Crawl-SS: Weapon & armor changes
lemuel...@gmail.com
weapon/armor base types in Stone Soup?
>From my casual observation, it seems that many, but not all, weapons
had their to-hit modifeirs greatly reduced. E.g. long swords went from
10+1 to 10-2, war axes went from 11+0 to 11-3, etc. On the other hand,
maces, daggers, hand axes and most other smaller weapons are unchanged.
And giant (spiked) clubs got a damage boost as well as a to-hit
penalty. (And they can also be wielded by non-large races -- a change I
am definitely NOT fond of. Giant clubs were a big part of what made
Ogres special.) I'm not saying these changes are bad, but I'd like to
know what the thinking was behind them. I don't recall a lot of
complaints about weapon base-type balance in Crawl -- in fact, other
than polearms, it seemed just about perfect. So why alter it so
radically?
Similarly, heavy armors got their evasion penalties raised
significantly. Were they really overpowered relative to dodging -- as I
recall, the consensus if anything leaned the other way. So what waas
the thinking?
Also, on the subject of evasion, it would be nice if the dispalyed
evasion took into account all factors, including skill and Dex. Where
do I file a feature request, again?
Lemuel
Darshan Shaligram
> Could Eric (or someone) explain the thinking behind the changes to
> weapon/armor base types in Stone Soup?
We've imported 4.1's item properties. This has introduced some bugs.
Oh, and s/Eric/Erik/. Erik is subtle and slow to anger, but on this
point he's quicker on the trigger than Xtahua. :-)
> And giant (spiked) clubs got a damage boost as well as a to-hit
> penalty. (And they can also be wielded by non-large races -- a
> change I am definitely NOT fond of.
This is a bug, we'll fix it for 0.1.2.
> I don't recall a lot of complaints about weapon base-type balance in
> Crawl -- in fact, other than polearms, it seemed just about
> perfect. So why alter it so radically?
In practise, you won't see much difference, especially as far as melee
weapons are concerned. Oh, if you minmax and work out your average
damage, there is a difference, but in actual gameplay, it's not a
killer, or even very noticeable.
Brent's general intent was to make it difficult for lucky finds of
high-end weapons to completely change the game for low-level
characters. In stone_soup (and 4.1), not only do you need some skill
to use the big weapons effectively, you likely also need some
enchantment on the weapon.
This parallels spellcasting. If you find, say, the Book of
Annihilations on D:2, it's not as if it completely changes your game
experience because you need lots of skill before it becomes anything
more than a paperweight.
> Similarly, heavy armors got their evasion penalties raised
> significantly. Were they really overpowered relative to dodging --
> as I recall, the consensus if anything leaned the other way. So what
> waas the thinking?
Which heavy armours have been badly EV-penalised? There's been a bit
of rebalancing in armour EV, certainly, but nothing extreme.
[...]
> Where do I file a feature request, again?
Go to http://crawl-ref.sourceforge.net and click on Feature Requests.
--
Darshan Shaligram <scin...@gmail.com> Deus vult
lemuel...@gmail.com
Darshan Shaligram wrote:
> Brent's general intent was to make it difficult for lucky finds of
> high-end weapons to completely change the game for low-level
> characters.
OK. But was that really a problem? Finding a longsword -- or a
doublesword or quickblade or demon trident -- helped enough in the
early game to be fun and exciting, but do you really think it changed
the game so dramatically? Isn't modifying the stats of most of the
weapons in the game kind of a big change to fix a small-to-nonexistent
problem?
Before, if I found a longsword on lvl 1 I'd think "Cool!" and maybe
make the character a longblader even if I'd been planning on something
else. But it was no more gamebreaking than finding a platemail, or +2
cloak, or wand of draining, or... Now, I think "Meh. My starting mace
is way better." Are you sure this is an improvement?
Lemuel
Denis
Darshan Shaligram wrote:
>
> In practise, you won't see much difference, especially as far as melee
> weapons are concerned. Oh, if you minmax and work out your average
> damage, there is a difference, but in actual gameplay, it's not a
> killer, or even very noticeable.
>
what I noticed was that great mace (a common weapon) was 16 damage and
executioner axe (very rare) was 20 damage.
now both are 18 damage, and it makes less sense.
Also a difference between 16 and 20 base damage is rather large when
you take into account the multipliers from strength and skills. I think
weapon damages need to be adjusted.
lemuel...@gmail.com
Denis wrote:
> what I noticed was that great mace (a common weapon) was 16 damage and
> executioner axe (very rare) was 20 damage.
>
> now both are 18 damage, and it makes less sense.
Yeah. It all makes less sense. Used to be a longsword was 10+3 and a
scimitar was 11+1 -- the long sword was slightly superior overall, but
if you really wanted more damage the scimitar might be a better choice
(and was a little more common, I think). Now, the long sword is 10-2
and the scimitar is 11-2 -- the scimitar is simply a superior weapon.
Why?
(I can't help but wonder if it was a great idea to import Brent's
weapon stats changes wholesale, when they were presumably (a) a work in
progress and (b) integrated with his larger changes to the combat code,
which you very wisely did not import.)
Darhsan, Erik, I don't mean to be critical -- you all did great work
and I thank you from the bottom of my crawl-playing heart. But for the
next version, could we please go back to the b26 weapon stats?
Lemuel
lemuel...@gmail.com
Change in: Damage, To-Hit, Speed (counting decreases as positive)
Knife: +1, +5, +1
Dagger: +1, 0, +1
Quickblade: 0, 0, -1
Short Sword: 0, -1, +1
Sabre: 0, 0, +1
Knives got a lot better, quickblades got worse; sabres are now strictly
better than short swords.
Falchion: 0, 0, 0
Long Sword: 0, -5, 0
Scimitar: 0, -3, 0
Demon Blade: 0, -5, 0
Katana: 0, -5, 0
Double Sword: 0, -8, 0
Great Sword: 0, -4, 0
Triple Sword: -2, -4, +2
Longblades (except falchions) got hosed; a double sword, one of the
rarest weapons in the game, now has worse stats than a glaive, which
every third gnoll drops.
Whip: +1, +1, +1
Club: 0, -1, -1
Hammer: 0, +1, 0
Mace: 0, 0 ,0
Flail: 0, 0, 0
Ancus: 0, +1, 0
Morningstar: 0, -4, 0
Demon Whip: 0, +1, +1
Spiked Flail: 0, -4, 0
Eveningstar: 0, -4, 0
Great Mace: +2, -2, -1
Giant Club: +3, 0, -4
Giant Spiked Club: +2, -3, -4
Hand Axe: 0, 0, 0
War Axe: 0, -3, 0
Broad Axe: 0, -7, 0
Battleaxe: -1, -5, 0
Executioner's Axe: -2, -4, +2
Good axes got worse; executioner's axe (a maybe-once-a-game item) is
now barely distinguishable from a battleaxe.
Spear: +1, 0, +1
Trident: 0, +4, +3
Demon Trident: -2, +4, +3
Halberd: 0, 0, +3
Scythe: 0, 0, +2
Glaive: 0, 0, 0
Halberd: 0, 0, +3
Polearms got better, but also more similar to each other.
Overall, weak weapons got better, good ones got worse. So with the new
base types, we're basically trying to cram more weapons into a smaller
volume of potential-weapon-space, meaning less distinction between
weapon types. Which, needless to say, does not seem like an
improvement.
I assume that Erik and Darshan have not really thought about this yet,
and just stuck Brent's numbers in as a placeholder while they focused
on more pressing changes. But I suspect if the new values stay in
place long, I won't be the only one hollering about them.
Lemuel
erisdiscordia
[analysis and for that matter analyses of the new weapon numbers]
Darshan's made some changes heavily in the direction of b26, and we're
reviewing them right now.
Let's consider this exchange part of that review. :-)
dot
dot
dot
(By the way, Lemuel, a lot of insights below are things that never
would have occurred to me, and are really important, so I'm really
eager to see you stay in this discussion :-) )
> [A] now strictly better than [B}
Those are the ones where you convince me the most. Heck, the reason why
I like the armour changes is that scale mail is no longer strictly
better than ring mail, for example. :-)
I do like the principle of needing to switch back to a smaller weapon
sometimes. I don't agree that needing training to properly handle a
really big weapon makes them not a happy find -- it's not like that
training isn't coming up if you want it, and it's not like those
weapons show up every game even by the *endgame*, so knowing from the
very beginning that they're there if you want to use them is hardly
without coolness.
However, the numbers do still need work, as your analysis shows.
> Longblades (except falchions) got hosed;
I have the feeling that, with all due respect, Brent let his emotions
rule his rulechanges in some areas. There's also, for example, huge
punishments applied in the code for the sin of having been born a
caster, or of having been born an elf (hey! Long blades again!), and
meanwhile Brent said once that (paraphrased) he likes to play -- if I
understand correctly -- magicless non-berserkers or wut, and
(paraphrased again) that he wants a game balanced for them... which
unwise desire he apparently applied very, very literally.
> a double sword, one of the
> rarest weapons in the game, now has worse stats than a glaive, which
> every third gnoll drops.
Wow.
> Good axes got worse;
There was an overall theme in 4.1 of hosing good things.
I *do* see the idea.
Good things often indicates no-brainers; no-brainers are some of the
most destructively imbalancing elements you can have because they make
whole classes of items (except for themselves) meaningless for the
purpose of decision-making.
On the other hand, there need to be good things, because good things
make you shout "cool!", and shouting "cool!" is one of the most
important parts of the fun. :-)
So the task is to find good things that are not no brainers. Brent
tended to reduce this task to "eliminate good things because they're
no-brainers," I'm afraid, and **unless** (attention!) there was
something about the 4.1 mechanics that *really* altered the impacts of
the various numbers, the new properties were such a case, nastily
hidden under a lot of numbers.
> Overall, weak weapons got better, good ones got worse.
I *am* in favor of that, really, in the sense of ensuring that to, well
not everything, but to many more things than before, there is a season.
That is, I'm in favor of ensuring that, unlike in b26, there are times
when you'll really *want* to pull out something smaller than an
executioner's axe even though you have an executioner's axe.
Let's face it, weapon choices are *not* as flavorful as they could be
in b26. That fact does make an already very hard game a pleasant bit
easier, but there are other ways to do that without reducing the
variety of the game.
Then again, making things more flavorful will probably be about more
than weapon numbers. Part of the reason why one enchants up a single
weapon most of the time is that it's usually the most efficient way to
do things. At present, one enchants up a mega-weapon. With a "various
needs" setup for the weapon numbers, we might just see people sticking
to just a *midrange* weapon for their enchanting-up. And that really
just puts us back where we started.
So I'm in favor of a "2-pronged attack" on the b26 situation:
- new weapon numbers that lead to a variety of needs -- which will
ultimately mean worse accuracies for big weapons, like in 4.1 / SS 0.1,
just better thought-out ones, and
- a system where it's worth enchanting and carrying multiple weapons,
not the One Weapon of Doom.
> Overall, weak weapons got better, good ones got worse. So with the new
> base types, we're basically trying to cram more weapons into a smaller
> volume of potential-weapon-space, meaning less distinction between
> weapon types.
Would you be interested in helping draft new numbers that spread across
a wider space and avoid no-brainers while maintaining "cool!"'s?
> I assume that Erik and Darshan
All in all, we're a five-man team by now. But during the discussion, or
rather non-discussion of the import from 4.1, we were 2.
> have not really thought about this yet,
> and just stuck Brent's numbers in as a placeholder while they focused
> on more pressing changes.
Partially. I also really thought Brent had probably gotten it right,
especially due to seeing the "create multiple needs" philosophy shining
out of the numbers.
e.
bwross
// Darshan Shaligram wrote:
//
// > Brent's general intent was to make it difficult for lucky finds of
// > high-end weapons to completely change the game for low-level
// > characters.
//
// OK. But was that really a problem? Finding a longsword -- or a
// doublesword or quickblade or demon trident -- helped enough in the
// early game to be fun and exciting, but do you really think it
changed
// the game so dramatically?
There were a lot of comments (and even some complaints) about how
starting weapons were largely irrelevant since the PC would just scoop
up some midrange weapon 3-4 levels down and start using it. One of
Linley's key design rules is: If it's not going to be relevant, we
don't include it. This is why we don't bother with light sources (food
sources can cover the same effect and more).
The idea is not so much that finding some midrange weapon wouldn't be
useful for anybody (you're giving the accuracy stat far more weight
that it deserves, especially in b26 combat)... it's that fighter types
would be in a better position to move into the new weapon (having the
important skills... the -2 on a long sword is far from a big deal to
overcome, it's still easy), and mage types would have to deside if they
wanted to practice to become a fighter mage or rather stick with easy
to use short swords and go the more pure mage route. In other words,
it was about giving a small advantage to fighters, additional early
choices to everyone, and the starter weapons were left as an easy to
use backup for pure mages.
// Isn't modifying the stats of most of the
// weapons in the game kind of a big change to fix a
small-to-nonexistent
// problem?
The changes are really more insignificant than you think. The factors
are not equal. Accuracy is practically ignorable (and can't be pushed
too far... especially in the low granularity, wash everything into dust
to-hit system of b26), base damage is good because of damage
magnification in b26, and speed is by far the most imporant in that a
low frequency of attacks effects your chance of getting double hit, and
the base damage/speed ratio is the most significant factor in
determining the damage you can lay out in a given amount of "game
seconds" (as opposed to player typing turns).
The change was added before 4.1 combat was even really started, most of
the stats are only slightly tweaked from values I reworked for b26.
They were essentially calculated using stats and some simulations to
figure out how much damage the weapons would be doing over a period of
time (which was about 10 regular moves long), considering the frequency
and resultant variance. Because of the way the factors work, changes
that might look like a huge hosing to you aren't really one at all...
the different between what looks to be very similar weapons can be very
huge. For ingame comparison you really want to compare base/speed
first, speed second (only if the speed is slow and your skill is low),
base third, and not consider the accuracy unless forced at gunpoint.
There really is a sizeable advantage to the executioner's axe over the
battleaxe... that two point bonus is a good thing for base * good thing
for ratio (1/1 or better rocks in crawl)... the one point difference in
acurracy isn't something you'd be able to notice in play (possible
delude yourself into believing given the randomness of combat).
// Before, if I found a longsword on lvl 1 I'd think "Cool!" and maybe
// make the character a longblader even if I'd been planning on
something
// else. But it was no more gamebreaking than finding a platemail, or
+2
// cloak, or wand of draining, or... Now, I think "Meh. My starting
mace
// is way better." Are you sure this is an improvement?
Well, some of the complaints did run along the lines that the type of
starting weapon didn't matter so much since it was the type of the
first "real" weapon you found that would really count. So I'd have to
say, "yes"... if only for those people.
And really... is your choice here any differnent than a weak
spellcaster turning down that platemail because they won't be able to
case their spells with it on? It sounds to me like your complaining
about the game giving you more choices to make that aren't as trivial
(and I'm pretty sure that's not an unimprovement).
Brent
Darshan Shaligram
[...]
Hey, welcome back!
What do you think of stone_soup's general direction? Also, are you
planning to revive 4.1 development?
Rubinstein
> [...]
Huh, I've just seen a ghost! *scared*
No, seriously, glad to see you back in r.g.r.m! :-)
And, btw, these are exciting times for a come back...
--
Rubinstein
bwross
// However, the numbers do still need work, as your analysis shows.
His analysis is at least incomplete, if not flawed. He weighs factors
uniformly and independantly that are neither. If you really want to
work the table you need to actually do stats grinding or simulations.
The numbers given in the table can easily lie and confuse you and lead
you into creating weapons which look reasonable but when you consider
their output, they become Clear Choices (tm). I wouldn't take this
posted analysis as a guide for development... some tweaking is probably
needed (the values were tweaked slightly for 4.1 features you don't
have), but his comments misrepresent the magnitude of the actual
changes.
Not that the comments shouldn't be made, I'd just encourge him to look
deeper (consider the actual differences in to-hit chances and the
amount of damage the character can do with a weapon at various skill
levels per unit time) and then make comments.
Disclaimer: people shouldn't take things I say too personally, I'm
often a bit blunt and like to devil's advocate and debate things
(another part of this is that I've typically not bothered to post on
things that are correct, but to correct things when they're wrong... if
you take everything I don't respond to as an affirmation, I'm extremely
nice:). Attentive people might have noticed that in the past, various
little things which have been suggested by other people and I
seemingly disagreed with and strongly debated have occasionally shown
up in the code in some form after things get hammered out.
// > Longblades (except falchions) got hosed;
//
// I have the feeling that, with all due respect, Brent let his
emotions
// rule his rulechanges in some areas. There's also, for example, huge
// punishments applied in the code for the sin of having been born a
// caster, or of having been born an elf (hey! Long blades again!)
Okay, I have no fondness for elves... but I never really went after
them in Crawl, I mostly left them alone. About the only significant
change I made was to cut their stats a bit when I started focusing on
making stats more relevant (because to not do that would have made them
uber... it was that or hose them in some other way like slow their
leveling up even further).
I declared no Jyhad on casters... if it looks like there is one, it's
because crawl is a spellcaster intensive game. That's where a lot of
the code is, player's are, and comments about balance issues came from.
It's a filter effect... with so much that needs/wants to be done, the
focus goes to where the action is (just look at the fighting code... it
doesn't change much over time until 4.1, most people just assumed that
things were working as they expected because things would fall over
dead when they hit them, and the system's so random that it's easy to
delude oneself into believing all kinds of things are relevant when the
code dramatically proves otherwise).
Oh, my personal (real life, gaming) preferences actually lean towards
long swords and polearms.
// meanwhile Brent said once that (paraphrased) he likes to play -- if
I
// understand correctly -- magicless non-berserkers or wut, and
// (paraphrased again) that he wants a game balanced for them... which
// unwise desire he apparently applied very, very literally.
That's pretty strong paraphrasing that mixes and matches many comments
in different contexts. Traditionally, in D&D I liked to play fighters
and priests over mages. I liked to think of interesting things that
they could do rather than the interesting things that get spoonfed to
mages (called "spells"). They required more work (well, less work now
that D&D added feats). I've also played (and really enjoy)
berserkers... especially with the template that I worked out that
combined the Viking Handbook with the Fighter's Handbook versions. It
mixed flavour (Viking shapeshifting) with some of the choice
disadvantages (hand over your character sheet and let the DM handle
your damage while feeding you such wonderful obscure hints at "The axe
tickles you!").
In roguelikes, I like to play a varity of characters in a varity of
styles (Real Man, Role Player, Loonie, and even Munchkin if I'm playing
a game for the first time (or just looking to see how much I can
brutalize an old one). I like variety (more than most), choices that
aren't braindead obvious, and limitations that real but not iron bars.
For example, the ability of controlled blink to allow a character to
quickly zip out of line of sight and essentially move at an
unbelievable speed across several levels to a place you can safely
rest... this was braindead easy and obvious. Adding the magical
contamination factor to controlled teleportation gives it a limitation
(can't use it endlessly without potentially paying for it in a nasty
way) but doesn't stop it from being used for the occasional one/two hop
out of line of sight.
My largest desire for fighter types in crawl was for mechanics that
were readable and maintainable (the two separate monster functions in
fight.cc were horrid and out of date with each other... the merger of
the two in 4.1 cleaned up so much there), factors that made sense (the
to-hit calculation in b26 and before is a joke that's been tweaked by
several people, again and again, without anyone bothering to understand
the whole... it lacks granularity and tries to deal with it in a way
that really prevents adding anything new or making things relevant...
the test_hit() system in 4.1 makes sense, provides a good diagnositic
for development and maintenance and furthermore was extended to be
consistant with other EV test systems like the beam code), and variety
of meaningful choice (as opposed to "Me trade up to bigger stick and
thicker skins as they come" mentality).
// > a double sword, one of the
// > rarest weapons in the game, now has worse stats than a glaive,
which
// > every third gnoll drops.
//
// Wow.
It's not... the speed factor is worth far more credit to a mid-level or
high-level character than he gives it credit. For a low-level
character they're fairly equivalent (other things like skills and
handedness will decide). Which is appropriate... a low level character
shouldn't really have as much variance in their weapon's output as a
high level character (who can better leverage the advantages of a
better designed weapon). Even for a low-level character the "worse" is
pretty illusionary... the speed disadvantage leads to more double-hits
in the discrete nature of a roguelike, which is the worst thing.
// On the other hand, there need to be good things, because good things
// make you shout "cool!", and shouting "cool!" is one of the most
// important parts of the fun. :-)
I can agree with that... but I don't think that here is a good place to
put that. The problem is that there's a narrow window to fit the stats
into... it looks large, but accuracy can't be tweaked very much and
base and speed are not independant with speed being too dominant (I was
tempted to completely hose it down in 4.1). There's room for a few
items "under the curve"... giving a monster a piece of bad equipment is
a useful thing for balancing. The other's need to be close enough to
each other (to prevent one item or class becoming the nobrainer) while
still not going over the top and resulting in insane damage leading to
insane monster HP inflation (this of course is one of the problems with
4.1... a lot of the problems with playablity come from me quickly
boosting monsters and adjusting things so that weenie casters could no
longer go around mowing down monsters with melee right from the start).
That's too much to ask from these stats (especially given the way
they're implemented). That's why I started adding (and trying to make
relevant) alternate properties to weapons to distiguish them and make
them cool: str/dex ratio, impact, 1-1/2 hand weapons, double weapons,
EV bonus, etc. There's more room to be creative there than there is in
to-hit, to-dam, and speed (where making the range of skills and stats
meaningful combined with the difference in speed from low-level to
high-level eat up so much of the flexibility).
// So the task is to find good things that are not no brainers. Brent
// tended to reduce this task to "eliminate good things because they're
// no-brainers," I'm afraid, and **unless** (attention!) there was
// something about the 4.1 mechanics that *really* altered the impacts
of
// the various numbers, the new properties were such a case, nastily
// hidden under a lot of numbers.
The real hiding is in the b26 system's calculation of to-hit and
to-dam. The numbers on that table will mostly still be relevant to
that system because the real factor is speed... and with the base
damage multiple system of b26 that just gets multiplied even further.
// Then again, making things more flavorful will probably be about more
// than weapon numbers.
That's the correct way to do it. Don't stretch the weapon numbers so
much, they just make a stick into a slightly different stick.
// - a system where it's worth enchanting and carrying multiple
weapons,
// not the One Weapon of Doom.
That was a large part of the reason for why I lowered the max
enchantment values in 4.1. Part of the problem is that enchanting
scrolls are a random and limited commodity which tends to promote the
behaviour of saving them for a single item. I was very tempted to cut
things down to the barest bones (2 or 3)... but figured 5 would be more
acceptable. This helps in a number of ways: one is the obvious (more
excess scrolls to use on extra weapons) the other biggie is that it
reduces the granularity of enchantment on weapons... which means that
swords found "in the wild" with what used to be small pluses are now a
lot closer to the top end.
// Would you be interested in helping draft new numbers that spread
across
// a wider space and avoid no-brainers while maintaining "cool!"'s?
The problem with the numbers that look "cool!" is that once someone
does the math and posts the results we'll all know which weapon is the
New Unarmed Combat (tm). Before you really mess around with one of the
weapons stats you should at least try to estimate it's damage potential
over something like 10 turns and then compare it with results other
weapons you don't plan on changing at all... or hack out the relevant
sections of the fight code you intend to use the table with and write a
little simulator (better yet, yank those sections out into separate
functions and compile the simulator as a test against the same object
file).
// > I assume that Erik and Darshan
//
// All in all, we're a five-man team by now. But during the discussion,
or
// rather non-discussion of the import from 4.1, we were 2.
I'm available for consult if you need. I've been meaning to try and
get things together enough to hand over 4.1 code to someone. I haven't
been able to work seriously on 4.1 very much for a few months now (I
was on medication this summer that completely knocked me out for over a
month), and was pretty much looking at seeing if someone (I was going
to suggest Darshan) would be interesting in trying to salvage the good
from it and advance from b26. Then I just noticed that you're doing it
anyways. Hoorah to you guys!
// Partially. I also really thought Brent had probably gotten it right,
// especially due to seeing the "create multiple needs" philosophy
shining
// out of the numbers.
Well, I'm far from perfect... but I did try to err or the sides of
being conservative here. Things could possibly be stretched a bit, but
I'd be careful because what you're playing with here involves speed
(which can easily bite your foot off if you don't keep your eye on it).
The other problem is that the ranges for the numbers are quite small.
You can't stretch speed too far because once you get to 20 you're
already talking about a weapon that's defaulting to 1 attack every two
rounds (which is dangerously slow). And you can't stretch damage to
much with b26 combat because of the way it gets multiplied into
skill... increasing the spread in the ratio of high-level weapon damage
to low just gets magnified further. The end result is that you start
making the low end into more and more garbage and forcing all PCs into
advancing up into the big weapons.
In the end, it probably doesn't matter so much... if people really want
to help increase the variety in the weapons they should be thinking up
ways to introduce their performance flavours into the game. For
example, pole arms aren't reasonable as fast or power weapons so I
introduced the idea of their ability to hold opponents outside of their
comfortable melee range (ie an EV bonus).
Brent
bwross
// What do you think of stone_soup's general direction? Also, are you
// planning to revive 4.1 development?
>From what I've seen and heard so far it's pretty much what I was going
to suggest be done because I haven't been able to really do much with
4.1 lately (I keep looking at it sporadically which interferes with any
sort of work on it because I keep having to refresh myself in order to
get up to speed and that eats up too much of the time I have had).
There are certainly lots of things that can be salvaged from 4.1, and I
am available to answer questions.
Oh, and about those draconians... I added them from Gavin's patch
because they sort of met one of my intents (which was to physically
beef up the the monster deep elves in some way because they just
weren't robust enough for occuring at the bottom of the main dungeon).
But I decided to fix one of the nagging things that he didn't hack into
it... by making sure that each got a proper colour, including the
specialist ones. This, of course, did make them stronger than they
were originally... and they were very strong (although some details did
get tones down). So I certainly support tuning them down.
Brent
erisdiscordia
> erisdiscordia wrote:
> nice:). Attentive people might have noticed that in the past, various
> little things which have been suggested by other people and I
> seemingly disagreed with and strongly debated have occasionally shown
> up in the code in some form after things get hammered out.
He called me attentive! Awwww!
(More in a while, I don't really want to answer this one in a hurry.
Well, except to say thanks for the enlightenment. Well, and -- could
you explain more about why you state that stretching the range from
lowest to highest accuracy wouldn't work? And if that's in the context
of the b26//SS0.1 code, the 4.1 code, or both?)
e.
Tina Hall
> lemuel...@gmail.com wrote:
> [analysis and for that matter analyses of the new weapon numbers]
Don't know where to put that in all that text, so:
My suggestion is putting the old (b26) numbers back in and then looking
at how to make the weapons (and armour) that would otherwise be
neglected more attractive. (Not that I see any of them neglected.)
The big weapons should keep their big power. It was hard enough to kill
something tough with them, after all.
>> [A] now strictly better than [B}
> Those are the ones where you convince me the most. Heck, the reason why
> I like the armour changes is that scale mail is no longer strictly
> better than ring mail, for example. :-)
I don't understand what you mean here.
I can't tell all the foo mail apart anyway, only plate mail stands out,
and in one recent game I remembered that banded mail was heavy.
Among the rest, I only remember that some weren't too bad on
spellcasters, which made the lighter ones an option, at least
intermediate.
> I do like the principle of needing to switch back to a smaller weapon
> sometimes.
I don't. I like the idea to slowly move up as I find better weapons.
Others only are an option if they're, for example, the only source of
poison resistance, for someone who doesn't use weapons to fight.
You only have 52 slots, and what you propose means wasting 2 or more on
extra weapons just in case. That's just a nuisance and no fun.
I'm speaking from experience because for most of the early and mid-game
I already drag all sorts of weapons around 'just incase', plus other
items 'just incase'. Leaving hardly any room for new loot.
I never knew what to wield with the old system. Improvement would be
some clarity.
>> Good axes got worse;
> There was an overall theme in 4.1 of hosing good things.
> I *do* see the idea.
I don't.
> Good things often indicates no-brainers; no-brainers are some of the
> most destructively imbalancing elements you can have because they make
> whole classes of items (except for themselves) meaningless for the
> purpose of decision-making.
Then make the other classes more attractive, rather than removing
powerful weapons.
>> Overall, weak weapons got better, good ones got worse.
> I *am* in favor of that, really, in the sense of ensuring that to, well
> not everything, but to many more things than before, there is a season.
> That is, I'm in favor of ensuring that, unlike in b26, there are times
> when you'll really *want* to pull out something smaller than an
> executioner's axe even though you have an executioner's axe.
I don't see the point in that. It is rare enough that when you find one,
it's something to keep. Now (going by lemuel's numbers), the "Neat!"
when I find one is completely smothered and nonexistent. Might as well
throw out half the weapons if they're all the same anyway.
> Let's face it, weapon choices are *not* as flavorful as they could be
> in b26. That fact does make an already very hard game a pleasant bit
> easier, but there are other ways to do that without reducing the
> variety of the game.
How did it make it easier if the really tough weapons were hard to get
in the first place, and you had to use a medium one? And then drag all
the others around (protection, holy wrath, whatnot), just incase you ran
into something where that would be preferable?
> Then again, making things more flavorful will probably be about more
> than weapon numbers. Part of the reason why one enchants up a single
> weapon most of the time is that it's usually the most efficient way to
> do things. At present, one enchants up a mega-weapon. With a "various
> needs" setup for the weapon numbers, we might just see people sticking
> to just a *midrange* weapon for their enchanting-up. And that really
> just puts us back where we started.
I don't understand your reasoning for why that should be done.
> - a system where it's worth enchanting and carrying multiple weapons,
> not the One Weapon of Doom.
What's the good in that? What's the point? How would that improve a
game? (The enjoyment, not making it easier.)
>> I assume that Erik and Darshan
> All in all, we're a five-man team by now.
Who are the others?
--
Tina the Chiller - a Priest of the Resplendent Notorious Glitter
Darshan Shaligram
> erisdiscordia wrote:
> // However, the numbers do still need work, as your analysis shows.
> If you really want to work the table you need to actually do stats
> grinding or simulations.
We've been doing this (simulations and stat-analysis) for missile
weapons, but we've not applied it to melee yet.
[...]
> I'm available for consult if you need.
Thank you! We'll probably take you up on that offer.
[...]
> I've been meaning to try and get things together enough to hand over
> 4.1 code to someone. I haven't been able to work seriously on 4.1
> very much for a few months now (I was on medication this summer that
> completely knocked me out for over a month), and was pretty much
> looking at seeing if someone (I was going to suggest Darshan) would
> be interesting in trying to salvage the good from it and advance
> from b26. Then I just noticed that you're doing it anyways. Hoorah
> to you guys!
Thanks again, we appreciate the encouragement. And 4.1's been very
useful - in addition to being the source for most of the new stuff,
it's been a great place to look for answers when something in 4.0's
source is puzzling.
R. Dan Henry
<scin...@gmail.com> wrote:
>lemuel...@gmail.com writes:
>
>> Could Eric (or someone) explain the thinking behind the changes to
>> weapon/armor base types in Stone Soup?
>
>We've imported 4.1's item properties. This has introduced some bugs.
>> I don't recall a lot of complaints about weapon base-type balance in
>> Crawl -- in fact, other than polearms, it seemed just about
>> perfect. So why alter it so radically?
>
>In practise, you won't see much difference, especially as far as melee
>weapons are concerned. Oh, if you minmax and work out your average
>damage, there is a difference, but in actual gameplay, it's not a
>killer, or even very noticeable.
>
>Brent's general intent was to make it difficult for lucky finds of
>high-end weapons to completely change the game for low-level
>characters.
Why? It didn't break the game, it was just a lucky find. A dagger of
venom will do more to keep you alive in the early game than a high-end
vanilla weapon.
>In stone_soup (and 4.1), not only do you need some skill
>to use the big weapons effectively, you likely also need some
>enchantment on the weapon.
So, even less reason for fighter-types to consider long blades.
>This parallels spellcasting.
Exactly. So, why is different good when it comes to long blades, but bad
when it comes to melee in general? In any case, I consider this one of
the perks of being a fighter. You don't get to throw around massive
energy surges of death, killing multiple opponents at once, before they
can ever reach you, at higher levels, but you might find a nice weapon
that gives you a boost in the early game.
>If you find, say, the Book of
>Annihilations on D:2, it's not as if it completely changes your game
>experience because you need lots of skill before it becomes anything
>more than a paperweight.
Being able to cast those spells *would* be game breaking. A triple sword
isn't. It's still safer to face an orc warrior by killing him with Throw
Frost before he closes with you, that to go toe-to-toe with a weapon
bigger than his.
--
R. Dan Henry
danh...@inreach.com
Darshan Shaligram
> On Tue, 26 Sep 2006 21:33:11 +0530, Darshan Shaligram
> <scin...@gmail.com> wrote:
>>Brent's general intent was to make it difficult for lucky finds of
>>high-end weapons to completely change the game for low-level
>>characters.
> Why? It didn't break the game, it was just a lucky find. A dagger of
> venom will do more to keep you alive in the early game than a
> high-end vanilla weapon.
No, it doesn't break the game, but if you're a short blade specialist
and luck into an executioner's axe and want to use that exclusively,
you should have a transition period where life is awkward. Those
changes aim to achieve that (they don't quite succeed because 4.0's
combat code is very forgiving). Most players will probably not even
notice the change unless they go around examining weapons with
'v'. I've played an axeman and a swordsman quite far in stone_soup,
and believe me, ladies and gentlemen, these changes are not going to
kill you.
In any case, we're removing the big discrepancies from 4.0 in 0.1.2
since they weren't reviewed by the stone_soup team but snuck in while
we were cleaning up code.
erisdiscordia
> On Tue, 26 Sep 2006 21:33:11 +0530, Darshan Shaligram
> <scin...@gmail.com> wrote:
> >lemuel...@gmail.com writes:
> So, even less reason for fighter-types to consider long blades.
Sorry to respond to just one part of what you wrote, but your "even
less" is the most interesting, or rather shocking, because frankly, if
you sat me down at gunpoint with b26 in front of me and said "ascend or
die," the likelihood that long blades would be a part of what I came up
with to save my skin is very, very high. Even higher than for short
blades. :-)
e.
Tina Hall
>>> Brent's general intent was to make it difficult for lucky finds of
>>> high-end weapons to completely change the game for low-level
>>> characters.
>> Why? It didn't break the game, it was just a lucky find. A dagger of
>> venom will do more to keep you alive in the early game than a
>> high-end vanilla weapon.
> No, it doesn't break the game, but if you're a short blade specialist
> and luck into an executioner's axe and want to use that exclusively,
> you should have a transition period where life is awkward.
Which was the case before you made heavy weapons no longer heavy and
thus no fun.
I'm getting ever more worried that you change too much and ruin a fine
game.
Don't fix what isn't broken; executioner's axes weren't. Now they are.
> Those changes aim to achieve that
Except they remove any reason for a transition period, when that axe is
no longer a nifty find.
> Most players will probably not even notice the change unless they go
> around examining weapons with 'v'.
I do that anyway, because I can never remember what's better. Same for
foo mail armour.
> I've played an axeman and a swordsman quite far in stone_soup,
> and believe me, ladies and gentlemen, these changes are not going to
> kill you.
They just kill the fun.
> In any case, we're removing the big discrepancies from 4.0 in 0.1.2
> since they weren't reviewed by the stone_soup team but snuck in while
> we were cleaning up code.
What are the big discrepancies from 4.0? (Confused about versions here,
and what you will remove; the changes between b26 and your version, or
something I can't even guess?).
--
Tina the Nimble - an Elder of the Reprobate Nutty General
Martin Read
>I can't tell all the foo mail apart anyway, only plate mail stands out,
>and in one recent game I remembered that banded mail was heavy.
All non-elven mail is heavy.
--
Martin Read - my opinions are my own. share them if you wish.
\_\/_/ http://www.chiark.greenend.org.uk/~mpread/dungeonbash/
\ / "the lights shine clear through the sodium haze the night draws near
\/ and the daylight fades" -- Sisters of Mercy, "Lights"
Tina Hall
> Tina...@kruemel.org (Tina Hall) wrote:
>> I can't tell all the foo mail apart anyway, only plate mail stands
>> out, and in one recent game I remembered that banded mail was heavy.
> All non-elven mail is heavy.
I mean heavy to carry, and IIRC one of the foo mail wasn't so bad for
spellcasting, either. (Ring mail?)
--
Tina the Yeoman - a Priest of the Reliably Nuclear Gothic
lemuel...@gmail.com
is all I really wanted -- and even provoked the reappearance of Brent!
I'm not as sold as you are on the 4.1 philosophy -- not that I'm
against it, it just seems to me that of the various things one might be
unhappy with in b26, the stats of melee weapons would be pretty low on
the list. Especially with shields improved, I actually don't think
there are a lot of no-brainers, and where there are, I would prefer to
see some more targeted fixes rather than the wholesale rearrangements
of 4.1/SS.
I also suspect that the best solution will involve *enlarging*
weapon-space, either by making variation in the existing stats more
important (accuracy especially) or introducing some new stats (as I
understand Brent did with the evasion bonus to polearms in 4.1, which
seemed like a promising idea.) Of course that will be much trickier to
balance and code -- but that's just another reason to keep the b26
weapons for now.
Brent is certainly right that to do this seriously you'd have to
actually simulate various combinations, and not just eyeball the
numbers.
I did want to respond to Brent on one point: damage/speed is the
relevant stat only if monster armor is unimportant. For heavily armored
monsters, you want more damage even if the damage/speed result is
worse. This dynamic seems like one of the easiest ways of making for
less obvious weapon choices. It should be easy to create a situaion
where for most monsters a war axe (e.g.) is the best choice, but for
stone trolls and other hard-skinned brutes, a battleaxe is much better.
But I don't want to start making lots of suggestions until Stone Soup
has stabilized a bit and there's some kind of consensus about the
direction of possible changes.
erisdiscordia wrote:
> Would you be interested in helping draft new numbers that spread across
> a wider space and avoid no-brainers while maintaining "cool!"'s?
Yeah. Very interested. Maybe I'll pull something together this weekend.
Lemuel
R. Dan Henry
Axes are very nice and by the time I can expect to find even a lousy
long blade, I'll have a lot of experience invested in my starting weapon
type. Monks and fighter-mages (who will be started as mages, so not
burning XP on weapons until a preferred one is found) look at long
blades as a possibility, but I find they rarely show up when they're
actually an option (whereas great swords like appearing early for
inappropriate characters).
Then again, if I had to ascend or die, I'd be dead.
R. Dan Henry
>Good things often indicates no-brainers; no-brainers are some of the
>most destructively imbalancing elements you can have because they make
>whole classes of items (except for themselves) meaningless for the
>purpose of decision-making.
The problem is that this isn't a problem for very rare items. These can
be quite good, no-brainer types, at least some of the time. (It is
unlikely to ever be a no-brainer except for a limited range of character
types.)
Items you can expect to find more often (once a game or better on
average), need to avoid being so good, but rarity matters. Oh, and
there's no problem with some items being no-brainers to replace ASAP,
like clubs. I also don't see why knives should be improved; they aren't
there to be weapons, but butchery tools, and they filled that role
nicely already.
Also, as long as there are multiple good things that are mutually
exclusive, they don't become no-brainers unless you get *really* lucky.
After I've invested heavily in one weapon type, I have to consider
heavily before switching to a cool new weapon, unless all the possible
weapons in my current type suck relative to this one. Which is maybe
true for polearms, and certainly for staves, in b26 compared to the top
long blades, but short blades, maces, and axes all have their own really
nifty types that one can hold out in hopes of.
>I *am* in favor of that, really, in the sense of ensuring that to, well
>not everything, but to many more things than before, there is a season.
>That is, I'm in favor of ensuring that, unlike in b26, there are times
>when you'll really *want* to pull out something smaller than an
>executioner's axe even though you have an executioner's axe.
Isn't an executioner's axe a two-hander? If you want a shield or
punching, you need a free hand. If punching hasn't been utterly nerfed
and with shield beefed up, that's reason enough two-handers should pack
noticeable punch and be the choice when you need raw damage. Doesn't
mean it'd be a no-brainer in all circumstances.
And I have to second the notion that reducing the range of weapon stats
and adding new weapons at the same time seems pretty silly. More names
replacing real variety isn't a very good exchange. Add new weapons if
you want to make one really different. Add a two-headed flail that rolls
two attacks instead of one at larger damage and you add an interestingly
different weapon, a sort of bogusly high-speed flail. Maybe the stats do
need to be squashed together into a narrower range for balance, but it's
silly to add more weapons, distinguished only by stats, when you're
doing that. Brent tried to add other differences in 4.1; I think waiting
to add new weapons to Stone Soup until you could do so would have been
wiser.
Also, how many no-brainers are there, really? Even just among those with
multiple wins, opinions vary widely on just about everything. Brent's
number-crunching is certainly useful, but you can't account for playing
style and personal preference. I doubt adding 2 to the damage rating of
every mace-class weapon would lure Tina over to a non-butchering weapon
class.
>- a system where it's worth enchanting and carrying multiple weapons,
>not the One Weapon of Doom.
Seriously, you'll need to ease weight limits if you want to have people
do this much. It's just a pain carrying stuff right now. Steamband fails
in this goal for that very reason -- for most characters, weight is an
even more oppressive problem than in Crawl, so the need for multiple
attack types is often met even by melee classes with wand-equivalents
instead of a second weapon.
And while I love the idea of swap weapons with my designer cap on, when
I switch to my player cap, I see it is mostly a pain and I'll readily
adopt a somewhat sub-standard strategy to avoid having to worry about
switching between weapons of the same general type (melee, missile,
magic staff). It's a matter of conserving "attention space", something
you certainly do yourself. Not saying you can't make swap weapons
worthwhile, but the combat system will need to be a lot more transparent
first.
R. Dan Henry
>The problem with the numbers that look "cool!" is that once someone
>does the math and posts the results we'll all know which weapon is the
>New Unarmed Combat (tm).
Except that no weapon will really be the New Unarmed Combat unless
anyone can start using it any time. A single "best weapon" in each class
isn't a problem if it is rare enough you find yours once every several
winning games. No-brainers aren't bad as *rarities* -- chance will
create no-brainer choices often enough as it is. Obviously, the stats
can't be so good combat becomes trivial, but clearly better than the
alternatives is not a bad thing, if it comes from a *truly* lucky find.
bwross
// bwross wrote:
// > erisdiscordia wrote:
//
// > nice:). Attentive people might have noticed that in the past,
various
// > little things which have been suggested by other people and I
// > seemingly disagreed with and strongly debated have occasionally
shown
// > up in the code in some form after things get hammered out.
//
// He called me attentive! Awwww!
I've heard that it's a Scorpio thing to deceive people into thinking
your beliefs are different as a tactic to trick them into revealing
more about themselves than one does about oneself.
// (More in a while, I don't really want to answer this one in a hurry.
// Well, except to say thanks for the enlightenment. Well, and -- could
// you explain more about why you state that stretching the range from
// lowest to highest accuracy wouldn't work? And if that's in the
context
// of the b26//SS0.1 code, the 4.1 code, or both?)
It's a granularity thing... the old system just doesn't really have a
lot of granularity for playing with such things. You can stretch it,
but your not going to get any exciting results out of it without
causing potential problems. The 4.1 code gives a lot more room to play
with it, but there's still the question of how much you really want to
stretch it... it's just not a very exciting stat to twiddle and it
basically just spreads and increases the weight of the to-hit plus. In
the end, it can never be particularly interesting because all weapons
need to meet a decent standard for reliability (or be regulated to
quaint feature or monster crippler). Early on it can be interesting
but only in that it can be used to make weapons impossible to use...
and since you're looking at enchant scrolls (rare, often treasured and
saved for much later) or skill (hard to get if you can't use the weapon
you just found... you need to find a training weapon as well) that can
be overly painful. By not going to such a large range as to make
things virtually imposible, the weapon itself becomes available for
it's own training (albeit by picking and choosing your practice targets
carefully... Summon Small Furry Target works well to start for Mages
and Fighters can typically pick up a few levels of a weapon skill
without too much problem because of their stats and robustness).
Brent