Semi-permanent engraving (YANI RFC)
Janis Papanagnou
but at the same time you can overwrite existing text with a completely
unrelated text, thereby erasing the old text. This is an inconsistency.
On the other hand, if you've already engraved a word and want to fix
just the erased characters you have to rewrite the word completely.
I think it would be a nice enhancement if an erased pattern "Elbcretn"
could be rewritten by "Elbereth" in just two turns[*], because you'd
actually have to rewrite only two characters; 'c'->'e' and 'n'->'h'.
What do you think about that proposal?
Janis
[*] I.e. it requires two turns if you use e.g. a hard gem to write it;
with an athame or magic marker you'd get the usual faster writing speed.
PS: Here's some prototype code in awk, testcases for illustration below...
# Author: Janis Papanagnou, Date: 2009-12-16, Code license: GPL
BEGIN { FS = "|"
blurmap["A"] = "^-/\\"
blurmap["B"] = "IPb["
blurmap["C"] = "("
blurmap["D"] = "I|)["
blurmap["E"] = "|FL[_-"
blurmap["F"] = "|-"
blurmap["G"] = "C("
blurmap["H"] = "|-"
blurmap["I"] = "|"
blurmap["J"] = "|"
blurmap["K"] = "|<"
# ...etc.
}
function inchain (ec, pc)
{
# currently no recursive set enclosure considered
return pc == " " || index(blurmap[ec], pc)
}
function fuzzymatch_part (es, ps, ec, pc, i, turns)
{
for (i=1; i<=length(ps); i++) {
ec = substr(es, i, 1)
pc = substr(ps, i, 1)
if (ec == pc)
continue
else if (inchain(ec, pc))
turns++
else {
turns = -1 ; break
}
}
return turns
}
function fuzzymatch (e, p, le, lp, i, turns)
{
# assert: length(e) >= length(p)
le = length(e)
lp = length(p)
for (i=1; i<=le-lp+1; i++) {
if ((turns = fuzzymatch_part(substr(e,i,lp), p)) >= 0) {
printf "index: %d turns: %d \n", i, turns+le-lp
}
}
}
{ fuzzymatch($1, $2) }
# Testcases:
#
# Hello|He index: 1 turns: 3
# Hello|Hell index: 1 turns: 1
# Hello|ell index: 2 turns: 2
# Hello|e|| index: 2 turns: 4
# Hello|cll index: 2 turns: 3
# Hello|cllc index: 2 turns: 3
# Hello|cllo index: 2 turns: 2
Patashu
wrote:
> It always puzzled me that you cannot erase a semi-permanent engraving
> but at the same time you can overwrite existing text with a completely
> unrelated text, thereby erasing the old text. This is an inconsistency.
>
> On the other hand, if you've already engraved a word and want to fix
> just the erased characters you have to rewrite the word completely.
>
> I think it would be a nice enhancement if an erased pattern "Elbcretn"
> could be rewritten by "Elbereth" in just two turns[*], because you'd
> actually have to rewrite only two characters; 'c'->'e' and 'n'->'h'.
>
> What do you think about that proposal?
>
> Janis
>
> [*] I.e. it requires two turns if you use e.g. a hard gem to write it;
> with an athame or magic marker you'd get the usual faster writing speed.
>
'Elberethless' isn't a conduct yet.
Derek Ray
> On Dec 16, 3:43 pm, Janis Papanagnou <janis_papanag...@hotmail.com>
>> could be rewritten by "Elbereth" in just two turns[*], because you'd
>> actually have to rewrite only two characters; 'c'->'e' and 'n'->'h'.
>>
>> What do you think about that proposal?
>>
> 'Elberethless' isn't a conduct yet.
Illiterate.
Original proposal is broken as all hell. Elbereth is already your
one-stop-shopping solution for the vast majority of monsters; a useful
enhancement would to make it un#defined by default, or removed entirely
from the game.
(I opted for a middle ground, personally, but there's little doubt that
it's one of the biggest design flaws in Nethack.)
--
Derek
Game info and change log: http://sporkhack.com
Beta Server: telnet://sporkhack.com
IRC: irc.freenode.net, #sporkhack
Janis Papanagnou
Thanks for your opinion. Elbereth is powerful, you're right. The wands
of fire and lightning support a single-turn permanent Elbereth, one that
you cannot overwrite; once you got those wands you've reached a milestone
in your game. Then there are those semi-permanent engravings; the best
you can have is an athame, which is IMO even better than those permanent
wands because you need no charges and can spread as many engravings as
you want, also in a single turn. Wands of digging come in abundance later
in the game, not bad as well. Magic markers are often a waste to use them
for Elbereths, especially if you have any other means to write that word
reliably. So these are the most powerful sources for the E-word. (Writing
in the dust is unreliable, and you cannot really act without erasing the
written word; hardly to be considered overpowered, IMO.)
But there are also those semi-permanent engravings from blades and gems;
they require *eight* turns to write the magic word, the blades get dull,
and typically you don't encounter hard gems in the early game. In tight
situations you cannot apply blades/gems; it lasts too long to write the
word. Using any of the respective wands (markers, athames) is much better.
That's where my proposal smooths the issue. Engraving the first time with
hard gems is still long lasting. Once you're done with the basic setup the
"refresh" needs, depending on the number of characters one turn (like the
wand of digging or athame) or maybe two or three turns (which makes it also
inferior to those other means). So I think the "more powerful" is relative,
if one considers circumstances the and other options.
The proposal would certainly remove the upthread menationed inconsistency,
though.
Rethinking about it; it might be an appropriate tweak to give a "penalty"
of one additional turn (say, for figuring where to update erased letters).
> It's kind of surprising 'Elberethless' isn't a conduct yet.
It's part of the illiterate conduct, I guess, but it would probably make
sense to isolate it in a separate conduct.
Janis
Derek Ray
> Patashu wrote:
>> And make Elbereth even more powerful?
> But there are also those semi-permanent engravings from blades and gems;
> they require *eight* turns to write the magic word, the blades get dull,
> and typically you don't encounter hard gems in the early game.
You aren't looking hard enough. Test all gems you find on the ground by
writing "x" or your choice; you'll find one very quickly, especially if
you do the Mines where gem generation chances are increased.
> In tight
> situations you cannot apply blades/gems; it lasts too long to write the
> word. Using any of the respective wands (markers, athames) is much better.
Ever think there's a reason for this?
> That's where my proposal smooths the issue.
What issue?
"I only ascend 60% of random characters, the game's too hard!" is not an issue.
>> It's kind of surprising 'Elberethless' isn't a conduct yet.
> It's part of the illiterate conduct, I guess, but it would probably make
> sense to isolate it in a separate conduct.
It would make sense to remove it from the game entirely, as it removes
the vast majority of the tactical challenge.
David Damerell
>On 2009-12-16, Janis Papanagnou <janis_pa...@hotmail.com> wrote:
>>But there are also those semi-permanent engravings from blades and gems;
>>they require *eight* turns to write the magic word, the blades get dull,
>>and typically you don't encounter hard gems in the early game.
>You aren't looking hard enough. Test all gems you find on the ground by
>writing "x" or your choice; you'll find one very quickly, especially if
>you do the Mines where gem generation chances are increased.
To be specific, one in 5 gems are valuable. Of the 13 types of valuable
gem that can be found on DL 1, 2 are hard. Hence about 1 in 32 gems are
hard valuable gems. This only gets worse as citrine (soft, DL4) and
turquoise (soft, DL 7) enter the mix; it is DL10 before additional hard
gems (at DL 10,13,16,19,22,25,28) are found.
Hence... you are going to have to look at quite a lot of gems in the early
game before you find a hard one, and it is quite possible that you will
not find one at all until you get down to the Mine's End, where hard gems
are guaranteed.
--
David Damerell <dame...@chiark.greenend.org.uk>
Clown shoes. I hope that doesn't bother you.
Today is Thursday, December.
Tomorrow will be Friday, December.
Doug Freyburger
> Janis Papanagnou <janis_pa...@hotmail.com> wrote:
>
>> But there are also those semi-permanent engravings from blades and gems;
>> they require *eight* turns to write the magic word, the blades get dull,
>> and typically you don't encounter hard gems in the early game.
Eight turns - Once in trouble it's too late to engrave. I make that
mistake all too often.
> You aren't looking hard enough. Test all gems you find on the ground by
> writing "x" or your choice; you'll find one very quickly, especially if
> you do the Mines where gem generation chances are increased.
There's no need to engrave even one character. Pick up any stone but
rocks or gray stones. (Usual discussion of gray stones here). Use "E"
to engrave with any unknown type. For soft types it will telling you
you are about to write in the dust and ask you what to engrave. For
hard types it will tell you you are about to etch in the stone and ask
you what to engrave. In either case <ESC> will about without taking
another turn.
I call gems with a pattern like "red sf br 3" or "blue hr lv 16". I
have a table of what non-branch levels which valuable gems can appear
first so if I'm in the main dungeon I can often tell what the gem is
just from its color, the level and the fact that it's hard. Some colors
just knowing it's hard is enough to know.
With soft gems once I am quick I call a type with a pattern like "white
glass" then I toss it at a co-aligned unicorn. If the unicorn keeps it
I know it's valuable so I call the rest of the stack "white val lv 16"
or similar. If the unicorn rejects it then I know it's glass and I
start figuring out the other types of that color.
But knowing the identities of gem types isn't all that important. Just
knowing at least one hard type is enough in some games. Knowing which
gems are valuable and which aren't is important to a sling user but
otherwise the game eventually gives you enough gold that gems are not
needed as sales items.
The first hard gem in the game I encounter I name it "engraver" and I
keep it in my main inventory most or all of the time from then on
until I get an athame. Sometimes I name the hard gem with a pattern
liked "pr 8831" for the last turn I prayed.
Before I've found my first hard gem I like to keep a pet-checked orcish
dagger named "engraver" for use with E and #force. They are easily
replaced when I break one forcing a lock on a chest.
Doug Freyburger
>
> To be specific, one in 5 gems are valuable. Of the 13 types of valuable
> gem that can be found on DL 1, 2 are hard. Hence about 1 in 32 gems are
> hard valuable gems. This only gets worse as citrine (soft, DL4) and
> turquoise (soft, DL 7) enter the mix; it is DL10 before additional hard
> gems (at DL 10,13,16,19,22,25,28) are found.
The level restrictions do not occur in branches if I read the code
correctly. On every mines level every type of gem can appear. That's
14 soft types and 8 hard types. The spoiler claims that permissible gem
types are all equally probable so 8/171 are hard and 14/171 are soft. I
see plenty of hard gems in the mines but I have no way of telling if
it's the 1-in-21 suggested by the spoiler.
> Hence... you are going to have to look at quite a lot of gems in the early
> game before you find a hard one, and it is quite possible that you will
> not find one at all until you get down to the Mine's End, where hard gems
> are guaranteed.
Very possible but there are more gems in the mines than on main dungeon
levels. It seems common to have at least one hard gem by mines end.
Maybe a third of the games by the time I reach mines end? Enough that
it is worth trying E on every unknown type to see if its hard.
Janis Papanagnou
>
> To be specific, one in 5 gems are valuable. Of the 13 types of valuable
> gem that can be found on DL 1, 2 are hard. Hence about 1 in 32 gems are
> hard valuable gems. This only gets worse as citrine (soft, DL4) and
> turquoise (soft, DL 7) enter the mix; it is DL10 before additional hard
> gems (at DL 10,13,16,19,22,25,28) are found.
>
> Hence... you are going to have to look at quite a lot of gems in the early
> game before you find a hard one, and it is quite possible that you will
> not find one at all until you get down to the Mine's End, where hard gems
> are guaranteed.
This describes exactly my experience. The problem is that I often have to
delay visiting the mines, or the area below minetown, and depending on the
actual game entering mines end quite late. It's not uncommon that I find a
wand of digging before I get my hands on a hard gem.
I wonder whether anyone cares to dig up all the gems that are embedded in
rock on a level? I usually leave them alone (because of food/time/equipment
constraints).
Janis
Janis Papanagnou
>
> With soft gems once I am quick I call a type with a pattern like "white
> glass" then I toss it at a co-aligned unicorn. If the unicorn keeps it
> I know it's valuable so I call the rest of the stack "white val lv 16"
> or similar. If the unicorn rejects it then I know it's glass and I
> start figuring out the other types of that color.
You don't need to name them twice. If you have a stack just throw one
from the stack and watch the message - "graciously" means glass, while
"gratefully" (or "hesitatingly" if not coaligned) means a gem - then
name the rest of the stack.
> Before I've found my first hard gem I like to keep a pet-checked orcish
> dagger named "engraver" for use with E and #force. They are easily
> replaced when I break one forcing a lock on a chest.
How do you use the dagger for engraving and avoid getting it dull? With
a +0 dagger I seem to not be able to write a single Elbereth completely.
Janis
Pat Rankin
> On Dec 16, 3:43 pm, Janis Papanagnou <janis_papanag...@hotmail.com>
>
> And make Elbereth even more powerful? It's kind of surprising
> 'Elberethless' isn't a conduct yet.
It wasn't simply overlooked. During the initial implementation
of conducts (which was probably in slash'em before being added to
nethack; I don't remember anymore), it was felt that revealing
"Elbereth" in end of game disclosure would be too much of a spoiler.
Even after explicit mention of it was added to nethack's Guidebook,
it still seems more spoiler than not, and no one has come up with a
reasonable way to phrase a conduct dealing with it.
If anyone thinks they want to implement changes to engraving,
converting it into an interruptible multi-turn occupation like
digging would be more worthwhile than fiddling with partial text
modifications.
Janis Papanagnou
>> On Dec 16, 3:43 pm, Janis Papanagnou <janis_papanag...@hotmail.com>
>> [ suggestion to support rewriting just parts of a scuffed engraving ]
>
> converting it into an interruptible multi-turn occupation like
> digging would be more worthwhile than fiddling with partial text
> modifications.
Those two are completely independent tasks. I presume the multi-turn fix
is nasty to implement because it has not been done yet? (The sugestion,
OTOH, is at least just a couple lines of code, certainly more terse in C
than in the awk "prototype", and not too much "fiddling"; the assertion in
the code requires two more lines of code, and the mapping tables are there
already.)
If you think partial text replacement is not worthwhile, there's of course
other much more simple ways to fix the current inconsistency; either just
don't let semi-permanent engravings be erased by engraving completely new
text (i.e. allow just adding[*]), or allow wiping the engraving without
adding new text (similar to dust engravings, but maybe requiring a couple
turns). Currently game behaviour is just unnecessarily inconsistent. YMMV.
Anyway...
Janis
[*] The partial text refreshing is certainly more appealing, if only for
avoiding "L||ere+| E||e c+h Elbe e+| El e cth [l e t| Elbcr +" artifacts
in this case.
Derek Ray
> Pat Rankin wrote:
>>> On Dec 16, 3:43 pm, Janis Papanagnou <janis_papanag...@hotmail.com>
>>> [ suggestion to support rewriting just parts of a scuffed engraving ]
>>
>> If anyone thinks they want to implement changes to engraving,
>> converting it into an interruptible multi-turn occupation like
>> digging would be more worthwhile than fiddling with partial text
>> modifications.
>
> Those two are completely independent tasks. I presume the multi-turn fix
> is nasty to implement because it has not been done yet?
It's not pretty to implement, no. It isn't precisely _hard_ to do so,
but it involves building a bit of framework so that it's actually
_possible_ to stop, and then (like unlocking a chest or eating) resume
from the point you left off, etc. etc.
I looked into it at one point, if only to stop the unpleasant-YAAD
that comes when someone tries to engrave with a gem and discovers that
they _can't_ stop when attacked -- unlike so many other multi-turn
occupations in NH. The problem is less significant in Spork though, as
Elbereth's general weakening there means that only the one-turn methods
get used any significant amount. Which, frankly, I approve of.
Derek Ray
> Pat Rankin <ran...@pactechdata.com> writes:
>
>> nethack; I don't remember anymore), it was felt that revealing
>> "Elbereth" in end of game disclosure would be too much of a spoiler.
>> Even after explicit mention of it was added to nethack's Guidebook,
>> it still seems more spoiler than not, and no one has come up with a
>> reasonable way to phrase a conduct dealing with it.
>
> Well, maybe that would generate too many questions from people who don't
> understand the hint.
Given that it's in the Guidebook, I can't see how knowledge of the magic
word could at all be considered a spoiler. However, a suitable conduct
might be a subset of illiterate -- ie. splitting it into "You never read
anything" and "You never wrote anything", and have "illiterate" be the
shorthand reporting format for both of those.
The latter covers E nicely as well as adding an implicit "no magic marker"
conduct, which is another abuse-prone and variety-removing item.
David Damerell
>David Damerell wrote:
>>To be specific, one in 5 gems are valuable. Of the 13 types of valuable
>>gem that can be found on DL 1, 2 are hard. Hence about 1 in 32 gems are
>>hard valuable gems. This only gets worse as citrine (soft, DL4) and
>>turquoise (soft, DL 7) enter the mix; it is DL10 before additional hard
>>gems (at DL 10,13,16,19,22,25,28) are found.
>The level restrictions do not occur in branches if I read the code
>correctly.
Aha, yes. So even the shortest expedition into the Mines should yield a
hard gem, and hence if you plan on using one a quick look in the first
Mines level is well worth having.
--
David Damerell <dame...@chiark.greenend.org.uk>
If we aren't perfectly synchronised this corncob will explode!
Today is Friday, December.
Tomorrow will be Saturday, December - a weekend.
Pasi Kallinen
>
> If anyone thinks they want to implement changes to engraving,
> converting it into an interruptible multi-turn occupation like
> digging would be more worthwhile than fiddling with partial text
> modifications.
...or else the next version won't have either change?
--
Pasi Kallinen
pa...@alt.org
http://bilious.homelinux.org/ -- NetHack Patch Database
TJR
as they like, without making it any harder /easier for the rest.
Please don't remove Elbereth entirely. That would be a steep increase.
I like the idea of splitting illiterate into several sub-conducts: You
never wrote / read anything are fine. Separating off Elbereth would be
even better.
Spoiling too much in the conducts screen for Elberethless-conduct: How
about adding a message "You engraved Elbereth" after the conduct has
been broken? While inconsistent with the other conducts, this approach
definitely doesn't give too much away.
Probability of hard gems: You can engrave with rings, too. In my
games, I often get an "engraver" ring from Sokoban first. However,
engraving in the dust gives faster results, so over time I have
stopped using hard gems/rings routinely altogether.
Derek Ray
> The nice part about conducts is everybody can play as difficult a game
> as they like, without making it any harder /easier for the rest.
The bad part about conducts is that asking players to tie one hand
behind their back just to get some challenge is far worse design than
simply making the game reasonably well-balanced in the first place.
This allows everyone to measure their progress against each other on the
same scale (trying to win the game), rather than having some people
measure on the scale of "can i ascend?" and others measure on the scale
of "how many times in a row can i ascend while standing on one leg and
farting the 'Hallelujah Chorus'?"
Janis Papanagnou
>
> Spoiling too much in the conducts screen for Elberethless-conduct: How
> about adding a message "You engraved Elbereth" after the conduct has
> been broken?
"You engraved Elbereth (3192 times)"
;-)
> Probability of hard gems: You can engrave with rings, too.
Specific rings, to be more precise.
> In my
> games, I often get an "engraver" ring from Sokoban first. However,
> engraving in the dust gives faster results, so over time I have
> stopped using hard gems/rings routinely altogether.
Hard gems are still helpful, e.g. to secure stashes, but also in other
application contexts.
Janis
Doug Freyburger
> Quoting Doug Freyburger <dfre...@yahoo.com>:
>>David Damerell wrote:
>
>>>To be specific, one in 5 gems are valuable. Of the 13 types of valuable
>>>gem that can be found on DL 1, 2 are hard. Hence about 1 in 32 gems are
>>>hard valuable gems. This only gets worse as citrine (soft, DL4) and
>>>turquoise (soft, DL 7) enter the mix; it is DL10 before additional hard
>>>gems (at DL 10,13,16,19,22,25,28) are found.
>
>>The level restrictions do not occur in branches if I read the code
>>correctly.
Leading to 1-in-21 gems encountered in branch levels being hard rather
than 1-in-32.
> Aha, yes. So even the shortest expedition into the Mines should yield a
> hard gem, and hence if you plan on using one a quick look in the first
> Mines level is well worth having.
Chortle. I've had plenty of characters who failed to get a hard gem
before mines end. Mines end is worth doing for either the luckstone or
the guaranteed hard gems. Finding both along the way may make it worth
abandoning the rest of the mines depending on how much you want the
other items in the mines. I don't ever remember a game where I got
enough candles without the lighting shop but there's always the Castle
wand and a wish.
Janis Papanagnou
> David Damerell wrote:
>
>> Aha, yes. So even the shortest expedition into the Mines should yield a
>> hard gem, and hence if you plan on using one a quick look in the first
>> Mines level is well worth having.
>
> Chortle. I've had plenty of characters who failed to get a hard gem
> before mines end.
Ditto. Being curious I've started a wizmode game and poly'd into a xorn
to get all the gems, even the embedded ones. This was just a single test
game, by far no significant data set. But, first, most of the gems on
the levels had been embedded in the surrounding stone, just a few lying
around. I found 6-7 gems per level, one soft on the first, and a diamond
on the second. I suppose object detection and a pickaxe are prerequisites
to obtain a somewhat sufficient amount of gem candidates.
BTW, I seem to have encountered a bug. Some gems were below surface, so I
had to dig them up. But the xorn was unable to pick things up...
"You fall into your pit!"
"You cannot reach the bottom of the pit."
Janis
Doug Freyburger
> Doug Freyburger wrote:
>
>> Chortle. I've had plenty of characters who failed to get a hard gem
>> before mines end.
>
> Ditto. Being curious I've started a wizmode game and poly'd into a xorn
> to get all the gems, even the embedded ones. This was just a single test
> game, by far no significant data set.
The closest I can get to data is trying to recall how many of my
characters have had a hard gem before they reach mines end. I can
recall a lot of them but it seems under half of all of my characters who
got to mines end.
Pat Rankin
wrote:
> BTW, I seem to have encountered a bug. Some gems were below surface, so I
> had to dig them up. But the xorn was unable to pick things up...
>
> "You fall into your pit!"
> "You cannot reach the bottom of the pit."
Sounds an awful lot like C343-16 to me....
Pat Rankin
> Spoiling too much in the conducts screen for Elberethless-conduct: How
> about adding a message "You engraved Elbereth" after the conduct has
> been broken? While inconsistent with the other conducts, this approach
> definitely doesn't give too much away.
That's reasonable once the hypothetical "didn't engrave
Elbereth" conduct has been broken, but it doesn't work prior
to that. Having no feedback when you've actually gone to
the effort to maintain the conduct would be pretty strange.
Unfortunately deliberately adhering to the conduct and being
ignorant of the possibility of breaking it produce the same
outcome: You never engraved Elbereth.
I was the one who added reference to engraving Elbereth
to nethack's Guidebook, specifically because it gave spoiled
players who use that such a huge advantage over those who
were unaware. Most spoilers require more effort or luck
before they can be taken advantage of, but this one can apply
at turn 1 (although resorting to it so early seems ludicrous).
The Guidebook entry isn't intended to encourage more
players to rely on that engraving, just to even the playing
field a bit. Evidently most players don't bother reading the
Guidebook, but they can only blame themselves if they later
discover that they've been missing out on something that
could have kept many dead characters alive a little longer.
Janis Papanagnou
Definitely. Since this part was just a note aside I haven't checked - sorry.
Janis
Janis Papanagnou
> TJR wrote:
>>
>> [...] How
>> about adding a message "You engraved Elbereth" after the conduct has
>> been broken?
>
> "You engraved Elbereth (3192 times)"
"You engraved Elbereth (3192 times, 1511 times successfully)" [*]
Just occurred to me. :-}
More seriously; in one way, a simple count of the number of engraving
attempts seems to be already sufficiently covered by the binary-valued
illiterate conduct, but then there's no diffenenciation between the full
illiterate conduct and having just abstained from applying the E-word.
Counting the actual successful "saved by the word" events might be more
appropriate. But in this case I'm not sure whether it's to be considered
a conduct any more.
Janis
[*] And this would have to consider engraving "Elbe" and "reth" in two
moves. And some more details would have to be considered if adding to
an engraving, while erasures and engraving failures may appear, and to
not count existing Elbereths on the current square twice. I sense a can
of worms...
Derek Ray
> I was the one who added reference to engraving Elbereth
> to nethack's Guidebook, specifically because it gave spoiled
> players who use that such a huge advantage over those who
> were unaware. Most spoilers require more effort or luck
> before they can be taken advantage of, but this one can apply
> at turn 1 (although resorting to it so early seems ludicrous).
The problem is that while it seems ludicrous, it's also one of the most
effective strategies for survival in the early- and mid-game: rather
than risk your character trying to melee while he doesn't have good DEX
or enough level to actually hit, engrave E in the dust a few times and
use whatever missile weapons are handy to soften things up; even rocks
do fine here.
It's also consistently one of the best strategies for taking care of
mildly out-of-depth monsters such as foxes on DL1, soldier ants and
leocrotta later, and really any case where something hits you a bit
harder than you were expecting, or where your melee combat is still too
random to rely on consistently. Combined with reflection (and the 50%
availability of the same in Sokoban), and used early on in fights rather
than as a last-ditch effort, it can make a character nearly invincible.
(Looking at the code, I do see that someone realizes this already: the
fact that rings of levitation can never be started with for a Wizard if
Elbereth is #defined is obviously not accidental.)
The difference between using it early and late is pretty simple to
quantify, too:
-- Chance of one dust-E failing is approximately 30%
-- Chance of two in a row failing reduces to 9%
-- Chance of three in a row failing is 2.7%
Combine this with its utility in blocking corridors to provide a retreat
mechanism from monsters you simply aren't prepared to handle (engrave E,
drop a gold piece, step behind it) and it becomes a tool that nobody
playing to win (as opposed to following conducts) fails to use.
Anyway, I know this has all been covered before. I personally find that
my fix in Spork works reasonably well for it (people still use it to
deter monsters, but it's not the 100% guaranteed-safety of vanilla) as
far as leaving it both available and less of a huge power-tool.
> The Guidebook entry isn't intended to encourage more
> players to rely on that engraving, just to even the playing
> field a bit.
I'm not sure there are a significant quantity of unspoiled players still
playing the game... at all. The #nethack channel on freenode routinely
warns people entering that it's not a spoiler-free zone, and also
routinely advises them to go ahead and spoil themselves, as there's just
too much information to parse-by-playing -- while technically the
difference between spoiler-backed ascension rates and non-spoilered
ascensions is immeasurable (no recorded instances of the latter), we've
seen people go from "noob" to "winner" in only a couple months using
spoilers, and even the most generous reading of an "unspoiled" player's
skills wouldn't put that kind of learning rate as possible.
David Damerell
>David Damerell wrote:
>>Quoting Doug Freyburger <dfre...@yahoo.com>:
>>>David Damerell wrote:
>>>>To be specific, one in 5 gems are valuable.
>>>correctly.
>Leading to 1-in-21 gems encountered in branch levels being hard rather
>than 1-in-32.
Not sure where you get that from, to be honest. Valuable gems are 9
hard:14 soft, and 1/5 * 9/14 is about 1/8.
--
David Damerell <dame...@chiark.greenend.org.uk> Kill the tomato!
Today is Saturday, December - a weekend.
Tomorrow will be Sunday, December - a weekend.
Erwin M.
> Janis Papanagnou wrote:
>> TJR wrote:
>>>
>>> [...] How
>>> about adding a message "You engraved Elbereth" after the conduct has
>>> been broken?
>>
>> "You engraved Elbereth (3192 times)"
>
> "You engraved Elbereth (3192 times, 1511 times successfully)" [*]
>
> Just occurred to me. :-}
>
> More seriously; in one way, a simple count of the number of engraving
> attempts seems to be already sufficiently covered by the binary-valued
> illiterate conduct, but then there's no diffenenciation between the full
> illiterate conduct and having just abstained from applying the E-word.
> Counting the actual successful "saved by the word" events might be more
> appropriate. But in this case I'm not sure whether it's to be considered
> a conduct any more.
"If the E-word is on a square and no monster is repelled by it, does it
make a difference?" Any more than weaponless conduct isn't broken if you
swing and miss.
Link
wrote:
> How do you use the dagger for engraving and avoid getting it dull? With
> a +0 dagger I seem to not be able to write a single Elbereth completely.
Quoted from elbereth_faq.html: "if you need to write Elbereth with
just one +0 non-athame, first engrave "Elb" and then add "ereth" to
the engraving."
Derek Ray
> Janis Papanagnou <janis_pa...@hotmail.com> writes:
>> "You engraved Elbereth (3192 times, 1511 times successfully)" [*]
>> Counting the actual successful "saved by the word" events might be more
>> appropriate. But in this case I'm not sure whether it's to be considered
>
> without Elbereth or not.
> The player may have an idea, but I don't think there's a way to
> programmaticaly be sure about it.
Well, you could track it a number of ways, it's just that none of them
are particularly reflective of what Janis wants to know.
You could track how many times a monster was repelled by any engraving,
but that leads to miscounts from the Soko zoo, the player's stash, and
times when you block corridors with a gold piece + E. It also doesn't
really reflect how many times the player was 'saved' per se, as a single
burned-in E may get a great deal of use, even at times when the player
isn't actually at threat. (Magicbane is useful to clear monsters out of
your way that you don't want to waste time fighting, for example.)
You can special-case out the first and third of those issues (ignore
Soko and only count cases where the player was standing on the square at
the time), but you still aren't getting an accurate count of actual
saves as opposed to number of critters repelled.
You can then build in code to track only the first monster any given "E"
square repels, at which point you've crossed my personal "waste of time
for benefit gained" threshold, and I'll file it in the "too-clever ideas
that people talk about instead of coding" bucket. :) (Already in the
bucket: cooking, player dismemberment, selfpolied player being able to
wield multiple weapon (marilith) or wear two helmets (ettin), etc.)
<plug>
(I preferred to address the real issue, which is that when Elbereth
isn't too powerful, it doesn't need to be a conduct anymore. ;)
</plug>
Doug Freyburger
> Quoting Doug Freyburger <dfre...@yahoo.com>:
>
>>Leading to 1-in-21 gems encountered in branch levels being hard rather
>>than 1-in-32.
>
> Not sure where you get that from, to be honest. Valuable gems are 9
> hard:14 soft, and 1/5 * 9/14 is about 1/8.
Thanks for the correction. Looks like I got lost in the arithmatic.
Pat Rankin
wrote:
> <plug>
> (I preferred to address the real issue, which is that when Elbereth
> isn't too powerful, it doesn't need to be a conduct anymore. ;)
> </plug>
Even when it's been nerfed, it's still a viable optional
conduct for players to use or deliberately abstain from using,
and there are competent, successful players who do abstain.
Somewhere along the line this discussion seems to have
mutated from "player did[not] engrave Elbereth" to "player
did[not] use an engraving of Elbereth for protection". Those
are two different things, since "Elbereth" is on the list of
random engravings potentially placed during level creation,
is deliberately placed on a couple of special levels to reduce
the chance of a monster intercepting some treasure (as noted
for Sokoban), and also could have been engraved by a previous
player on what is now a bones level in current player's game.
Tracking whether someone created the engraving is far
simpler, even if the writing is done in pieces rather than
all at once, than tracking whether they derived any benefit
from such an engraving. And conduct-wise, I think the simpler
form is the not just the only viable one to actually track but
also the only one of the two that makes any sense to attempt
to track. Engraving the magic word is an attempt _by the
player_ to gain protection even if the RNG decides that no
monster will ever be affected by that particular engraving.
Janis Papanagnou
>> Counting the actual successful "saved by the word" events might be more
>> appropriate. But in this case I'm not sure whether it's to be considered
>
> without Elbereth or not.
No, by "saved by the word" I didn't mean that we should determine whether
he would have survived, rather whether the monsters got scared and fled,
and did not attack the player because of the E-word, in at least one turn.
Sorry if I have been unclear in my wording.
Janis
Derek Ray
> On Dec 18, 9:33 am, Derek Ray <de...@moot.its.only.a.spamtrap.org>
> wrote:
>> <plug>
>> (I preferred to address the real issue, which is that when Elbereth
>> isn't too powerful, it doesn't need to be a conduct anymore. ;)
>> </plug>
>
> Even when it's been nerfed, it's still a viable optional
> conduct for players to use or deliberately abstain from using,
> and there are competent, successful players who do abstain.
Uh huh. Name three, and how long their longest ascension streak is.
However, this is beside the original point, which is that requiring the
player to avoid using all the tools available to him in order to be
challenged isn't really a good thing. Hence the above change in Spork,
bringing Elbereth back in line as a "useful tool" rather than
"all-purpose solution".
fire monkey
wrote:
> On 2009-12-19, Pat Rankin <ran...@pactechdata.com> wrote:
>
> > On Dec 18, 9:33 am, Derek Ray <de...@moot.its.only.a.spamtrap.org>
> > wrote:
> >> <plug>
> >> (I preferred to address the real issue, which is that when Elbereth
> >> isn't too powerful, it doesn't need to be a conduct anymore. ;)
> >> </plug>
>
> > Even when it's been nerfed, it's still a viable optional
> > conduct for players to use or deliberately abstain from using,
> > and there are competent, successful players who do abstain.
>
> Uh huh. Name three, and how long their longest ascension streak is.
>
cowardly, and a failed dust-E is a pathetic way to die. So I've
avoided it for a long time, except for tactical late-game (liches,
Demogorgon, farming, stashes) and often gone completely Elberethless.
Hence my ratio sucks. 15-20% on 60 NAO ascensions, with the longest
streak being 3, as part of a 5 for 6. I regularly ascend in the high
20Ks to low 30Ks with a high speed of 21318. I don't care about
conducts either, so I'm not an elite player in any sense. Ironically,
that made my transition to Spork easier than some exceptional players
who relied on E. Its still available as a late game tool, so no big
deal. The god-gifted armor and racial armor bonuses balance the
defense well enough.
- firemonkey
Derek Ray
> On Dec 19, 12:14 pm, Derek Ray <de...@moot.its.only.a.spamtrap.org>
>> > On Dec 18, 9:33 am, Derek Ray <de...@moot.its.only.a.spamtrap.org>
>> >> (I preferred to address the real issue, which is that when Elbereth
>> >> isn't too powerful, it doesn't need to be a conduct anymore. ;)
>> >> </plug>
>> > Even when it's been nerfed, it's still a viable optional
>> > conduct for players to use or deliberately abstain from using,
>> > and there are competent, successful players who do abstain.
>> Uh huh. Name three, and how long their longest ascension streak is.
> I'll offer myself as an example. I have always found E to be
> cowardly, and a failed dust-E is a pathetic way to die. So I've
> avoided it for a long time, except for tactical late-game (liches,
> Demogorgon, farming, stashes)
I have to point out that this is not precisely an example of
"abstention", but more "selective usage".
> and often gone completely Elberethless.
This, of course, is such an example.
> Hence my ratio sucks. 15-20% on 60 NAO ascensions, with the longest
> streak being 3, as part of a 5 for 6.
And this is, again, really my point. It's possible to ignore E and not
use it. But making it into a conduct rather than fixing it is simply
the wrong approach to the problem, which is that players are literally
required to not use all the available tools in order to challenge
themselves -- and it is patently obvious that passing up E is far from
the optimal method of play (which you note yourself).
This kind of design bothers me. Using all the tools available shouldn't
lead directly to a playstyle that cuts out a large portion of the game's
content.
> who relied on E. Its still available as a late game tool, so no big
> deal. The god-gifted armor and racial armor bonuses balance the
> defense well enough.
I'd also note that the enhancements to ranged weaponry help here to a
degree as well; since a sling is actually more worthwhile than a thrown
rock in Spork, it's easier to off a gnome and pick up a basic ranged
weapon with relatively-infinite ammo.
</replug>
Kent Paul Dolan
> I'd also note that the enhancements to ranged
> weaponry help here to a degree as well; since a
> sling is actually more worthwhile than a thrown
> rock in Spork, it's easier to off a gnome and pick
> up a basic ranged weapon with relatively-infinite
> ammo.
It makes exactly no sense that digging with a
pickaxe through solid rock doesn't leave rubble
behind. Why not have it leave rocks, making that
"relatively-infinite ammo" even more readily
available? 3d10 per cell excavated should be plenty,
so that if all one is doing is digging to make
more rocks for ammo, it doesn't have to get tedious.
xanthian.
David Damerell
>On 2009-12-19, Pat Rankin <ran...@pactechdata.com> wrote:
>>and there are competent, successful players who do abstain.
>Uh huh. Name three, and how long their longest ascension streak is.
Well, "competent, successful" means something different to what it did
when I was playing vanilla regularly, but I've never used Elbereth and
IIRC I was 3 from 7 in the 2004 Tournament.
--
David Damerell <dame...@chiark.greenend.org.uk> Kill the tomato!
Today is First Gouday, December.
Tomorrow will be First Chedday, December - a public holiday.
TJR
> >> appropriate. But in this case I'm not sure whether it's to be considered
> Well, you could track it a number of ways, it's just that none of them
> are particularly reflective of what Janis wants to know.
>
> You could track how many times a monster was repelled by any engraving,
> but that leads to miscounts from the Soko zoo, the player's stash, and
> times when you block corridors with a gold piece + E. It also doesn't
> really reflect how many times the player was 'saved' per se, as a single
> burned-in E may get a great deal of use, even at times when the player
> isn't actually at threat. (Magicbane is useful to clear monsters out of
> your way that you don't want to waste time fighting, for example.)
>
> You can special-case out the first and third of those issues (ignore
> Soko and only count cases where the player was standing on the square at
> the time), but you still aren't getting an accurate count of actual
> saves as opposed to number of critters repelled.
[snip]
Well, I'd definitely like to know how many times my acts scared a
monster. This proposed conduct would include all Elbereth on squares I
engraved upon, all musical instruments, all read scrolls of scare
monster, Magicbane, but not monsters retreating due to low health. I'm
not sure about dropped scare scrolls. This "never scared a monster"
type conduct would be split off from illiterate.
Of course, the developers' opinions are what counts, and I'm not the
one doing the work.
> <plug>
> (I preferred to address the real issue, which is that when Elbereth
> isn't too powerful, it doesn't need to be a conduct anymore. ;)
> </plug>
That, of course, is also a solution to the issue at hand: The
illliterate conduct lumps too many things together, so in the end it
is quite a steep increase in difficulty.
I think the underlying assumption of Derek's brilliant rhethoric ("how
many times in a row can i ascend while standing on one leg and
farting the 'Hallelujah Chorus'?" ) is that conducts always have to be
an ugly, bolted-on way of introducing a difficulty scale into a game.
I imagine they could be evolved into a quite natural mechanism.
Probably, this has much to do with communication, with the interface
forming player expectations, and the more interface-related conducts
(e.g. pacifist) would have to be modified.
Janis Papanagnou
>> [...]
>
> That, of course, is also a solution to the issue at hand: The
> illliterate conduct lumps too many things together, so in the end it
> is quite a steep increase in difficulty.
>
> I think the underlying assumption of Derek's brilliant rhethoric ("how
> many times in a row can i ascend while standing on one leg and
> farting the 'Hallelujah Chorus'?" ) is that conducts always have to be
> an ugly, bolted-on way of introducing a difficulty scale into a game.
Hmm.. - "bolted on"? - Trying to imagine concepts that are not bolted-on
in a (highly artificial) game like Nethack. :-)
I don't play conducts (I think), but conducts give that opportunuity of
introducing difficulty, as you say, and in a way that people can choose
their level. It would certainly be a bad situation if Nethack would be
an easy game and all its difficulty would stem primarily from conducts,
but as an add-on it looks like a reasonable approach. In what other ways
could game difficulty be kept a variable factor over a wide scale so that
a large community of players would still enjoy the game? Optional branches
(with or without goodies?), maybe? What else?
> I imagine they could be evolved into a quite natural mechanism.
It would be interesting to hear more details about your thoughts.
> Probably, this has much to do with communication, with the interface
> forming player expectations, and the more interface-related conducts
> (e.g. pacifist) would have to be modified.
[kept for context]
Janis
Ray Kulhanek
> On Dec 16, 1:37 am, Patashu <patas...@gmail.com> wrote:
>> On Dec 16, 3:43 pm, Janis Papanagnou <janis_papanag...@hotmail.com>
>> [ suggestion to support rewriting just parts of a scuffed engraving ]
>>
>> And make Elbereth even more powerful? It's kind of surprising
>> 'Elberethless' isn't a conduct yet.
>
> It wasn't simply overlooked. During the initial implementation
> of conducts (which was probably in slash'em before being added to
> nethack; I don't remember anymore), it was felt that revealing
> "Elbereth" in end of game disclosure would be too much of a spoiler.
> Even after explicit mention of it was added to nethack's Guidebook,
> it still seems more spoiler than not, and no one has come up with a
> reasonable way to phrase a conduct dealing with it.
http://bilious.homelinux.org/?368
How about something like this? This patch keeps track of how many
times you've engraved Elbereth, but doesn't display it in #conducts
until the count is at least one (except if in wizard mode).
That way, a player can't learn about Elbereth by checking #conducts,
but does get a running count once he's already engraved it at least
once. As it stands, a player could still be spoiled by reading a
YAAP with the conduct in it, but reading those in the first place
basically guarantees a few spoilers.
I would have liked to display it at the end of game in the
event of an ascension even if the count is zero, but the int final
parameter doesn't distinguish between ascended and escaped, so I
didn't. I could probably make a modified version that does have final
distinguish between types of survivals if there's interest in that.
More details about exactly when the count is incremented are included
in the patch description.
James Of Tucson
> How about something like this? This patch keeps track of how many
> times you've engraved Elbereth, but doesn't display it in #conducts
> until the count is at least one (except if in wizard mode).
> That way, a player can't learn about Elbereth by checking #conducts,
> but does get a running count once he's already engraved it at least
> once. As it stands, a player could still be spoiled by reading a
> YAAP with the conduct in it, but reading those in the first place
> basically guarantees a few spoilers.
If it's in the Guidebook, it isn't a Spoiler.
Ray Kulhanek
> If it's in the Guidebook, it isn't a Spoiler.
I would have thought so too, but in the post I quoted, Pat Rankin said
"Even after explicit mention of it was added to nethack's Guidebook,
it still seems more spoiler than not, and no one has come up with a
reasonable way to phrase a conduct dealing with it." So I made a
point of phrasing the conduct in such a way that it didn't give
anything away.
James Of Tucson
There are some who would argue that a separate challenge for Elbereth
would nerf illiterate pacifist.
Kent Paul Dolan
> How about something like this? This patch keeps
> track of how many times you've engraved Elbereth,
> but doesn't display it in #conducts until the
> count is at least one (except if in wizard mode).
> That way, a player can't learn about Elbereth by
> checking #conducts, but does get a running count
> once he's already engraved it at least once. As
> it stands, a player could still be spoiled by
> reading a YAAP with the conduct in it, but reading
> those in the first place basically guarantees a
> few spoilers.
> I would have liked to display it at the end of
> game in the event of an ascension even if the
> count is zero, but the int final parameter doesn't
> distinguish between ascended and escaped, so I
> didn't. I could probably make a modified version
> that does have final distinguish between types of
> survivals if there's interest in that. More
> details about exactly when the count is
> incremented are included in the patch description.
One doesn't have to be too cutesy. Just saying in
the ascension details: "An engraving-related conduct
(undisclosed to avoid spoiling an unspoiled player)
was achieved".
"You know what it was if you did it deliberately".
There could indeed be several such conducts achieved
but hinted at rather than disclosed at game end.
Also, finding ways to subdivide existing conducts
for scoring purposes might go a long way toward
replacing NetHack's current broken scoring system.
Geno'd nothing gets a max score bump, Geno'd six or
more species gets no score bump, values in between
cover a range of scores less than max and greater
than zero.
Similar gradients of score components could exist
for stuff like speed of ascension (which turns out
to have really bizarre low values, use of a log
function might be appropriate as part of computing
that score component), numbers of species extincted
(and a big bump for extinguishing all of them except
your own), et cetera.
Of course, we chewed all this to death and back
again a few years ago, with no fruitful result.
FWIW
xanthian.
TJR
> Also, finding ways to subdivide existing conducts
> for scoring purposes might go a long way toward
> replacing NetHack's current broken scoring system.
> Geno'd nothing gets a max score bump, Geno'd six or
> more species gets no score bump, values in between
> cover a range of scores less than max and greater
> than zero.
>
> Similar gradients of score components could exist
> for stuff like speed of ascension (which turns out
> to have really bizarre low values, use of a log
> function might be appropriate as part of computing
> that score component), numbers of species extincted
> (and a big bump for extinguishing all of them except
> your own), et cetera.
I have been musing a similar idea in the past: an ascendability index
instead of score. Banks collect certain data about their customers and
then use statistics to find out how likely they are to repay their
loans, and then given them according conditions. This system seems to
work quite well (unless it isn't used, see credit crunch). Given the
huge amounts of data available from NAO, it should be theoretically
possible to define an ascendability index in a similar manner. One
would first define a huge set of all concievable questions (e. g. max
HP, has MR by turn x, AC, knows identities of key items, ever split a
pudding, split >1000 puddings, controls Luck while praying, typically
injured x% of maxHP, same of Elbereth-resistant monsters, carries
known escape item, ...). Next, one would take all games alive at turn
t and fit the unknown "did later ascend" to the known answers to these
questions. (Probably, only a few of these variables will be relevant,
so it should probably not be least-squares.) The result is an estimate
of ascension chance.
For those games that did not end in ascension, the predicted ascension
chance would replace the current score. For ascensions, all followed
conducts would give bonus points. The size of the bonus should be
related to how much more difficult the combination of conducts made it.
Ray Kulhanek
> I have been musing a similar idea in the past: an ascendability index
> instead of score. Banks collect certain data about their customers and
> then use statistics to find out how likely they are to repay their
> loans, and then given them according conditions. This system seems to
> work quite well (unless it isn't used, see credit crunch). Given the
> huge amounts of data available from NAO, it should be theoretically
> possible to define an ascendability index in a similar manner. One
> would first define a huge set of all concievable questions (e. g. max
> HP, has MR by turn x, AC, knows identities of key items, ever split a
> pudding, split >1000 puddings, controls Luck while praying, typically
> injured x% of maxHP, same of Elbereth-resistant monsters, carries
> known escape item, ...). Next, one would take all games alive at turn
> t and fit the unknown "did later ascend" to the known answers to these
> questions. (Probably, only a few of these variables will be relevant,
> so it should probably not be least-squares.) The result is an estimate
> of ascension chance.
>
> For those games that did not end in ascension, the predicted ascension
> chance would replace the current score. For ascensions, all followed
> conducts would give bonus points. The size of the bonus should be
> related to how much more difficult the combination of conducts made it.
Personally, I think this is overcomplicated, but if it were to be
done, here are my suggestions.
Instead of using parameters like "has MR by turn x", I'd suggest just
making it "has MR" and have the turn be an independent parameter.
There would be a strong positive correlation between "turn" and
ascension probability (likewise with "max dlvl reached" and ascension
probability), but a negative correlation if various other parameters
are still 0 when turn and max dlvl are high, and of course increases
due to turn would level out eventually as it starts to get into
extinctionist game length. The exact optimal turn to have MR on isn't
something any of us know; we just know to get it as soon as possible.
Let the system learn the approximate correlation by itself.
You'd also want the final score to be some sort of aggregate of the
score on all prior turns or the last n turns. Otherwise, the game
would see "at 2% of max HP, can't pray, out of escape items, and
Demogorgon just put on an AoLS" and decide that the probability of
ascending is 0.3% no matter how well the player had been doing
previously.
But of the various scoring proposals I've seen, I still like
the other Ray's suggestion from this thread.
http://groups.google.com/group/rec.games.roguelike.nethack/browse_thread/thread/76ce9bf0ec4d4a9e/e6e03c7017473236?tvc=2&#e6e03c7017473236
Ray
> I have been musing a similar idea in the past: .... Given the
> huge amounts of data available from NAO, it should be theoretically
> possible to define an ascendability index.... One
> would first define a huge set of all concievable questions (e. g. max
> HP, has MR by turn x, AC, knows identities of key items ...).
> Next, one would take all games alive at turn
> t and fit the unknown "did later ascend" to the known answers to these
> questions. (Probably, only a few of these variables will be relevant,
> so it should probably not be least-squares.) The result is an estimate
> of ascension chance.
Way too complicated. Just use a distance measure. In order to win,
a character has to go do their quest, come back to the branch level,
go down to the amulet level, get the amulet, come back up, and
traverse the planes. Just track the furthest point along that journey
so far reached and award "completion points" according to the
percentage of completion.
I think completion points, rather than experience, should dominate
scores.
Bear
TJR
> > huge amounts of data available from NAO, it should be theoretically
> > possible to define an ascendability index.... One
> Way too complicated. Just use a distance measure. In order to win,
> a character has to go do their quest, come back to the branch level,
> go down to the amulet level, get the amulet, come back up, and
> traverse the planes. Just track the furthest point along that journey
> so far reached and award "completion points" according to the
> percentage of completion.
>
> I think completion points, rather than experience, should dominate
> scores.
My idea amounts to letting statistics determine what the important
milestones are you want to award completion points for. The current
score system is broken because it is unrelated to ascending, and any
human is bound to be off to smaller or larger degree.
Patashu
One problem I saw with the setup is that it would be biased by the
fact that, for instance, masters of nethack play with conducts more
often then newbies do. Imagine if there was an extremely hard conduct,
conduct X, that master Y plays with and he ascends with it every time.
How is the system to know that the conduct is actually really hard?
How do you unbias the sample?
sid vicious
>
--
sid
... i don't fuckin' care -- sid
Ssulian
The problem with this idea is that it would affect the
realisticness of the game. If it were always a multiple
of 3 rocks that dropped in a rubble square, players
would wonder at the strange reason for the value
of the number 3, and even the designers might be
called into question as some kind of numerologists.
My solution is to drop 3d10 + d3 rocks per rubble
square, so that there is no clear factor to the number
of rocks dropped.
\
And yes, it is possible to create a d3 die in physical
form.
Ssulian
> Also, finding ways to subdivide existing conducts
> for scoring purposes might go a long way toward
> replacing NetHack's current broken scoring system.
> Geno'd nothing gets a max score bump, Geno'd six or
> more species gets no score bump, values in between
> cover a range of scores less than max and greater
> than zero.
In my opinion, players should get higher scores for the more
species genocided. Frankly, a non-broken scoring system
would reward a straight 1000 points for each genocide, and
no other action in the game would generate points.
--
S
Ssulian
> For those games that did not end in ascension, the predicted ascension
> chance would replace the current score. For ascensions, all followed
> conducts would give bonus points. The size of the bonus should be
> related to how much more difficult the combination of conducts made it.
As I see it, if the game did not end in an
ascension then the predicted ascension
chance should be 0 %.
\
Ssulian
So I guess according to this, everyone
who dies in the Gnomish mines gets a
score of zero.
\
news.telesweet.net
3d10 means roll three 10-sided dice. It doesn't mean roll
one 10-sided die and multiply by 3.
--
Dan G
Ssulian
Oh, good on you for the correction.
--
S
Doug Freyburger
>
> The problem with this idea is that it would affect the
Digging monsters leave behind boulders and rocks. Player monsters do
not. Digging monsters dig twice as fast as player monsters.
I think it was a deliberate decision by the DevTeam that doing your own
digging should not be a way to generate objects. This might be a matter
of playing style or balance, a matter of not wanting to be able to
generate infinite missiles, or not wanting to be able to generate
infinite items because of the memory and disk costs.