a 4-6-2 Pacific.
>
> >combat reduces playability. Having un-realistic combat reduces the
> >suspension of disbelief, and otherwise makes the game less fun.
>
> _Completely_ unjustified assertion. There are a great many fun roguelikes
> out there; they all have unrealistic combat. Realistic combat would mean
> the inevitable small chance of being flat-out killed - how would that be
> fun?
>
> >The "solution", in vague and broad terms, is to "Strive for
> >Balance!" (fence sitting FTW).
>
> The solution is to note that there isn't a problem.
>
> >representation of combat. Dice rolling. It adds randomness to a
> >fight, which is both fun and realistic.
>
> Jeff Lait has done some fun games with deterministic combat; and how
> exactly is the normal roguelike situation, where when the average monster
> rushes up to unwounded-you and takes a swing and has _no chance at all_ of
> incapacitating you, realistic?

If the game is being moderated at a "sufficiently fast" timescale,
combat often ends up being *very* deterministic as the main sources of
realistic error are mental/programmatic modeling errors, and materials
property variation. To-hit rolls basically stop being relevant.

> >probably the best stab we have at being realistic. But I'd have stress
> >that every wound impairs you in some way. The balance? Well, how about
> >a game that reduces your stats by a factor of your hit points over
> >your max hit points?
>
> So once you start to lose you keep on losing. Just _great_ for the "single
> hero against the hordes" so typical in roguelikes.

It worked fine for the 7DRL Urban Warfare. At larger scales, you
wouldn't see that trope because it had been designed out.

heckr...@yahoo.com

unread,
Jun 17, 2008, 3:05:48 PM6/17/08
to
On Jun 17, 11:27 am, David Damerell <damer...@chiark.greenend.org.uk>
wrote:

> Quoting  I Own The Letter O  <lord_h...@yahoo.co.uk>:
>
> >my poison resistance. Then I will fill my ears full of wax before
> >luring a banshee to an underground cave full of living trees to get
> >the treasure that they somehow carry with them!!!
>
> That's the spirit. Sod realism.
> --
> David Damerell <damer...@chiark.greenend.org.uk> Distortion Field!

> Today is Second Stilday, June - a weekend.

So would you be in favor of a surreal roguelike? Or do you know of
any?
And the topic is realistic combat, so you'd want some surreal form of
attacks. Some sort of combat chefs, or killer cuddling, or a literal
sharp wit. You could make "Twisp and Catsby's adventure in
Rogueaira".

I guess the most surreal RL I can think of would be Numbers.

Martin Read

unread,
Jun 17, 2008, 4:42:03 PM6/17/08
to
heckr...@yahoo.com wrote:
>And the topic is realistic combat, so you'd want some surreal form of
>attacks. Some sort of combat chefs, or killer cuddling, or a literal
>sharp wit. You could make "Twisp and Catsby's adventure in
>Rogueaira".

A melting clock engulfs your head. You cannot breathe. You have become
a small teapot shaped like a fish. Do you want your humdrum trappings of
reality identified?
--
\_\/_/ turbulence is certainty turbulence is friction between you and me
\ / every time we try to impose order we create chaos
\/ -- Killing Joke, "Mathematics of Chaos"

jot...@hotmail.com

unread,
Jun 17, 2008, 5:17:28 PM6/17/08
to
IMO detailed damage systems are just icing on the cake. And you still
have no cake, because "rolling" is too shallow to become a tasty cake.
It's just icing. What you really need is a detailed combat model.

Jotaf

Paul Donnelly

unread,
Jun 17, 2008, 5:33:34 PM6/17/08
to
David Damerell <dame...@chiark.greenend.org.uk> writes:

> Quoting <heckr...@yahoo.com>:
>>against each other. That is, you could theoretically have a very
>>realistic game that is very playable,
>
> Let's see it in roguelikes before we accept that assertion. The only place
> I think realism for its own sake works is simulation (flight, driving,
> train driving, whatever) and devotees of those games tend to want graphics
> - rightly, for once, I feel, because if I'm devoting myself to the real
> problems of driving a 4-6-2 Pacific I _do_ want it to look like I'm
> driving a 4-6-2 Pacific.
>
>>combat reduces playability. Having un-realistic combat reduces the
>>suspension of disbelief, and otherwise makes the game less fun.
>
> _Completely_ unjustified assertion. There are a great many fun roguelikes
> out there; they all have unrealistic combat. Realistic combat would mean
> the inevitable small chance of being flat-out killed - how would that be
> fun?

Completely unjustified _assumption_. There's a small chance that I'll
quantum-tunnel to the sidewalk next time I leave the house rather than
using the door. I'm much too large for it to be likely, but it could
happen. Is the probability high enough to implement it in a RL? I don't
think so. The same goes for a one-hit kill from the average monster if
its level is near that of your character. It would make some sense when
monsters are much higher in level than your character, but in that case
you're screwed anyway. Which you are in the HP system too. It's not
unrealistic to ignore the possibility of uncommon events when you don't
think they will add to the game.

Billy Bissette

unread,
Jun 17, 2008, 9:20:22 PM6/17/08
to
jot...@hotmail.com wrote in news:063042ec-c669-4da6-b430-e51dede45675@
27g2000hsf.googlegroups.com:

> IMO detailed damage systems are just icing on the cake. And you still
> have no cake, because "rolling" is too shallow to become a tasty cake.
> It's just icing. What you really need is a detailed combat model.

You need a detailed combat model that:
1) Matters
2) Doesn't kill the game

You could come up with something that made each individual battle a
complex and logical but still thrilling dance of death. A single foe
might take hundreds of considered commands (not just mashing in the
same action turn after turn) and the battle might last ten or twenty
minutes, but it still can be interesting. Until you put it into a
game where on average you will face 10,000 battles before you can
reach victory.


Omega let you pre-program an attack string. You had a certain
number of attacks in a turn, and could program them for high or low
or defense or riposte. Theoretically, it would make combat more
interesting than just bumping a creature. In practice, it just
didn't work. While beginners didn't know how to take advantage of
the system, it quickly becomes apparent that it didn't really matter.
As long as you have some variety, you'll eventually blast through
everything anyway (particularly once you start using things other
than basic combat). On the other hand, even if it did matter and
attack string design was a critical factor, then you'd have had
the problem of the player needing to redesign their combat string
every time they face a different enemy.

You can make a game where combat is a massive simulation of
behind-the-scenes calculations. I'd once thought of a system
where players could learn various forms of combat, each with
their own attack patterns, bonuses, weaknesses, special features,
and the like. Further, the game would automatically cobble
together a "best combination" of styles based on the situation at
hand (weapons used, equipment worn, and enemy details.) The
question is, for the amount of work done, would the player even
notice? Furthermore, would the player still hit spots where
they griped? (And would some of those situations be warranted?)

Ido Yehieli

unread,
Jun 18, 2008, 5:55:42 AM6/18/08
to
On Jun 17, 9:05 pm, heckrule...@yahoo.com wrote:
> So would you be in favor of a surreal roguelike? Or do you know of
> any?

Nethack?

Ido Yehieli

unread,
Jun 18, 2008, 5:58:58 AM6/18/08
to
On Jun 17, 9:05 pm, heckrule...@yahoo.com wrote:
> I guess the most surreal RL I can think of would be Numbers.

Are you confusing *abstract* with *surreal*?

Talking about surreal, an expanded version of Alan's Psychedelic
Journey (7drl) might fit the bill.

-Ido.

Gerry Quinn

unread,
Jun 18, 2008, 11:32:23 AM6/18/08
to
In article <a7fb4730-d662-4ec0-a3be-
562000...@e53g2000hsa.googlegroups.com>, heckr...@yahoo.com
says...

> But I'd have stress
> that every wound impairs you in some way. The balance? Well, how about
> a game that reduces your stats by a factor of your hit points over
> your max hit points?

Problem: the first hit tends to win. This may be realistic, but does
not tend to make for exciting battles where the outcome is in doubt. At
best, it can give a system where you try to get the first hit in, and if
you fail, it becomes "can I get out of this alive"? That's not too bad,
especially for certain character types, but somewhat restrictive to have
in your design before you even start building the game...

> Outputting the effects to the user. Should our combat formula be open
> or closed to the player?

I lean towards full information in roguelikes. After all, what would a
realistic game be without it?

-> The Giant swings and chops off one of your legs.
-> Kick Giant
-> Kick Giant using which leg - left(L) or right(R)?
-> ummm..... R
-> I am sorry, you do not have a right leg
-> The Giant swings and chops off your head.
-> You die.
-> You had 1056 gold pieces

> Dying means permadeath, and it's a realistic aspect of roguelikes that
> I fondly embrace. I think it makes it fun, but I know plenty of people
> who would argue that point. But how many of them peruse RGRD?

I think we can have roguelikes without permadeath.

> Limiting abilities is an important factor of realism that shouldn't be
> broken. For such games that HAVE ability scores (STR, DEX, INT, ETC)
> most of them have some sort of maximum that you can reach and go no
> further. This is realism in play. There are limits to how much a man
> can lift and how fast he can run.

I hate games where you can max them all out, like in Angband.

> Magic. Now, by definition, I believe that magic is unrealistic. It's
> rather pointless to get into any specific ruleset, so I'm just going
> to have a blanket argument that selectivly breaking the rules of
> reality is what makes magic systems fun.

It can be fun just as another form of combat (with a different flavour
from swords) too.

- Gerry Quinn
--
Lair of the Demon Ape (a coffee-break roguelike)
<http://indigo.ie/~gerryq/lair/lair.htm>

zai...@zaimoni.com

unread,
Jun 18, 2008, 12:12:43 PM6/18/08
to
On Jun 18, 10:32 am, Gerry Quinn <ger...@indigo.ie> wrote:
> In article <a7fb4730-d662-4ec0-a3be-
> 562000b33...@e53g2000hsa.googlegroups.com>, heckrule...@yahoo.com
> says...

> > Outputting the effects to the user. Should our combat formula be open
> > or closed to the player?>
> I lean towards full information in roguelikes. After all, what would a
> realistic game be without it?

It depends on what "full information" means. My opinion is wargame-
ish: the overall combat formula should be fully exposed, as it *is*
fully exposed to @ by immersion. The player is not sensorily
immersed, so the implementor should compensate.

(SSCrawl pushes the limit here. It *barely* qualifies, mainly because
going from the verbal descriptions to the algorithm provides no useful
information.)

On the other hand, an unfun application of realism is to not render
anything @ cannot see/hear/etc. This is *very* bad for playability if
you lock down the player to only consciously aware information (5-9
chunks for a normal human, adjust to the time scale of the game),
rather than allowing anything in iconic memory to be known because the
player is playing both conscious and subconscious minds of the @.
[Note that 360-degree vision at large time scales (2+ seconds) is not
that unrealistic; it abstracts both "lighthouse syndrome" and hearing.]

David Damerell

unread,
Jun 18, 2008, 12:37:54 PM6/18/08
to
Quoting Paul Donnelly <paul-d...@sbcglobal.net>:

>David Damerell <dame...@chiark.greenend.org.uk> writes:
>>_Completely_ unjustified assertion. There are a great many fun roguelikes
>>out there; they all have unrealistic combat. Realistic combat would mean
>>the inevitable small chance of being flat-out killed - how would that be
>>fun?
>Completely unjustified _assumption_. There's a small chance that I'll
>quantum-tunnel to the sidewalk next time I leave the house rather than
>using the door. I'm much too large for it to be likely, but it could
>happen. Is the probability high enough to implement it in a RL? I don't
>think so. The same goes for a one-hit kill from the average monster if
>its level is near that of your character.

Except that - unlike quantum tunneling - a single hit being fatal or so
incapacitating as to render a human incapable of fighting is a perfectly
plausible result when everyone's waving bloody great weapons around. People
are incapacitated by single stab wounds from tiny shivs; that being bashed
over the head with a mace could kill you is not a highly improbable effect
unworthy of implementation at all.
--
David Damerell <dame...@chiark.greenend.org.uk> Kill the tomato!
Today is Potmos, Presuary.

Joni Toivanen

unread,
Jun 18, 2008, 1:28:51 PM6/18/08
to
Ido Yehieli wrote:
[snip]

> Talking about surreal, an expanded version of Alan's Psychedelic
> Journey (7drl) might fit the bill.
>
> -Ido.

I'm actually working on a sequel to the game, but I'm not exactly sure
how surreal it's going to be. That being said, I am trying to keep the
feeling at least somewhat similar to the first game :)

The name is going to be "Fungi Island" and there will be a short story
in the game. But no more details, as the game is at a very early stage
of development...

Joni.

David Damerell

unread,
Jun 18, 2008, 1:49:12 PM6/18/08
to
Quoting <heckr...@yahoo.com>:
>On Jun 17, 11:27=A0am, David Damerell <damer...@chiark.greenend.org.uk>

>>Quoting =A0I Own The Letter O =A0<lord_h...@yahoo.co.uk>:
>>>my poison resistance. Then I will fill my ears full of wax before
>>>luring a banshee to an underground cave full of living trees to get
>>>the treasure that they somehow carry with them!!!
>>That's the spirit. Sod realism.
>So would you be in favor of a surreal roguelike?

Surreal versus realistic is an obvious false dichotomy. Try harder.

jot...@hotmail.com

unread,
Jun 18, 2008, 2:22:08 PM6/18/08
to
On 18 Jun, 02:20, Billy Bissette <bai...@coastalnet.com> wrote:
> jota...@hotmail.com wrote in news:063042ec-c669-4da6-b430-e51dede45675@

> 27g2000hsf.googlegroups.com:
>
> > IMO detailed damage systems are just icing on the cake. And you still
> > have no cake, because "rolling" is too shallow to become a tasty cake.
> > It's just icing. What you really need is a detailed combat model.
>
> You need a detailed combat model that:
> 1) Matters
> 2) Doesn't kill the game
>
> You could come up with something that made each individual battle a
> complex and logical but still thrilling dance of death. A single foe
> might take hundreds of considered commands (not just mashing in the
> same action turn after turn) and the battle might last ten or twenty
> minutes, but it still can be interesting. Until you put it into a
> game where on average you will face 10,000 battles before you can
> reach victory.

There's an obvious fix to this. More interesting combat means less
foes to fight around. I'd say that fighting 4 enemies at the same time
without precise control and more emphasis on dodging than on hitting
would be impossible. You'd be better off cobbling ways to take them
out one at a time instead. A full-frontal assault would be more
demanding of the player, but then so is a big boss battle in Final
Fantasy. I think this is only good for the game, care to disagree?

>
> Omega let you pre-program an attack string. You had a certain
> number of attacks in a turn, and could program them for high or low
> or defense or riposte.

> ...

But you're assuming that combat is just a string of optimal commands.
My view is more of highly reactive combat, where you need to
constantly adjust to the situation, not mindless repetition. It's the
opposite of a grinding game, where you hit the same key over and over
again, or the same combination of keys (smells like a MMORPG...). IMO
it's a step forward.

>
> You can make a game where combat is a massive simulation of
> behind-the-scenes calculations.

> ...

The player must be presented with that information, in order to adjust
the combat technique. There's no point in hiding anything. If the
enemy is steadily improving its position against the player, the
player should know. Otherwise it's no better than the "oops, the dice
says I die". Jice is developing a combat prototype where he shows the
status bar with the "attack cooldown" of the enemies; while playing, I
found myself instinctively backing off from an enemy when his cooldown
was almost over. It's a work in progress, but I already like where
this is going.

Jotaf

heckr...@yahoo.com

unread,
Jun 18, 2008, 3:12:37 PM6/18/08
to

Probably. Even after the going through the wiki entry on surreal I
still don't "get it". I think it's best described as "dreamlike".
Abstract still isn't the right word however, as we use real-world
equivlents in roguelikes: Swords, monsters, dungeons, stairs, etc. But
we are talking about combat system specifically here so abstract might
work just fine. However I think any sort of calculation to represtent
combat is an abstraction of violence. Simple calculations may be MORE
abstract however. Hmm...

David, I think you're missing the point: That a balance between
realism and playability is important. There is no ultimate realism.
It's just something to work towards. Often to the extent that you kill
the game. It's a lot like ultimate replayability. Everything has a
certain degree of realism. You can't just throw one or the other out
the window.

I honestly don't understand what you would be trying to make if you
wanted to completely abandon reality.

-HeckRuler

Jeff Lait

unread,
Jun 18, 2008, 4:23:18 PM6/18/08
to
On Jun 18, 3:12 pm, heckrule...@yahoo.com wrote:
>
> I honestly don't understand what you would be trying to make if you
> wanted to completely abandon reality.

Numbers is a really cool roguelike. Its combat system completely
abandons reality. Indeed, it is one of the most challenging combat
systems I've played in a roguelike.

heckr...@yahoo.com

unread,
Jun 18, 2008, 5:19:59 PM6/18/08
to
On Jun 18, 11:12 am, zaim...@zaimoni.com wrote:

> [Note that 360-degree vision at large time scales (2+ seconds) is not
> that unrealistic; it abstracts both "lighthouse syndrome" and hearing.]

Well, allow me to agree that 360-degree FOV, which is the standard, is
definatly realistic enough. I don't discount the whole genre. But I'd
have to say that a system with hearing, directional FOV, and a small
timeslice (moving to turning 45deg ratio of 1:4) would be MORE
realistic. It's more realistic because you explicitly impliment two
features of anatomy, directional vision and hearing. Some, maybe most,
would say that this reduces playability, but I don't mind the
lighthouse syndrome. You need a smaller timeslice so you can make it
work, and that increases the play-time it takes to do some basic
things, which is a general hit to playability. But I think if you made
a small time-slice system that played as fast as a regular RL, we'd
have a happy medium of real and playable.

It's more complicated to design and build around, that's for sure. It
gives you more tools to work with in terms of game balance, but AI
complexity increases.

I'd love to point out mazecrawl as a fun game that impliments these
things, but the site is down. How long do I have to wait before
declaring it abandonware?
-HeckRuler

Billy Bissette

unread,
Jun 18, 2008, 6:16:11 PM6/18/08
to
jot...@hotmail.com wrote in news:d70d44ff-481f-466d-978b-bdd104b0fa95
@e53g2000hsa.googlegroups.com:

I'm not being negative to the idea.

The first paragraph above is showing how you can easily botch #2,
with an example of having achieved a complicated but interesting
battle system, and then blowing it by plugging it into a standard
Roguelike encounter rate designed for standard Roguelike bashing.

The second paragraph is an example of a real game that botched #1.
Omega had the basis for more user input, but it ultimately didn't
matter. Mind, the same paragraph warns that if Omega's combat
programming had mattered, then as implemented it would still risk
botching #2 (by being too much a hassle.)

>>
>> You can make a game where combat is a massive simulation of
>> behind-the-scenes calculations.
>> ...
>
> The player must be presented with that information, in order to adjust
> the combat technique. There's no point in hiding anything.

The final example that you cut arguably could work with hiding
at least some of the details. But then it called for having the
game engine itself pick the optimal attacks. Players would have
input in what attack forms they would chose to pursue, and what
weapons they chose to equip. They wouldn't necessarily need to
know all the details and numbers, just the general ideas of what
to expect.

(If you wonder why that idea called for automation, it was a
way to keep from bogging down combat on the player's side. It
was a system where the player still could just repeatedly bump
into an enemy, but still be more than just random checks of
to-hit versus armor. And for different enemy details to matter,
without having the player manually change tactics on each round
of combat based on what he was facing. It also ran with the
idea that forms were sets of moves that a fighter trained to
perform in sequence. Some forms would even be designed to
branch based on conditions. (Such as if you fought with spear
and net, then successfully netting an opponent would
automatically trigger a different follow-up attack than failing
to do so would trigger.) Mind, this idea had an alternate
version where the player could program his own move trees, with
various conditional branches, built from various moves and
condition checks that he had learned.)

Derek Ray

unread,
Jun 18, 2008, 7:13:07 PM6/18/08
to
On 2008-06-18, heckr...@yahoo.com <heckr...@yahoo.com> wrote:
> Well, allow me to agree that 360-degree FOV, which is the standard, is
> definatly realistic enough. I don't discount the whole genre. But I'd
> have to say that a system with hearing, directional FOV, and a small
> timeslice (moving to turning 45deg ratio of 1:4) would be MORE
> realistic.

And significantly less fun. I'm playing a roguelike because it's _not_
an FPS, after all; I don't want to worry about constantly turning around
to see what's behind me.

It's possible to implement hearing without removing 360-degree FOV.

> features of anatomy, directional vision and hearing. Some, maybe most,
> would say that this reduces playability,

If you're looking for things that might be "good" features, this is an
excellent sign that you have made a "bad" feature.

> but I don't mind the lighthouse syndrome.

It's not _you_ who's going to be playing your game most of the time.

> things, which is a general hit to playability. But I think if you made
> a small time-slice system that played as fast as a regular RL, we'd
> have a happy medium of real and playable.

This is tantamount to saying "If I had wings that worked, I could fly."
It doesn't mean anything until someone actually makes some penguin
wings that work. In this case, the wings are "a non-360 FOV system
that plays as fast as a regular RL and is just as fun for the player, as
opposed to getting in his way and forcing him to spin round like a top
constantly lest he get snuck-up-behind."

Trying to force realism into games often removes the fun. This is one
such case.

--
Derek

Game info and change log: http://sporkhack.com
Beta Server: telnet://sporkhack.com
IRC: irc.freenode.net, #sporkhack

Numeron

unread,
Jun 18, 2008, 8:05:39 PM6/18/08
to
I tend to agree that the lighthouse syndrome removes alot of the fun
from the game: particularly because it makes walking through empty
areas so much more tedious. However Ive always considered that
directional FOV has been given way too much of a bashing here. It
certainly has potential, its jsut all about the implementation.

consider a system that has 360 degree view with a radius of 3 tiles,
and a 90 degree 6 tile directional cone on top. Its keeps the realism
of directional FOV, but its also somewhat more functional. Or perhaps
instead of circular FOV centered on the player, it could be centered a
few tiles infront of the player (or be an oval shape) so that they can
see more of whats infront and less behind but can still see behind.

The problem is rarely realism itself but poor implementation.

-Numeron

Paul Donnelly

unread,
Jun 18, 2008, 8:24:17 PM6/18/08
to
Derek Ray <de...@moot.its.only.a.spamtrap.org> writes:

> On 2008-06-18, heckr...@yahoo.com <heckr...@yahoo.com> wrote:
>> Well, allow me to agree that 360-degree FOV, which is the standard, is
>> definatly realistic enough. I don't discount the whole genre. But I'd
>> have to say that a system with hearing, directional FOV, and a small
>> timeslice (moving to turning 45deg ratio of 1:4) would be MORE
>> realistic.

How so? I think it would be better in this case to say "more precise",
since having 360-degree awareness within the space of a turn is not at
all unreal.

> And significantly less fun. I'm playing a roguelike because it's _not_
> an FPS, after all; I don't want to worry about constantly turning around
> to see what's behind me.
>
> It's possible to implement hearing without removing 360-degree FOV.
>
>> features of anatomy, directional vision and hearing. Some, maybe most,
>> would say that this reduces playability,
>
> If you're looking for things that might be "good" features, this is an
> excellent sign that you have made a "bad" feature.
>
>> but I don't mind the lighthouse syndrome.
>
> It's not _you_ who's going to be playing your game most of the time.

Maybe it will be, if the interface is bad. :P

>> things, which is a general hit to playability. But I think if you made
>> a small time-slice system that played as fast as a regular RL, we'd
>> have a happy medium of real and playable.

I bet a quick-turn key (like in the old FPSes) and wide FOV would work
wonders. Make it a question of whether to check behind you for pursuers
rather than save your turn for movement or whatever. A perfect fit for
your Spy vs. Spy RL with sneaky enemies.

Paul Donnelly

unread,
Jun 18, 2008, 8:38:11 PM6/18/08
to
David Damerell <dame...@chiark.greenend.org.uk> writes:

What if my character encounters only low-level beasts before acquiring
enough combat skill and/or armor to effectively guard his head against
foes of his level? As the designer, can't I exert at least a little
influence over the situations players will find themselves in, perhaps
keeping the Ecstatic Knights of the Church of the Severed Head deep in
the dungeon where they belong? I don't know where the idea that realism
means "worst-case scenarios, all the time" came from, but it's not what
I have in mind.

zai...@zaimoni.com

unread,
Jun 18, 2008, 9:35:33 PM6/18/08
to
On Jun 18, 4:19 pm, heckrule...@yahoo.com wrote:
> On Jun 18, 11:12 am, zaim...@zaimoni.com wrote:
>
> > [Note that 360-degree vision at large time scales (2+ seconds) is not
> > that unrealistic; it abstracts both "lighthouse syndrome" and hearing.]
>
> Well, allow me to agree that 360-degree FOV, which is the standard, is
> definatly realistic enough. I don't discount the whole genre. But I'd
> have to say that a system with hearing, directional FOV, and a small
> timeslice (moving to turning 45deg ratio of 1:4)

Classic. If I was copying X-COM 1/2's TU system, that would be it.
[I'm not, as I want the benefits of the TU/Reactions hack in that
series without the weaknesses.]

> .... You need a smaller timeslice so you can make it


> work, and that increases the play-time it takes to do some basic
> things, which is a general hit to playability.

A *very* serious hit to playability. The only solution I have thought
of so far is proper pathfinding, on the assumption that you drop the
player into the ~0.1-second reaction-time mode only when absolutely
necessary.

There would need to be some modes to the pathfinding (fastest-path vs.
cover/concealment-seeking path, etc.).

zai...@zaimoni.com

unread,
Jun 18, 2008, 9:45:11 PM6/18/08
to
On Jun 18, 7:05 pm, Numeron <irunsofastineedafinonmyh...@hotmail.com>
wrote:

> I tend to agree that the lighthouse syndrome removes alot of the fun
> from the game: particularly because it makes walking through empty
> areas so much more tedious.

Walking *and* doubling back. Urban Warfare made this work by making
doubling back almost certainly game-losing, and ensuring that no enemy
is faster than you are. (A decent strategy is to immediately zoom out
identify the fastest direction out, then zoom in back to the "largest
scale that supports keypad diagonals". With careful attention to
moving with cover, lighthouse syndrome happens only when it's
intuitive: stepping out past a corner.)

Billy Bissette

unread,
Jun 19, 2008, 12:07:06 AM6/19/08
to
Paul Donnelly <paul-d...@sbcglobal.net> wrote in
news:87lk125...@plap.localdomain:

The problem is random chance, with multiple rounds to a battle,
multiple battles, multiple games, and multiple players means that
some people are going to hit those worst-case scenarios.

What percentage chance should it be? One in a hundred? One
hundred attacks can happen really quickly. That is only twenty
enemies getting five attacks each. (Sure, you might go two or
three hundred attacks without it happening. But it also might
happen on the first.)

One in a thousand? Face a few hundred enemies, and you can be
on the receiving end of a thousand attacks. It still may be common
enough to be a "certainty" in your game.

Sure, you can make it even more uncommon. One in ten thousand.
One in one hundred thousand. Now you are in the territory of blind
chance ending a game. And existing Roguelikes *will* dance this
particular dance, but it is a risky thing to play with, to get to
the point that players don't feel robbed if it happens.

And if you shove it off to some deep part of the dungeon, then it
just becomes another form of deadly special attack.

Inuga...@gmail.com

unread,
Jun 19, 2008, 3:32:27 AM6/19/08
to
On Jun 18, 9:07 pm, Billy Bissette <bai...@coastalnet.com> wrote:
> Sure, you can make it even more uncommon. One in ten thousand.
> One in one hundred thousand. Now you are in the territory of blind
> chance ending a game. And existing Roguelikes *will* dance this
> particular dance, but it is a risky thing to play with, to get to
> the point that players don't feel robbed if it happens.
>
> And if you shove it off to some deep part of the dungeon, then it
> just becomes another form of deadly special attack.

It's something the player can't reduce, reverse, or prevent that leads
directly to death, which makes it unfair towards the player. Fights
shouldn't be completely entrusted to luck, as it would be gambling,
and in gambling, the player is certain to eventually "win" death
(lose) when given enough time (in slots, it would be hitting 777)
unless there is a way to figure out how to beat the system (counting
in blackjack, on-the-fly physics calculations in roulette) or the
player stops gambling completely (fighting in our case).

Also, having an instant kill option for monsters (against the player)
would be disasterous, as the player would soon learn NOT to fight due
to negative reinforcement ("If I fight, I am certain I will die
eventually"). The player would be TERRIFIED of the smallest things if
the player isn't able to kill them before they attack the player. The
problem is, to kill things quickly in order to not get one-hit-killed,
the player would have to fight things that might kill him/her. Another
problem is, this encourages farming ("I have killed the little things
to the point that I can kill them before they can attack me, now I
will keep on killing the little things to the point that I can safely
kill the bigger things before they can attack me").

Giving an instant kill option for players (against monsters) is also
disasterous (slight less than giving it to monsters), because it gives
the player too much freedom and encourages reckless gameplay. It might
make players make stupid decisions hoping to get a lucky break or make
stupid players make stupid decisions and get away with it (there's a
difference). One player's going to do something stupid hoping to get
lucky, but isn't guaranteed, the other is playing like an idiot and
gets a lucky break. It's not good for learning. Also, it makes
everything in the dungeon seem to become more or less powerful, from
monsters, to weapons/items, and even player classes. Dangerous
monsters get less powerful because they can be killed instantly at any
moment when being attacked by the player. Powerful weapons/items
become less powerful the stronger they are and weaker weapons/items
grow in power (why wield that 2d7 long sword when you can dual wield
1d1 multi-hit nun-chucks that are faster and get in more hits to
hopefully kill monsters?). Wizards and Rangers would also become MUCH
stronger than Fighters because now their weak mele attacks as well as
their distanced attacks can instantly kill.

I say to COMPLETELY avoid this idea as it severely downplays the
tactical aspect of RLs (winning and losing are dependant on player's
actions) and turns it into pure gambling (winning and losing are
independant of the player's actions), disregarding player skill.

POWDER plays fair because any negative random event-- be it reading a
bad scroll, drinking a bad potion, getting punished by a pissed god,
et cetera-- is survivable as long as the player takes precautions in
order to survive (getting HP, IDing, getting gods on his/her side and
doing less to piss off others in early game).

Ido Yehieli

unread,
Jun 19, 2008, 5:22:33 AM6/19/08
to
On Jun 19, 2:05 am, Numeron <irunsofastineedafinonmyh...@hotmail.com>
wrote:

> consider a system that has 360 degree view with a radius of 3 tiles,
> and a 90 degree 6 tile directional cone on top. Its keeps the realism
> of directional FOV, but its also somewhat more functional.

I believe I implemented exactly that (modulus the number of tiles) in
a game I abandoned a few years ago.
Worked quite well iirc.

-Ido.

I Own The Letter O

unread,
Jun 19, 2008, 5:33:11 AM6/19/08
to
I like the idea of a limited 360 degree fov and an extended 90/180
degree fov in the implicit/explicit facing of the character. I'm
currently looking at my computer screen but have somewhere between 180
and 270 vision including peripherals, to turn my head and look behind
me takes... no time at all, I just did it, not even a second. If I see
something I might look a bit longer.

If you add hearing to a system like this then you could have '?' in
the unseen areas but within hearing range. This could alert you to
something out-side of your visual range, or even round a corner. If
you're playing a stealth rogue (MetalGearRL?) then this would be a
great thing to implement.

You don't necessarily need to implement every good idea in every game.
If you're having a pure hack -n- slash then full 360 degree fov is
great as you can see the hordes approaching. If you're playing
something a bit more indepth, perhaps with fewer monsters, more
challenging fights and the location-specific damage mentioned before
then the above idea might fit better. Horses for courses as an
equestrian might say.

Jeff Lait

unread,
Jun 19, 2008, 11:25:15 AM6/19/08
to
On Jun 19, 3:32 am, Inugami...@gmail.com wrote:
>
> POWDER plays fair because any negative random event-- be it reading a
> bad scroll, drinking a bad potion, getting punished by a pissed god,
> et cetera-- is survivable as long as the player takes precautions in
> order to survive (getting HP, IDing, getting gods on his/her side and
> doing less to piss off others in early game).

I still think the flamestrike punishment is too much like an out-of-
the-blue insta-death for my tastes. Yes, you can mitigate it in many
ways, but maybe I should do something like have an explicit warning
before gods toast you so you get a few turns to rethink and prep for
possible punishment...

Derek Ray

unread,
Jun 19, 2008, 11:57:56 AM6/19/08
to

The biggest problem I'm noting is that you can get yourself flamestruck
by following one particular god in the early game. That's no big deal,
since theoretically if you're Tsloshed or Paxie, you will be defended by
that god -- except that if you're following either of those gods
effectively, you are _also_ pissing off H'ruth, meaning that you may
burn god points defending from H'ruth and leaving yourself open for a
flamestrike or polymorph later.

Something I might suggest looking into (I haven't yet) is whether the
positive piety for certain gods (Tlosh and Pax) is enough to offset the
negatives for certain actions properly -- ie. casting "Heal" is worth +2
piety for Pax, but shouldn't be worth -2 or more hostility from Tlosh
because then the player isn't actually making any progress -- since he's
annoying H'ruth at the same time. (Or casting "heal" should be worth
more piety, perhaps, given that the piety will be eventually burned by
Tlosh's anger).

Now admittedly, the player shouldn't be able to perpetually keep himself
ahead of ALL the other gods; if he wants to be noisy, melee a lot, use
healing magic, etc., then he may well get himself behind the eight-ball
eventually. But having it be inherently negative just for trying to
keep one particular god happy is probably not a good way to go. The
inverse of this appears to be true as well; using too many Raise Undeads
or Summon Imps to keep yourself alive will eventually get you
flamestruck by Pax _or_ poisoned/prove-your-worthed by H'ruth.

Gerry Quinn

unread,
Jun 19, 2008, 12:10:23 PM6/19/08
to
In article <3cd6afcf-d60e-43df-987e-b612e86c00f2@
27g2000hsf.googlegroups.com>, heckr...@yahoo.com says...

> I'd love to point out mazecrawl as a fun game that impliments these
> things, but the site is down. How long do I have to wait before
> declaring it abandonware?

Wait till the author dies, and then another 70 years :-)

- Gerry Quinn

Paul Donnelly

unread,
Jun 19, 2008, 5:39:21 PM6/19/08
to
Billy Bissette <bai...@coastalnet.com> writes:

But if it's uncommon enough to be unexpected when it happens, no one
will miss its absence.

> What percentage chance should it be? One in a hundred? One
> hundred attacks can happen really quickly. That is only twenty
> enemies getting five attacks each. (Sure, you might go two or
> three hundred attacks without it happening. But it also might
> happen on the first.)
>
> One in a thousand? Face a few hundred enemies, and you can be
> on the receiving end of a thousand attacks. It still may be common
> enough to be a "certainty" in your game.
>
> Sure, you can make it even more uncommon. One in ten thousand.
> One in one hundred thousand. Now you are in the territory of blind
> chance ending a game. And existing Roguelikes *will* dance this
> particular dance, but it is a risky thing to play with, to get to
> the point that players don't feel robbed if it happens.

How about none percent? If the player is normally in situations where
being insta-killed would be surprising, it couldn't possibly be
unrealistic not to make it happen.

> And if you shove it off to some deep part of the dungeon, then it
> just becomes another form of deadly special attack.

Maybe you misunderstand me. I'm saying that by not confronting the
player with enemies who would be expected to insta-kill them (based on
their level, the enemy's level, and that enemy's description),
insta-killing can be left out of the game completely without stretching
realism. The player enounters enemies appropriate to their character's
level, rather than meeting high-level foes before being prepared. By the
time the player is deep in the dungeon, these foes can't insta-kill
because the player is too tough or too protected by armor and
amulets. If the player somehow gets there too soon, it's academic
whether the monsters down their instantly kill the player by dealing a
lot of damage, or with a special insta-kill feature. I mean, what's a
normal starting HP? 10 or 15, right? There are probably monsters in the
game that can deal much more damage than that (an insta-kill to a
low-level player), but they are deep in the dungeon.

R. Dan Henry

unread,
Jun 19, 2008, 9:01:27 PM6/19/08
to
On Tue, 17 Jun 2008 07:20:38 -0700 (PDT), I Own The Letter O
<lord...@yahoo.co.uk> wrote:

>Then I will fill my ears full of wax before
>luring a banshee to an underground cave full of living trees to get
>the treasure that they somehow carry with them!!!

That's not unrealistic. They keep it in the tree houses that Dwarven
children build in their branches.

--
R. Dan Henry = danh...@inreach.com
If you wish to put anything I post on your website,
please be polite enough to ask first.

zai...@zaimoni.com

unread,
Jun 19, 2008, 9:07:11 PM6/19/08
to
On Jun 19, 4:39 pm, Paul Donnelly <paul-donne...@sbcglobal.net> wrote:
> Billy Bissette <bai...@coastalnet.com> writes:

> > And if you shove it off to some deep part of the dungeon, then it
> > just becomes another form of deadly special attack.
>
> Maybe you misunderstand me. I'm saying that by not confronting the
> player with enemies who would be expected to insta-kill them (based on
> their level, the enemy's level, and that enemy's description),

Except that this is mostly designed out by going for a low/no-hp
system. Character level/abstracted skill, like effective armor, has
to do with *preventing* the potentially instant kill attack from even
happening. It doesn't do anything about the damage when it does
happen.

> insta-killing can be left out of the game completely without stretching
> realism. The player enounters enemies appropriate to their character's
> level, rather than meeting high-level foes before being prepared.

But the *situations* will be unrealistic. It's not enough to weaken
the enemies; a working game has to guarantee that with perfect play,
the instant-kill attack never happens. (E.g., one could have
effective armor that stops *one* attack and then has a probability of
failing that goes up with the number of attacks it has already
stopped. Then scenarios where perfect play involves only one attack
being attempted become fit for a game.)

In practice, without the unrealism of hp what ends up being
unrealistic is the AI (either in-game or in the scenario design).
Unless the point is to make a un-fun simulation rather than a game, in
which case all of the YOU LOSE elements in the design are tolerable
even though they break the simulation as a game.

Paul Donnelly

unread,
Jun 19, 2008, 11:26:30 PM6/19/08
to
zai...@zaimoni.com writes:

> On Jun 19, 4:39 pm, Paul Donnelly <paul-donne...@sbcglobal.net> wrote:
>> Billy Bissette <bai...@coastalnet.com> writes:
>
>> > And if you shove it off to some deep part of the dungeon, then it
>> > just becomes another form of deadly special attack.
>>
>> Maybe you misunderstand me. I'm saying that by not confronting the
>> player with enemies who would be expected to insta-kill them (based on
>> their level, the enemy's level, and that enemy's description),
>
> Except that this is mostly designed out by going for a low/no-hp
> system. Character level/abstracted skill, like effective armor, has
> to do with *preventing* the potentially instant kill attack from even
> happening. It doesn't do anything about the damage when it does
> happen.

It won't happen. The player will not encounter enemies they expect to be
insta-killed by, so there will be no provision in the game engine for
insta-killing.

>> insta-killing can be left out of the game completely without stretching
>> realism. The player enounters enemies appropriate to their character's
>> level, rather than meeting high-level foes before being prepared.
>
> But the *situations* will be unrealistic. It's not enough to weaken
> the enemies; a working game has to guarantee that with perfect play,
> the instant-kill attack never happens. (E.g., one could have
> effective armor that stops *one* attack and then has a probability of
> failing that goes up with the number of attacks it has already
> stopped. Then scenarios where perfect play involves only one attack
> being attempted become fit for a game.)

So... I don't write the code for instant kills, and it doesn't
happen. Problem solved.

The *issue* is how to maintain combat's verisimilitude without instant
kills. I don't think it's a stretch, with enemies near your level, for
insta-kills not to happen.

> In practice, without the unrealism of hp what ends up being
> unrealistic is the AI (either in-game or in the scenario design).

Is the player a mind-reader to know why their head hasn't been chopped
off? To insta-kill, the attacker needs to want to insta-kill (accidents
are possible but even more unlikely), have enough skill to hit a vital
spot, the defender needs to be too unskilled to deflect the blow, the
attacker needs to be armed with something appropriate, the defender
can't be armored where the attacker strikes, and there's an element of
chance as well. A vital blow can go wrong at any one of those steps, and
RLs, abstract as they are, don't need to make excuses for failed
attacks.

Giving the player a minor run of good luck in this one regard isn't
going to destroy realism. If the game is any good, why would any player
be thinking, "Man, I never get insta-killed. This game is so
unrealistic!"?

zai...@zaimoni.com

unread,
Jun 20, 2008, 1:08:23 AM6/20/08
to
On Jun 19, 10:26 pm, Paul Donnelly <paul-donne...@sbcglobal.net>
wrote:

> zaim...@zaimoni.com writes:
> > On Jun 19, 4:39 pm, Paul Donnelly <paul-donne...@sbcglobal.net> wrote:
> >> Billy Bissette <bai...@coastalnet.com> writes:

> So... I don't write the code for instant kills, and it doesn't
> happen. Problem solved.

Asserting a tautology does not constructively demonstrate that the
tautology permits a given objective.

GURPS has been out for some time, and is a constructive demonstration
that a realistic combat system can be completely unsuitable for heroic
fantasy. The vaporware system I've been working on since 2003-ish
behaves similarly, with a far more geometric interpretion of "to-hit/
to-aim" rolls.

> The *issue* is how to maintain combat's verisimilitude without instant
> kills.

Yes.

> I don't think it's a stretch, with enemies near your level, for
> insta-kills not to happen.

Having casually studied Tai Chi, and thought through whether it would
make sense for me to apply for a KS CCW license: I can assume this for
unarmed combat, or similar situations involving at least partially
effective non-bypassable armor -- assuming no surprise. (This
describes a fencing match fairly well.)

It would, of course, be possible to contrive levels and incompetent AI
such that no-surprise is a guaranteed assumption with correct play.
Non-bypassable armor is harder to backstory, except in science fiction
settings. For sake of discussion, let's assume that correct play
guarantees no surprise, and that surprise without fully effective
armor implies non-zero probability of an attack being instant-kill.

> > In practice, without the unrealism of hp what ends up being
> > unrealistic is the AI (either in-game or in the scenario design).
>
> Is the player a mind-reader to know why their head hasn't been chopped
> off? To insta-kill, the attacker needs to want to insta-kill (accidents
> are possible but even more unlikely), have enough skill to hit a vital
> spot,

Which is pathetically easy with minimal training. Training away
mental blocks causing automatic misses isn't so minimal, let alone
hoplophobia and its lower-technology analogs.

> the defender needs to be too unskilled to deflect the blow,

Out of position to deflect/evade the blow/shot is enough. Skill would
refer to staying *in* position, being aware, and possibly reflexively
rolling with the blow.

> the attacker needs to be armed with something appropriate, the defender
> can't be armored where the attacker strikes,

Fully effectively armored. (This also describes a fencing match
fairly well).

Again, skill shows up here -- if the defender has reaction time,
evasion can include turning to present an unfavorable angle of attack
(possibly upgrading partially effective armor to fully effective
armor).

> ....

> Giving the player a minor run of good luck in this one regard isn't
> going to destroy realism. If the game is any good, why would any player
> be thinking, "Man, I never get insta-killed. This game is so
> unrealistic!"?

Simulationist players. [E.g., one time I tweaked the top-down shooter
Meteor ( http://www.jbgames.com/games/meteor/ ) to let me play with
only 15 hp (what normal soldiers had) rather than 100. You can get
moderately far in the game before the fact the levels were designed
for hp inflation catches up with you.]

David Damerell

unread,
Jun 20, 2008, 1:16:45 PM6/20/08
to
Quoting <heckr...@yahoo.com>:
>David, I think you're missing the point:

No, I don't. I've seen the realism fetish dozens of times, and it's
unmitigated drivel. You've got nothing new.

>That a balance between realism and playability is important.

Er, no. Playability is king. Why would you want to balance it with
anything?

[Of course _part_ of playability is world consistency, but that _is not
realism_.]

>There is no ultimate realism.
>It's just something to work towards. Often to the extent that you kill
>the game.

Why on earth would you want to work towards something that kills the game
and has to be balanced off against playability?


--
David Damerell <dame...@chiark.greenend.org.uk> Kill the tomato!

Today is Oneiros, Presuary.

David Damerell

unread,
Jun 20, 2008, 1:28:37 PM6/20/08
to
Quoting Paul Donnelly <paul-d...@sbcglobal.net>:
>It won't happen. The player will not encounter enemies they expect to be
>insta-killed by, so there will be no provision in the game engine for
>insta-killing.

A fine and sensible game design decision. But we're not arguing with the
game designers; we're arguing with the realism fetishists; and what you
are doing is sacrificing a modicum of realism for gameplay.

David Damerell

unread,
Jun 20, 2008, 1:26:37 PM6/20/08
to
Quoting Paul Donnelly <paul-d...@sbcglobal.net>:
>David Damerell <dame...@chiark.greenend.org.uk> writes:
>>Except that - unlike quantum tunneling - a single hit being fatal or so
>>incapacitating as to render a human incapable of fighting is a perfectly
>>plausible result when everyone's waving bloody great weapons around. People
>>are incapacitated by single stab wounds from tiny shivs; that being bashed
>>over the head with a mace could kill you is not a highly improbable effect
>>unworthy of implementation at all.
>What if my character encounters only low-level beasts before acquiring
>enough combat skill and/or armor to effectively guard his head against
>foes of his level?

Works fine in most of the unrealistic combat systems we have. Doesn't work
so well in the realistic ones that were being advocated, because
_realistically_ the human body, while tough and resilient, has a few
extremely vulnerable areas, like the eyes; and _realistically_ to mount a
perfect defence requires a vastly inferior opponent.

The problem for the realism fetish comes in trying to make opponents
challenging. If I can be assured that an opponent's (say) ranged attack is
so feeble that it can never pose a threat of hitting me in the eye and
blinding me - what is it, a child with a Nerf gun? How can such an attack
pose me any challenge at all?

R. Dan Henry

unread,
Jun 20, 2008, 4:39:45 PM6/20/08
to
On Mon, 16 Jun 2008 15:54:58 -0700 (PDT), heckr...@yahoo.com wrote:

>Having un-realistic combat reduces the suspension of disbelief, and
>otherwise makes the game less fun.

I would not say that it reduces suspension of disbelief (which I'm not
sure is even an applicable idea in a roguelike). It isn't as if even the
experts in historical combat agree on what would be realistic in the
first place. What you mean is not realism, but something that meets your
assumptions about combat, which might be entirely wrong.

>But I'd have stress
>that every wound impairs you in some way. The balance? Well, how about
>a game that reduces your stats by a factor of your hit points over
>your max hit points? Area-specific damage is a leap forward in both
>realism AND playability, but, well, it's hard and I'm lazy.

Wounds that impair tend to lead to "he who hits first, wins", which is
almost certainly realistic, but much like an instadeath that torments
you by not playing out instantly. Hit locations are generally not
handled in a genuinely realistic way, but have the possibility of
enhancing play (and the possibility of greatly degrading play).

>Outputting the effects to the user. Should our combat formula be open
>or closed to the player?

That has nothing to do with how realistic combat is.

>I think the open format would be better, although most games are
>closed. I'm not sure, I just like to know what is going on. Then
>again, I would probably enjoy a game implemented in excel.

Many roguelikes are open source, which means that the combat formulas
are available to anyone who wants to look at them.

>"Going Epic". In any situation where a human goes up against the
>stereotypical dragon and survives, nay, can EXPECT to survive, realism
>has been thrown right out the window.

But at that point, you are playing a HERO, and heroes traditionally *do*
kill monsters, including dragons.

>So in this, I have to agree that realism can take
>it in the pants.

In a realistic roguelike, a dragon is a big lizard from an island and as
long as I shoot it before its infectious bite gets me, I'll be fine.

>Limiting abilities is an important factor of realism that shouldn't be
>broken. For such games that HAVE ability scores (STR, DEX, INT, ETC)
>most of them have some sort of maximum that you can reach and go no
>further. This is realism in play. There are limits to how much a man
>can lift and how fast he can run.

And the limits in these games are generally beyond those limits. The
limits are there for gameplay reasons, not realism.

zai...@zaimoni.com

unread,
Jun 20, 2008, 5:57:07 PM6/20/08
to
On Jun 20, 12:26 pm, David Damerell <damer...@chiark.greenend.org.uk>
wrote:
> Quoting Paul Donnelly <paul-donne...@sbcglobal.net>:

>
> >David Damerell <damer...@chiark.greenend.org.uk> writes:
> >>Except that - unlike quantum tunneling - a single hit being fatal or so
> >>incapacitating as to render a human incapable of fighting is a perfectly
> >>plausible result when everyone's waving bloody great weapons around. People
> >>are incapacitated by single stab wounds from tiny shivs; that being bashed
> >>over the head with a mace could kill you is not a highly improbable effect
> >>unworthy of implementation at all.
> >What if my character encounters only low-level beasts before acquiring
> >enough combat skill and/or armor to effectively guard his head against
> >foes of his level?

Armor, as combat skill will *not* be sufficient.

> ....


>
> The problem for the realism fetish comes in trying to make opponents
> challenging. If I can be assured that an opponent's (say) ranged attack is
> so feeble that it can never pose a threat of hitting me in the eye and
> blinding me - what is it, a child with a Nerf gun? How can such an attack
> pose me any challenge at all?

Precisely. Once the choice for realism over playability is made,
opponents with such feeble attacks are *themselves* unrealistic and
thus banished by consistent game design; the player's armor cannot be
uniformly fully effective. Other means of guaranteeing zero attacks
against not-fully-effective armor are required.

Paul Donnelly

unread,
Jun 20, 2008, 7:54:26 PM6/20/08
to
David Damerell <dame...@chiark.greenend.org.uk> writes:

> Quoting Paul Donnelly <paul-d...@sbcglobal.net>:
>>It won't happen. The player will not encounter enemies they expect to be
>>insta-killed by, so there will be no provision in the game engine for
>>insta-killing.
>
> A fine and sensible game design decision. But we're not arguing with the
> game designers; we're arguing with the realism fetishists; and what you
> are doing is sacrificing a modicum of realism for gameplay.

That's true, although I'm also adding [what I consider to be more]
realism through the use of whatever combat system it is that I'm
supposed to be proposing.

I guess my argument with realism fetishists is whether disallowing
insta-kills is such a blow to realism that it's out of the question. My
position being that it's just fine and the player will never remark it,
and theirs being... something else, not succinct. I guess that players
will scoff at any attempt at realism that doesn't include insta-kills.

Different definitions of realism, perhaps. Mine being that mechanisms
that can't even be described in real-world terms (HP) are unrealistic,
and others are realistic to varying degree. Theirs being complicated,
not fun, and involving accurate hangnail simulation.

Paul Donnelly

unread,
Jun 20, 2008, 8:07:16 PM6/20/08
to
David Damerell <dame...@chiark.greenend.org.uk> writes:

> Quoting Paul Donnelly <paul-d...@sbcglobal.net>:
>>David Damerell <dame...@chiark.greenend.org.uk> writes:
>>>Except that - unlike quantum tunneling - a single hit being fatal or so
>>>incapacitating as to render a human incapable of fighting is a perfectly
>>>plausible result when everyone's waving bloody great weapons around. People
>>>are incapacitated by single stab wounds from tiny shivs; that being bashed
>>>over the head with a mace could kill you is not a highly improbable effect
>>>unworthy of implementation at all.
>>What if my character encounters only low-level beasts before acquiring
>>enough combat skill and/or armor to effectively guard his head against
>>foes of his level?
>
> Works fine in most of the unrealistic combat systems we have. Doesn't work
> so well in the realistic ones that were being advocated, because
> _realistically_ the human body, while tough and resilient, has a few
> extremely vulnerable areas, like the eyes; and _realistically_ to mount a
> perfect defence requires a vastly inferior opponent.

I think it would work fine in a variety of combat systems which could be
classed as realistic relative to other existing combat systems which
have in the past been described as less realistic than might be
desirable to a person hoping for a combat system which is a real-world
near-analog.

> The problem for the realism fetish comes in trying to make opponents
> challenging. If I can be assured that an opponent's (say) ranged attack is
> so feeble that it can never pose a threat of hitting me in the eye and
> blinding me - what is it, a child with a Nerf gun? How can such an attack
> pose me any challenge at all?

I've got some of those nifty side shields on my glasses?

I see what you're getting at there, and I agree with you in the case of
a "stuck in a cave full of horrible beasts" simulator. But I want to
make it clear that I am not and never was advocating a futile attempt at
total realism. Just a combat and damage system based on the shape of the
body and the general effects of wounds in different places (arms, legs,
head). Would it be just another one of "the unrealistic combat systems
we have"? Compared to the real world, I suppose so, and if that's your
holy grail I'll just leave you [of course not meaning you, personally,
David] to seeking it. Compared to HP I think it would be a pretty good
increase in realism and potentially more fun as well.

I think it's possible to devise a combat sytem that a reasonable person
would call realistic, but which doesn't include every single one of the
unpleasant possibilities of the real world.