A request for Vladimir
Mikko Sorsa
Hello, Vladimir
I'd like to ask for you to practice your 1337 hax0ring skills
and look at the item generation code for the dungeon levels and
especially see what things affect it and whether there are for
example race/profession/level specific considerations in there (thinking
of wizards/spellbooks here) etc. Please remember to add spoiler
space before the answer if necessary as this kind of knowledge
might lead to some serious scumming and I imagine some people
here might want to avoid that ; ).
Thanks,
Mikko
Vladimir Panteleev
@ihatespam.welho.com> says...
>
>
> Hello, Vladimir
>
> I'd like to ask for you to practice your 1337 hax0ring skills
heh >:]
> and look at the item generation code for the dungeon levels and
> especially see what things affect it and whether there are for
> example race/profession/level specific considerations in there (thinking
> of wizards/spellbooks here) etc. Please remember to add spoiler
> space before the answer if necessary as this kind of knowledge
> might lead to some serious scumming and I imagine some people
> here might want to avoid that ; ).
Who do you take me for? When has it happened that I forgot to add
spoiler space?
.
.
.
.
...
.
.
.
.
..
.
.
.
.
...
.
.
.
.
Ok. It's well known that every dungeon level has its own danger value.
This danger value can be modified by scrolls of peace/danger. In the ID,
the danger value is equal to the current depth (level nr).
Each item type has its own "creation probability" value, as well as a
"quality" value. The first represents the overall odds for an item to be
generated.
When choosing an item type, ADOM first calculates the maximum item
quality allowed for this level. This is done so:
MaxItemQuality=Max(1,Min(50,LevelDangerValue))
If the PC is in the ID, the maximum item quality is reduced to one
third:
MaxItemQuality=Max(1,MaxItemQuality/3)
This means that the maximum item quality level is reached at ID:16.
Items whose quality value is higher than the level's maximum item
quality can't be generated. So you won't be able to find anything cool
on ID:1 no matter how much you try (AoLS's danger level is 2, so is
potion of water's).
When choosing an item, ADOM(*) adds up the "creation probabilities" of
items whose quality value is not above the level's MaxItemQuality. Then,
it selects a random value from 1 to that number and sees which item's
creation probability was added at that spot.
Ugh...sounds confusing. Even I got confused. Here, let me picture it.
Suppose we have just a few items in the game: a dagger, a pretty good
sword and a very rare artifact.
ADOM piles their creation probabilities first, and remembers which
item's quality value was added:
| dagger | sword | |
^--- the cool artifact
Now, if ADOM were to pick a random spot, it would mostly choose the
dagger, sometimes the sword, and very rarely the artifact. That's the
way item types are selected.
There are a few restrictions when choosing them. For example, fluff
balls can't be generated in the WDL, flamable items in the ToEF, corpses
at anytime (banshee level corpses are generated manually).
After choosing the item type, the function calls another function, which
attunes the item's properties. There are many detail here, such as dice
rolls for the number of items such as arrows, quarrels etc. but I won't
list them here.
The first spellbook generated for wizards born in Samalamder always
contains Firebolt. If the wizard was born in another month, he gets
either of magic missile, acid bolt, fire bolt, frost bolt or lightning
bolt.
Necromancers always get frost bolt. Druids get lightning bolt.
That's about all I found. I know that there are rumors that spell-
casters find spellbooks more often, and I'm not saying I didn't miss
anything. However, as someone said, the human mind often sees patterns
where none exist. Maybe it was because spellcasters have lower spell
costs, and they don't need to refresh their spells that often. If you'll
insist, I'll look more carefully.
(*) Note:
That's not actually what ADOM does, but it works this way. ADOM uses a
little more optimized method.
--
Knowledge belongs to the World.
That's why I'm here for.
Mikko Sorsa
"Vladimir Panteleev" <pv_wer...@programist.ru> wrote
> Mikko Sorsa <msorsa4> <@ihatespam.welho.com> says...
>
> Who do you take me for? When has it happened that I forgot to add
> spoiler space?
>
Just making sure, it's mainly me who forgets these things ; )
I've actually programmed a little and even a little RPG, so this makes
perfect sense. Actually that is how I imagined how it would be.
> That's about all I found. I know that there are rumors that spell-
> casters find spellbooks more often, and I'm not saying I didn't miss
> anything. However, as someone said, the human mind often sees patterns
> where none exist. Maybe it was because spellcasters have lower spell
> costs, and they don't need to refresh their spells that often. If you'll
> insist, I'll look more carefully.
Yes, I'm quite sure this is just a case of seeing patterns where there are
none. But I'm quite surprised that there are no special considerations
on different dungeons and such (except for the ID). I wonder if
pickpocketing uses similar system to create the loot? Thanks again,
Vladimir, it's nice to have someone around to check out these things!
Mikko
Malte Helmert
>
> Who do you take me for? When has it happened that I forgot to add
> spoiler space?
>
> .
> .
> .
> .
> ...
> .
> .
> .
> .
> ..
> .
> .
> .
> .
> ...
> .
> .
> .
> .
>
(1):
> In the ID, the danger value is equal to the current depth (level nr).
(2):
> MaxItemQuality=Max(1,Min(50,LevelDangerValue))
(3):
> If the PC is in the ID, the maximum item quality is reduced to one
> third:
> MaxItemQuality=Max(1,MaxItemQuality/3)
(4):
> This means that the maximum item quality level is reached at ID:16.
No, if (1), (2) and (3) are correct, we obtain the following:
ID 16 => danger level 16 by (1)
=> MaxItemQuality = Max(1, Min(50, 16)) = 16 by (2)
=> MaxItemQuality = Max(1, 16 / 3) = 5 by (3)
Similarly:
ID 48 => danger level 48 by (1)
=> MaxItemQuality = Max(1, Min(50, 48)) = 48 by (2)
=> MaxItemQuality = Max(1, 48 / 3) = 16. by (3)
So it looks like at least one of the four statements above is wrong.
I cannot say which, of course.
Malte
DavidByron
>
> .
> .
> .
> .
> ...
> .
> .
> .
> .
> ..
> .
> .
> .
> .
> ...
> .
> .
> .
> .
>
> Items whose quality value is higher than the level's maximum item
> quality can't be generated. So you won't be able to find anything cool
> on ID:1 no matter how much you try (AoLS's danger level is 2, so is
> potion of water's).
Which implies that the optimum level to search for AoLS in the ID is
level 6/7. Or does the formula round to the nearest making it 5/6? 2
seems a very small number for the AoLS. What sort of stuff has levels
higher than 16?
A table of these weighted probabilities and the levels would be most
helpful, as would more information on special items not dropped in
various places. For example I don't think I've ever seen scrolls or
spellbooks in the WDL.
What about item generation due to monster kills?
Vladimir Panteleev
<hel...@informatik.uni-freiburg.de> says...
You're absolutely right. I have noticed the mistake in calculus this
morning and thought of correcting it ASAP.
Number 4, of course. ID 48. >:]
Vladimir Panteleev
<davidb...@yahoo.com> says...
Right, completely forgot about that.
You can ask ToGu to update his data dumps which are already in HTML
tables, or get my own datadump, which is rather unorganized:
http://www.geocities.com/adombot/ items.txt
Please copy the link to your browser's address bar if you're using
Google Groups, as GeoCities accused me once to posting direct links to
files on different forums and such.
Note that this datadump is already distributed together with AdomBot,
but I have only recently added item danger values.
> What about item generation due to monster kills?
Same old... except for some items specific for some monsters. Like frog
legs for giant frogs, and special items from special monsters (like the
axe of the minotaur emperor from the minotaur emperor etc.).
DavidByron
> In article <588db677.04032...@posting.google.com>, DavidByron
> <davidb...@yahoo.com> says...
> > Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1acbfba7d...@news.individual.net>...
> >
> > >
> > > .
> > > .
> > > .
> > > .
> > > ...
> > > .
> > > .
> > > .
> > > .
> > > ..
> > > .
> > > .
> > > .
> > > .
> > > ...
> > > .
> > > .
> > > .
> > > .
> > >
>
> > level 6/7. Or does the formula round to the nearest making it 5/6? 2
> > seems a very small number for the AoLS. What sort of stuff has levels
> > higher than 16?
And the optimum for spellbooks or !oGA is 1/2 which I figured
anyway.... but getting the extra water and a chance of the AoLS is
probably worth going to 6/7 unless the only things left on the
shopping list are spellbooks.
> > What about item generation due to monster kills?
>
> Same old... except for some items specific for some monsters. Like frog
> legs for giant frogs, and special items from special monsters (like the
> axe of the minotaur emperor from the minotaur emperor etc.).
By "same" do you mean that monster-kill drops are dependent only on
the level the monster dies in and not on the level of the monster say?
The items with the very high danger levels appear to mostly be higher
metal weapons and armor, so the 'ID university' turns out very good
for wizard types but not so good for fighters.
Danger level > 16 (not found in ID?)
adamantium chain mail
adamantium plate mail
eternium scale mail
eternium chain mail
eternium plate mail
large eternium shield
tower eternium shield
eternium cap
large eternium hammer
etc
scroll of chaos resistance = level 10
potion of cure corruption = level 2
7LBs = level 3
It says a scroll of danger is level 1. Has anyone found one on the
low levels of the ID?
Because the ring of djinn summoning is level 8 there's a slightly
better chance of getting one with potion of exchange at level 8
dungeons. Probabilities for most rings are similar and at level 8,
according to the theory that exchange would just regenerate rings at
the level of the dungeon, minus the chnace of getting the same potion,
the odds of getting djinn summoning rings would be about 1 in
29.25-31.5 which should be good enough for a wish engine with potions
of exchange and a small amount of holy water. Since this doesn't
tally with my experience (although I'll give it another try now)
something else may be going on.
I also wonder about the potions that the fool gives you - is there a
danger level limit on those items?
Przemyslaw Brojewski
: Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1acbfba7d...@news.individual.net>...
I get the impression that such a table would be influenced by Mana.
Why I get the impression... Well, this is a topic for another post.
look for the scummer's diary next week.
brojek.
Juho Julkunen
Two comments:
> Who do you take me for? When has it happened that I forgot to add
> spoiler space?
>
> .
> .
> .
> .
> ...
> .
> .
> .
> .
> ..
> .
> .
> .
> .
> ...
> .
> .
> .
> .
>
> Necromancers always get frost bolt. Druids get lightning bolt.
It at least used to be possible for a Necromancer to start with a Tract of
Chaos. (g16pre2 or maybe 1.0.0.) Possibly it has been fixed since.
>
> That's about all I found. I know that there are rumors that spell-
> casters find spellbooks more often, and I'm not saying I didn't miss
> anything. However, as someone said, the human mind often sees patterns
> where none exist. Maybe it was because spellcasters have lower spell
> costs, and they don't need to refresh their spells that often. If you'll
> insist, I'll look more carefully.
I'm pretty sure spellcasters do find more books, or possibly certain classes
find less than average.
--
Juho Julkunen
Przemyslaw Brojewski
: Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1accbd8a3...@news.individual.net>...
But of course. Someone has discovered this, pacified Dwarven Halls with
gremlins and water traps, and went happy pickpocketing. Only
corruption there gets in a way.
brojek.
Vladimir Panteleev
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ..
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > >
> better chance of getting one with potion of exchange at level 8
> dungeons. Probabilities for most rings are similar and at level 8,
> according to the theory that exchange would just regenerate rings at
> the level of the dungeon, minus the chnace of getting the same potion,
> the odds of getting djinn summoning rings would be about 1 in
> 29.25-31.5 which should be good enough for a wish engine with potions
> of exchange and a small amount of holy water. Since this doesn't
> tally with my experience (although I'll give it another try now)
> something else may be going on.
Don't forget about the item "creation probability" value. Rings of
Djinni Summoning have the creation probability value of 200, which isn't
very small, but is countered by their danger value. Their danger value
8, meaning they should start to appear around ID:24.
> I also wonder about the potions that the fool gives you - is there a
> danger level limit on those items?
Never noticed one, except the danger value of the current level. The
danger level in Terinyo is 1 >:[ You can always of course drag him to
the DH >:]
DavidByron
> DavidByron <davidb...@yahoo.com> wrote:
> : Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1accbd8a3...@news.individual.net>...
> :> In article <588db677.04032...@posting.google.com>, DavidByron
> :> <davidb...@yahoo.com> says...
> :> > Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1acbfba7d...@news.individual.net>...
> :> >
> :> > >
> :> > > .
> :> > > .
> :> > > .
> :> > > .
> :> > > ...
> :> > > .
> :> > > .
> :> > > .
> :> > > .
> :> > > ..
> :> > > .
> :> > > .
> :> > > .
> :> > > .
> :> > > ...
> :> > > .
> :> > > .
> :> > > .
> :> > > .
> :> > >
> : By "same" do you mean that monster-kill drops are dependent only on
> : the level the monster dies in and not on the level of the monster say?
>
> But of course. Someone has discovered this, pacified Dwarven Halls with
> gremlins and water traps, and went happy pickpocketing. Only
> corruption there gets in a way.
How would you discover that by game play unless you spent a ton of
time producing statistics to ensure the distributions were the same?
At any rate pick-pocketing is something else again but if it the
distribution does remain the same then pick-pocketing from water trap
Gremlins while on level 2 would probably represent the optimum way to
find an AoLS quickly.
DavidByron
> Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1accbd8a3...@news.individual.net>...
> > > >
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ..
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > >
I revisited the potion of exchange / ring of djinn summoning wish
engine and it seemed to kick in this time. Using the data which
Vladimir provided to create the optimum conditions I went to level 8
and over the course of 100 exchanges got djinn rings 5 times (and also
again on the 101th turn) which was more than I was expecting, but that
could be chance (I had lucky and fate smiles).
I started the engine in earnest (ie without saving / reloading) with
50-odd potions of exchange and 19 ordinary rings. I ran through 4
sets of djinn rings and ended up with a stock of about 200 potions of
exchange.
Once you get rolling it's a bit tedious but it's actually only about
10 actions per wish you end up with is a pretty good deal by any
measure.
(Dip rings in !oEx) x about 30
Bless 19 rings of djinn summoning
(Use ring to wish for "exchanges") x 10
(Use ring to wish for whatever you want) x 9
(Bless !oEx 4 at a time) x 7
repeat by dipping the 19 brass rings in !oEx, etc
And in theory you'd need to ocassionally wish for some water and bless
the water to keep your level of holy water up. Initially you'll want
to build up a stock of !oEx of about 2-300 I would think, using your
'whatever you want' spare wishes. After that you'd be getting 6-9
useful wishes every iteration which takes about 60 turns each.
To get started you will probably need a lot of !oEx but even one will
do to have a chance at it. What you really need is 19 rings of the
same kind. Presumably from milking a ring shop or the casino.
S Sawchuk
DavidByron wrote:
> davidb...@yahoo.com (DavidByron) wrote in message news:<588db677.04032...@posting.google.com>...
> > Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1accbd8a3...@news.individual.net>...
> > > > >
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ...
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ..
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ...
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > >
>
> I revisited the potion of exchange / ring of djinn summoning wish
> engine and it seemed to kick in this time.
[snip explanation]
>
>
> To get started you will probably need a lot of !oEx but even one will
> do to have a chance at it. What you really need is 19 rings of the
> same kind. Presumably from milking a ring shop or the casino.
Interesting information that you present, but just curious, or ineducated, why 19 rings of the same kind? Also, it seems
collecting ~30 potions of exchange for an effective trial would also prove difficult. Most of the time by Dwarftown I'm lucky
to have 1 or 2. Then again I'm not exactly looking for them. Odds are unusually favored as you say, you may have started a new
wave of fun for thr r.g.r.a.!
--S Sawchuk
Brian Eriksen
Very chancy. It easily stops. I started one with 75 holy! and 7 wishes. No
!oEx. Used the initial wishes for !oEx on ID8 and started
5 dips for 1st stack
29 dips for 2nd stack
12 dips for 3rd stack
at this point i am rolling and up 40 wishes. wishing only for !oEx to
continue. plenty of rings
91!! dips for 4th stack (thats 30-40 wishes worth of !oEx right there)
50 dips for 5th stack
79 dips and the engine is depleted costing me 70 vials of holy! and 7
wishes in total.
By all means go for it. Just don't expect a miracle. It's a gamble. I
demands too much to start it to be worth the risk. Like on the roulette
you should also stop while you are ahead.
DavidByron
more replying to myself....
On item levels, dungeon levels, danger levels...
> > >
> > > >
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ..
> > > > .
> > > > .
> > > > .
> > > > .
> > > > ...
> > > > .
> > > > .
> > > > .
> > > > .
> > > >
The rings produced by dipping into potions of exchange on level 8 of
the CoC fit the hypothesis that the exchanged items are just the same
as any other drop. All of the rings but one (21/22) that had danger
level 8 or less were produced in 100 tries. None of the rings with
danger level greater than 8 were produced.
If we assume pick pocketing produced items follow the same rules then
it might be preferable to try to gain items that are small enough
through this means. (1) because about half the weight of item
creation probabilities are >10 stones so they won't occur (2) because
you can produce an engine for item creation by pickpocketing fairly
easily whereas an item production engine for >10 stones would be...
tricky. Hmm.
For <=10 stones you can drop a gremlin bomb in the WDL and you don't
have to take actions to kill the gremlins because they drown, meaning
all you have to do is sit there picking pockets. I'm not sure what
the level of the WDL is but from previous trials of the gremlin bomb /
pick-pocket thing it must be at least level 8 because I picked up a
scroll of great identify. That means you can pick pocket rings of
djinn summoning there, and since they have quadruple the chances of
creation of an amulet of life saving, and the increase in "weight"
from level 2 (AoLS) to level 8 (=oDS) is not all that great (probably
less than doubled) I would now suspect that the quickest way to get a
wish and hence an AoLS - after you've drunk all the pools in Darkforge
maybe.
I think someone said the pools have a 1% probability of giving a wish.
I'd guestimate the probability of a =oDS from a sucessful pick
pocketing in the WDL at around 0.1% Since you don't always get
anything at all it could take a few thousand attempts. Hmm. I
suspect the gremlins would probably thin out by then - distrubting
their 200 max across the level, unless you block their spread with
multiple doors perhaps?
A shop of rings or wands would provide a less tedious solution
although they might take more game turns.
That prompts the question - do shops restock with items based on the
level they are in? I think in the archives someone said
HoleInTheWall's shop had 7LBs in it once so it would have to be at
least level 3. That would imply you could wait for an AoLS to turn up
in the shop there if you just waited long enough - and if you finessed
the restock function (I think there's more to it than just picking
everything up and dumping it in the corner of the shop - I found
better results in the casino shop if I bought everything - if I just
piled stuff up the restock would only partially fill the floor space).
I am not sure about an item engine for >10 stones. Probably the
easiest would be to get some level with a lot of monster production
and then get some background effect to kill all the creatures for you
so all you have to do is pick up stacks of treasure. WDL/gremlin bomb
doesn't work because the gremlins don't drop - perhaps no 'summoned'
creatures do? I'll have to try Banshee/Living Forest (suggested in
the archives). If you could get a good >10 stone item engine it might
be faster (in actions) than the WDL/gremlin bomb since you could grab
stacks of items with a single action.
N. Williams
>On Fri, 25 Mar 2004, DavidByron wrote:
>
>> davidb...@yahoo.com (DavidByron) wrote in message news:<588db677.04032...@posting.google.com>...
>> > Vladimir Panteleev <pv_wer...@programist.ru> wrote in message news:<MPG.1accbd8a3...@news.individual.net>...
>> > > > >
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > ...
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > ..
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > ...
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > > .
>> > > > >
>>
>Very chancy. It easily stops. I started one with 75 holy! and 7 wishes. No
>!oEx. Used the initial wishes for !oEx on ID8 and started
Did you mean CoC8?
With a little more inside information or a lot more experimentation one could
apply a bit of statistics to determine the size ring stack and initial !oEx
stack necessary to create a reliably runaway wish engine. Of course, characters
able to create (and survive) !oRaw Chaos could catalyze the process somewhat,
assuming that !oEx and !oRC have the same item generation qualities.
Looking forward to a troll barbarian Archmage (Artificer?) post.
DavidByron
Because that's how many you can dip into the same potion of exchange.
Someone just compared it to roulette but it is a roulette game that's
very heavily slanted in your favour. To give yourself the best odds
you must maximize the usage of the potions which means using 19 rings.
> collecting ~30 potions of exchange for an effective trial would also
> prove difficult.
Not really. Once you have 19 rings you can try to exchange them every
time you get a potion. Odds are not good but if you get lucky you
have 19 wishes which gives you 57 potions of exchange. 57 will
certainly not guarantee that you have enough "stock" to avoid unlucky
breaks but it does have a good chance of giving you the lock.
There is a finite probability starting from 57 potions that you'll get
to 300 potions before you go broke. I'd guess the probability at
around 1/2 to 2/3 in your favour, but if you run out of potions you've
lost nothing but the !oEx you used initially and can try it all again
from scratch if you find another potion.
If you do get to 300 potions in stock I'd guess you a safe to start
using the wishes on other stuff.
So you'd be talking about
(1) putting in the effort to get the 19 rings - quite some effort
(2) upon finding a !oEx deciding to use it for the wish engine instead
of -- say -- attribute re-arrangement, preparing to be disapointed as
it's likely to take many tries
(3) upon getting the first set of 19 wishes you have to decide whether
to cash in some or maximise your chance at the wish engine, if you go
for the wish engine then you have a good chance at that stage, but a
decreasing chance if you decide to use a couple of the wishes on other
stuff.
So all in all, definately not a method that is easy to implement.
Which is as it should be considering the rewards.
My stochastic process theory is a little rusty or I'd calculate the
odds on getting arbitrarily large number of wishes from the initial
19.
Léon Planken
on Thu, 25 Mar 2004 19:23:44 -0400, S Sawchuk wrote:
<snip />
could you fix your line length to about 76 characters?
Thank you,
Léon (aka Oliphaunt)
--
"Linux is very fast.
It can do an infinite loop in about 5 seconds."
-- Linus Torvalds
Adam Alexander
"DavidByron" <davidb...@yahoo.com> wrote in message news:588db677.04032...@posting.google.com...<Snip>
> That prompts the question - do shops restock with items based on the
> level they are in? I think in the archives someone said
> HoleInTheWall's shop had 7LBs in it once so it would have to be at
> least level 3. That would imply you could wait for an AoLS to turn up
> in the shop there if you just waited long enough - and if you finessed
> the restock function (I think there's more to it than just picking
> everything up and dumping it in the corner of the shop - I found
> better results in the casino shop if I bought everything - if I just
> piled stuff up the restock would only partially fill the floor space).
>
I have had AoLS generated in the Holeinthewall shop many times.
John S
news:<40636A00...@sympatico.ca>...
> > DavidByron wrote:
> >
> > > davidb...@yahoo.com (DavidByron) wrote in message
news:<588db677.04032...@posting.google.com>...
> > > > Vladimir Panteleev <pv_wer...@programist.ru> wrote in message
news:<MPG.1accbd8a3...@news.individual.net>...
> > > > > > >
p
o
i
l
e
r
s
p
a
c
e
.
.
.
.
.
.
.
.
.
.
<massive wish engine discussion snip>
> My stochastic process theory is a little rusty or I'd calculate the
> odds on getting arbitrarily large number of wishes from the initial
> 19.
I wrote a little program to test this. It relies on a few assumtions:
1. On CoC 8, there are 22 available rings, and each one is generated
equally. Therefore, the odds of a successful generation are 1/22.
2. You get 3 PoEX every wish. Therefore, you get 57 PoEX if you spend a
wish with every one.
The way I calculated it is that if you get 10000 PoEX, you're in the clear
for as many wishes as you want. I did 100000 trials.
Result? Starting with 57 PoEX (ie after a successful attempt, you have
about a 91% chance to get a wish engine going if you use every wish on PoEX.
Starting with 1 (ie you just found a PoEX), you have about a 4% chance.
Apparantly this is good enough to keep a reliable wish engine going. I'd
almost suggest this be RFEd to change somehow, as this isn't all that much
work for a nearly unlimited amount of wishes.
--
John S.
S Sawchuk
Léon Planken wrote:
> I failed miserably to keep a straight face when I read that
> on Thu, 25 Mar 2004 19:23:44 -0400, S Sawchuk wrote:
>
> <snip />
>
> could you fix your line length to about 76 characters?
I've been asked about this before, perhaps the r.g.r.a can find an
answer to my problem? I know some of you are good at this. My line
length is set to 72 characters, however it doesn't seem to work! I have
been asked many a time to change my line length, I'll look up my
preferences, see they are set to 72, and my line length will continue to
come out wrong!
>
>
> Thank you,
> Léon (aka Oliphaunt)
> --
>
Your welcome?
--S Sawchuk
Stephen White
> Léon Planken wrote:
>
>>I failed miserably to keep a straight face when I read that
>>on Thu, 25 Mar 2004 19:23:44 -0400, S Sawchuk wrote:
>>
>><snip />
>>
>>could you fix your line length to about 76 characters?
>
> I've been asked about this before, perhaps the r.g.r.a can find an
> answer to my problem? I know some of you are good at this. My line
> length is set to 72 characters, however it doesn't seem to work! I have
> been asked many a time to change my line length, I'll look up my
> preferences, see they are set to 72, and my line length will continue to
> come out wrong!
I think 72 is too short if you are posting a screenshot or flg... *checks*
Verily, the maximum line output in a .flg is 79 characters, so you need
at least 80, and that's assuming that your reader is smart enough not to
re-wrap when quoting something, even if quoting takes it over the normal
limit.
Al
> davidb...@yahoo.com (DavidByron) wrote in message news:<588db677.04032...@posting.google.com>...
>
> more replying to myself....
> On item levels, dungeon levels, danger levels...
>
> > > >
> > > > >
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ...
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ..
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > > ...
> > > > > .
> > > > > .
> > > > > .
> > > > > .
> > > > >
> easiest would be to get some level with a lot of monster production
> and then get some background effect to kill all the creatures for you
> so all you have to do is pick up stacks of treasure. WDL/gremlin bomb
> doesn't work because the gremlins don't drop - perhaps no 'summoned'
> creatures do? I'll have to try Banshee/Living Forest (suggested in
> the archives). If you could get a good >10 stone item engine it might
> be faster (in actions) than the WDL/gremlin bomb since you could grab
> stacks of items with a single action.
For an Item Engine, its better to lure the Banshee to the BDC.
If you have a couple of wishes under your belt, you can also
create multiple "banshee boxes" with door creation.
-Al
The Beerslayer
> I've been asked about this before, perhaps the r.g.r.a can find an
> answer to my problem? I know some of you are good at this. My line
> length is set to 72 characters, however it doesn't seem to work! I have
> been asked many a time to change my line length, I'll look up my
> preferences, see they are set to 72, and my line length will continue to
> come out wrong!
What news client are you using, and on what operating system? I cannot
tell from your headers...
-- Jeff
-- aka The Eternal Newbie :)
---------------------------------------------------------------
Doubts are more cruel than the worst of truths. -- Molière
---------------------------------------------------------------
S Sawchuk
Stephen White wrote:
O.K. I changed that over, perhaps an addition to the r.g.r.a FAQ, as a new line
length, as it has been 72-76 as far as I can remember.
--S Sawchuk
DavidByron
> 1. On CoC 8, there are 22 available rings, and each one is generated
> equally. Therefore, the odds of a successful generation are 1/22.
The weighting of the 22 are as follows:
7: 500
14: 200 including the djinn ring)
1 : 50
total weight is 6350 which gives odds of 31.75:1 but recall that with
!oEx you can't get the same thing you started with making the odds
either 29.25:1, 30.75:1 or 31.5:1 depending on what you start with.
On avergae then it's about 30:1 odds.
I tried the engine before but lacking the specific figures above I
couldn't say for sure how likely it was to catch.
> Result? Starting with 57 PoEX (ie after a successful attempt, you have
> about a 91% chance to get a wish engine going
Yeah now if your odds are 30:1 instead of 22:1 what is it? I guessed
50-67%
> Starting with 1 (ie you just found a PoEX), you have about a 4% chance.
Which would be 2% with the new figures.
> Apparantly this is good enough to keep a reliable wish engine going. I'd
> almost suggest this be RFEd to change somehow, as this isn't all that much
> work for a nearly unlimited amount of wishes.
Not all that much work? If you are a wizard all you have to do is go
to the library, take great book caster trait and wait for
Silvernight. This engine BALANCES the game a little more towards
non-Wizards. Hence the title of this thread. It's not exactly easy
to find 19 rings of the same type either. And then at 2% per !oEx on
AVERAGE you'd need to find 50 of then to get this to work. 19 rings
and 50 !oEx and you call that easy !?!?
S Sawchuk
The Beerslayer wrote:
> S Sawchuk (any relation to Terry?) wrote:
>
Absoultely none, get asked that frequently. He spells his last name Sawchuck
however. But, we are both Ukrainian, and it's a Ukrainian name.
>
> > I've been asked about this before, perhaps the r.g.r.a can find an
> > answer to my problem? I know some of you are good at this. My line
> > length is set to 72 characters, however it doesn't seem to work! I have
> > been asked many a time to change my line length, I'll look up my
> > preferences, see they are set to 72, and my line length will continue to
> > come out wrong!
>
> What news client are you using, and on what operating system? I cannot
> tell from your headers...
The Netscape newsreader, comes with Sympatico, on Windows 98. Any ideas?
Anyone?
--S Sawchuk
John S
Closer to 77%, actually, using 30:1.
>
> > Starting with 1 (ie you just found a PoEX), you have about a 4% chance.
>
> Which would be 2% with the new figures.
2.5%.
>
> > Apparantly this is good enough to keep a reliable wish engine going.
I'd
> > almost suggest this be RFEd to change somehow, as this isn't all that
much
> > work for a nearly unlimited amount of wishes.
>
> Not all that much work? If you are a wizard all you have to do is go
> to the library, take great book caster trait and wait for
> Silvernight. This engine BALANCES the game a little more towards
> non-Wizards. Hence the title of this thread. It's not exactly easy
> to find 19 rings of the same type either. And then at 2% per !oEx on
> AVERAGE you'd need to find 50 of then to get this to work. 19 rings
> and 50 !oEx and you call that easy !?!?
Compared to the reward, yes. 19 rings aren't THAT hard to get. Find one
ring store (yes, I know, not guarenteed) and it's easy. Plus, you don't
NEED 19 rings. Here are some stats:
19 rings: 77%
18 rings: 74%
17 rings: 70%
16 rings: 65%
15 rings: 59%
14 rings: 52%
13 rings: 43%
12 rings: 32%
11 rings: 18%
10 rings: 0.275%
17 or 18 would be enough for a good chance. I suppose you're right about
the initial chance for startup, though. 40 PoEX is a lot.
--
John S.
matija
> I've been asked about this before, perhaps the r.g.r.a can find an
> answer to my problem? I know some of you are good at this. My line
> length is set to 72 characters, however it doesn't seem to work! I
> have been asked many a time to change my line length, I'll look up my
> preferences, see they are set to 72, and my line length will continue
> to come out wrong!
ISTR that old crappy version of netscape having these exact
same problems when i used it (about 4 years ago). i've since
switched to gravity (which was nice, but became obsolete very
soon), and for the last year or more i've been using dialog,
maybe even the best newsreader out there. it has some really
nifty things, like easy wrap toggling and inserting internal
wrap override characters, which is just about perfect for
posting (F/V)LG's - they can get posted without wrapping,
and the rest of your text wraps at the set margin.
anyway, give different newsreaders a try, there are dozens
out there which perform their task better than ol' netscape:
gravity, agent, dialog, the new mozilla/firebird thingie, ...
--
there is a cheer. the gnomes have learned a new way to say hooray. [-shpongle]
address is scrambled - remove the superfluous "x" marks to reply
The Beerslayer
>
> I asked:
>> (any relation to Terry?)
>
> Absoultely none, get asked that frequently. He spells his last name Sawchuck
> however. But, we are both Ukrainian, and it's a Ukrainian name.
The Terry Sawchuk I am thinking of did not spell his name "Sawchuck",
but "Sawchuk", exactly like yours. See (http://tinyurl.com/25lld) for
verification. Is there another famous "Terry Sawchuck" that I don't
know about?
>>> My line
>>> length is set to 72 characters, however it doesn't seem to work!
>>
>> What news client are you using, and on what operating system? I cannot
>> tell from your headers...
>
> The Netscape newsreader, comes with Sympatico, on Windows 98. Any ideas?
There's your problem right there. Netscape's newsreader is, well, let's
be kind and say "not very good". I have no idea how to properly
configure it to wrap the way I might want, so I cannot help you with
that. Sorry.
As Matija pointed out back in the original thread, there are a lot of
good newsreaders for Windows out there. I use the same one Matija does,
called "40tude Dialog", available here (http://www.40tude.com/dialog/).
It is still in development and has a number of strange features to it,
but is overall well designed and fairly stable.
A lot of people swear by Free Agent
(http://www.forteinc.com/agent/index.php), though personally I always
swore at it rather than by it. Free version is too limited and omits a
number of very key features, making it unusuable for me. If you don't
find those same limitations, it may work for you.
In a pinch, even Microsoft's Outlook Express can function as a
newsreader, although it has some security limitations.
Malte Helmert
> "DavidByron" <davidb...@yahoo.com> wrote in message
> news:588db677.04032...@posting.google.com...
>
>
Ah, a one-dimensional discrete random walk process.
> I wrote a little program to test this. It relies on a few assumtions:
> 1. On CoC 8, there are 22 available rings, and each one is generated
> equally. Therefore, the odds of a successful generation are 1/22.
Is this uniform distribution a realistic assumption?
> 2. You get 3 PoEX every wish. Therefore, you get 57 PoEX if you spend a
> wish with every one.
>
> The way I calculated it is that if you get 10000 PoEX, you're in the clear
> for as many wishes as you want. I did 100000 trials.
>
> Result? Starting with 57 PoEX (ie after a successful attempt, you have
> about a 91% chance to get a wish engine going if you use every wish on PoEX.
> Starting with 1 (ie you just found a PoEX), you have about a 4% chance.
Where do you get enough holy water?
But anyway, as a little mathematical exercise I have analyzed your model
analytically. These are the exact probabilities (no sampling), rounded
to six decimals. If you start with 57 PoEX, the chance is actually
91.1169%. ;-)
initial potion number probability of reaching 10,000 PoEX
0 0.000000
1 0.041584
2 0.081440
3 0.119638
4 0.156247
5 0.191334
6 0.224963
7 0.257192
8 0.288082
9 0.317687
10 0.346060
11 0.373254
12 0.399317
13 0.424297
14 0.448237
15 0.471182
16 0.493173
17 0.514249
18 0.534449
19 0.553809
20 0.572363
25 0.654185
30 0.720352
35 0.773858
40 0.817127
45 0.852117
50 0.880412
55 0.903293
57 0.911169
60 0.921797
65 0.936760
70 0.948860
75 0.958645
80 0.966557
85 0.972956
90 0.978131
95 0.982315
100 0.985699
110 0.990648
120 0.993884
130 0.996001
140 0.997385
150 0.998290
160 0.998882
170 0.999269
180 0.999522
190 0.999687
200 0.999795
210 0.999866
220 0.999913
230 0.999943
240 0.999963
250 0.999976
260 0.999984
270 0.999990
280 0.999993
290 0.999996
300 0.999997
310 0.999998
320 0.999999
330 0.999999
340 0.999999
342 1.000000
> Apparantly this is good enough to keep a reliable wish engine going. I'd
> almost suggest this be RFEd to change somehow, as this isn't all that much
> work for a nearly unlimited amount of wishes.
I disagree. 57 potions of exchange don't grow on trees.
I think this is a nice and inventive idea; it should remain possible.
Malte
John S
"Malte Helmert" <hel...@informatik.uni-freiburg.de> wrote in message
news:4064d3e8$1...@olaf.komtel.net...
Hmm. Holy water. That makes things a lot harder.
Well, the odds are closer to 1 in 30 due to weighting.
<from DavidByron>
> The weighting of the 22 are as follows:
>
> 7: 500
> 14: 200 including the djinn ring)
> 1 : 50
>
Also, he pointed out you can't get what you started with.
<snip>
> > Apparantly this is good enough to keep a reliable wish engine going.
I'd
> > almost suggest this be RFEd to change somehow, as this isn't all that
much
> > work for a nearly unlimited amount of wishes.
>
> I disagree. 57 potions of exchange don't grow on trees.
> I think this is a nice and inventive idea; it should remain possible.
>
I didn't mean break it, I meant make it less likely.
However, factoring in holy water (which I completely forgot) and the reduced
odds, I'd say it does drop to balanced levels. Scumming the ID would bring
in the startup gear, but you could also scum for almost everything you can
wish for. I guess it's fine as is. I think I'm going to see if I can get
it started by just heading to the ID...
--
John S.
Malte Helmert
> "DavidByron" <davidb...@yahoo.com> wrote in message
> news:588db677.04032...@posting.google.com...
>
>>"John S" <gtg...@prism.gatech.edu> wrote in message
>
> news:<c420ko$9ln$1...@news-int2.gatech.edu>...
>
>>>
>>>
>>>s
>>>p
>>>o
>>>i
>>>l
>>>e
>>>r
>>>
>>>s
>>>p
>>>a
>>>c
>>>e
>>>.
>>>.
>>>.
>>>.
>>>.
>>>.
>>>.
>>>.
>>>.
>>>.
>>>
>>><massive wish engine discussion snip>
>>
>>>Result? Starting with 57 PoEX (ie after a successful attempt, you have
>>>about a 91% chance to get a wish engine going
>>
>>Yeah now if your odds are 30:1 instead of 22:1 what is it? I guessed
>>50-67%
>
> Closer to 77%, actually, using 30:1.
77.9513%. In other words, in one of four cases, you'll spend 19 wishes
on nothing. :-)
Some other values:
For a better than fifty-fifty chance, you need at least 27 potions.
For 90%, you need 87 potions.
For 99%, you need 174 potions.
>>>Starting with 1 (ie you just found a PoEX), you have about a 4% chance.
>>
>>Which would be 2% with the new figures.
>
> 2.5%.
2.6176% :-)
> Compared to the reward, yes. 19 rings aren't THAT hard to get.
Just getting 19 rings isn't really enough. Even if you start off with 19
identical rings plus 27 potions of exchange (which *are* hard to get),
you will still lose everything 50% of the time.
> Find one ring store (yes, I know, not guarenteed) and
> it's easy. Plus, you don't NEED 19 rings. Here are some stats:
>
> 19 rings: 77%
> 18 rings: 74%
> 17 rings: 70%
> 16 rings: 65%
> 15 rings: 59%
> 14 rings: 52%
> 13 rings: 43%
> 12 rings: 32%
> 11 rings: 18%
> 10 rings: 0.275%
Hmmm... not sure what you're calculating here. If you are unlucky enough
not to have any other rings, you might possibly want to spend one of
your early wishes on rings.
Malte
John S
<snip more>
> Hmmm... not sure what you're calculating here. If you are unlucky enough
> not to have any other rings, you might possibly want to spend one of
> your early wishes on rings.
Wishing for more rings. Duh. I guess that makes that chart obsolete. But
anyway, it's just the results from the tests I ran from the program. Doing
the math is of course more reliable
--
John
Brian Eriksen
[SNIP: Even more engine talk]
>
> Some other values:
> For a better than fifty-fifty chance, you need at least 27 potions.
> For 90%, you need 87 potions.
> For 99%, you need 174 potions.
>
I cannot claim to fully understand how you reach these numbers, but if I
understand correctly there is 99% chance of claiming a stack of 19 wishes
with 174 !oEx. Since 19 wishes will give you and estimate of 45 !oEx there
is hence a considerable loss involved there.
I think the overall chance of propagation is a more valid number. That is
how is the chances of getting MORE wishes out of 45 !oEx. Everything seems
to be irrelevant as you cannot create more than 19 wishes a time anyhow.
We need to find a formula that describes the probability of creating an
infinately self-sustaining engine. That is the sum of probabilities up to
iteration n for n going to positive infinate. All I can think of is
SUM(1-n)[Prob(45)^n] , where Prob is the probability of getting a stack on
45 !oEx. I think it is incorrect. I am not a mathmatician.
Sorry if this input is redundant or wrong. I have just been unable to
sustain the engine on two occations even though I got the first couple of
iteration through.
[Snip: Talk on chance to get 19 wishes on one !oEx. Talk on wishes for
rings to feed the engine.]
-maddog
S Sawchuk
The Beerslayer wrote:
> S Sawchuk wrote:
> >
> > I asked:
> >> (any relation to Terry?)
> >
> > Absoultely none, get asked that frequently. He spells his last name Sawchuck
> > however. But, we are both Ukrainian, and it's a Ukrainian name.
>
> The Terry Sawchuk I am thinking of did not spell his name "Sawchuck",
> but "Sawchuk", exactly like yours. See (http://tinyurl.com/25lld) for
> verification. Is there another famous "Terry Sawchuck" that I don't
> know about?
>
No, I was just misinformed. I recount my father telling my about Terry as a child,
however I remembered him saying he spells his last name differently.
>
> >>> My line
> >>> length is set to 72 characters, however it doesn't seem to work!
> >>
> >> What news client are you using, and on what operating system? I cannot
> >> tell from your headers...
> >
> > The Netscape newsreader, comes with Sympatico, on Windows 98. Any ideas?
>
> There's your problem right there. Netscape's newsreader is, well, let's
> be kind and say "not very good". I have no idea how to properly
> configure it to wrap the way I might want, so I cannot help you with
> that. Sorry.
>
> As Matija pointed out back in the original thread, there are a lot of
> good newsreaders for Windows out there. I use the same one Matija does,
> called "40tude Dialog", available here (http://www.40tude.com/dialog/).
> It is still in development and has a number of strange features to it,
> but is overall well designed and fairly stable.
>
> A lot of people swear by Free Agent
> (http://www.forteinc.com/agent/index.php), though personally I always
> swore at it rather than by it. Free version is too limited and omits a
> number of very key features, making it unusuable for me. If you don't
> find those same limitations, it may work for you.
>
I'll take a stab at 40tude, someone also suggested the new Mozilla/Firebird, thanks
for the aid!
--S Sawchuk
S Sawchuk
The Beerslayer wrote:
> S Sawchuk wrote:
> >
> > I asked:
> >> (any relation to Terry?)
> >
> > Absoultely none, get asked that frequently. He spells his last name Sawchuck
> > however. But, we are both Ukrainian, and it's a Ukrainian name.
>
> The Terry Sawchuk I am thinking of did not spell his name "Sawchuck",
> but "Sawchuk", exactly like yours. See (http://tinyurl.com/25lld) for
> verification. Is there another famous "Terry Sawchuck" that I don't
> know about?
>
> >>> My line
> >>> length is set to 72 characters, however it doesn't seem to work!
> >>
> >> What news client are you using, and on what operating system? I cannot
> >> tell from your headers...
> >
> > The Netscape newsreader, comes with Sympatico, on Windows 98. Any ideas?
>
> There's your problem right there. Netscape's newsreader is, well, let's
> be kind and say "not very good". I have no idea how to properly
> configure it to wrap the way I might want, so I cannot help you with
> that. Sorry.
>
> As Matija pointed out back in the original thread, there are a lot of
> good newsreaders for Windows out there. I use the same one Matija does,
> called "40tude Dialog", available here (http://www.40tude.com/dialog/).
> It is still in development and has a number of strange features to it,
> but is overall well designed and fairly stable.
>
>
Yeah that appears to be a good program. I downloaded it and gave it a try, however
I still cant get mail as I don't know what to put in as the Host Server. Does
anybody happen to know. I tried re.games.rogulike.adom, figuring it wanted a
default server to look up messages for, that's not it. I tried sympatico.ca [my
ISP] and that had no avail.
--S Sawchuk
anxious triffid
> I downloaded it and gave it a
> try, however I still cant get mail as I don't know what to put in as
> the Host Server. Does anybody happen to know. I tried
> re.games.rogulike.adom, figuring it wanted a default server to look up
> messages for, that's not it. I tried sympatico.ca [my ISP] and that
> had no avail.
>
Well, there is a page here which gives a list of servers for sympatico.ca:
http://www.smr-usenet.com/tech/sympatico.shtml
I would try "news.sympatico.ca" and see how that works. I can also
recommend Xnews as being a very serviceable newsreader, with quite a
shallow learning curve.
S Sawchuk
Actually, thanks to your help I managed to figure out how to use 40tude.
It's a nice system, lots of options should I need them. And it's not as
aesthetically ugly as Netscape. [What an eyesore].
--S Sawchuk
matija
[re dialog]
> Yeah that appears to be a good program. I downloaded it and gave it a
> try, however I still cant get mail as I don't know what to put in as
> the Host Server. Does anybody happen to know. I tried
> re.games.rogulike.adom, figuring it wanted a default server to look
> up messages for, that's not it. I tried sympatico.ca [my ISP] and
> that had no avail.
uhm... you have some things mixed up. first of all, your news
server probably isn't your POP3 or SMTP server. judging by
your headers, news.bellglobal.com or news20.bellglobal.com
should be your NNTP server. your POP3 server is most likely
pop3.sympatico.ca or mail.sympatico.ca, and SMTP smtp.sympatico.ca
or mail.sympatico.ca. you can also apply for a free NNTP account
at fu-berlin.de.
anyway, look through netscape's options, you'll surely find your
servers listed somewhere.
Carl Chisholm
>
Here's a question. *Does* the B/U/C status of both the !oEx and rings
affect the outcome? Common ADOM sense says yes, but it might be worth
checking just how it affects things - if the status of one or the other has
minimal effect, it could be worthwhile for resources to skip. Since we
*are* talking about a whole lot of holy water. Any math on that?
John S
"Carl Chisholm" <temp...@flarg.sympatico.ca> wrote in message
news:Qii9c.40697$A_2.1...@news20.bellglobal.com...
The PoEX has to be blessed, or they won't work. The djinni rings have to be
blessed, or no wish. I'm not sure how many you can bless of each per holy
water, but my guess would be 19. Also, you have to bless the water you wish
for, as it will usually not be blessed. The holy water is not
insignifigant, though I think 1 wish for water in the 19 will cover it.
DavidByron
Nah. What really makes it obsolete is that if you scum a ring shop
until you have 10 rings of say searching or one of the other highest
probability ring types and you exchange those ten a few times pretty
soon you'll exchange them into the type of ring you got 2nd most of -
say 8 or 9 of them. From then on you'll have the 19 rings you need.
This comes at the price of wasting about 50% of the usefulness of the
first 3-6 potions of exchange.
Malte Helmert
Nonono, the correct interpretation is that with 174 !oEx, there is a 99%
chance of getting 10,000 (= practically infinite) !oEx from the engine.
> We need to find a formula that describes the probability of creating an
> infinately self-sustaining engine.
Exactly, but that's what the numbers above mean.
That is the sum of probabilities up to
> iteration n for n going to positive infinate. All I can think of is
> SUM(1-n)[Prob(45)^n] , where Prob is the probability of getting a stack on
> 45 !oEx. I think it is incorrect. I am not a mathmatician.
The formula I used is the following:
If you have 0 potions of exchange, you lose:
chance(0) = 0
If you have 10,000 or more potions of exchange, you win:
For n >= 10,000: chance(n) = 1
For other values of n, you get 56 more potions of exchange if you are
lucky (losing one for dipping the rings, getting 57 new ones from the
rings), but lose one otherwise [1]. I write "p" for the chance of
getting rings of djinni summoning (currently estimated at 1/30):
For 1 <= n <= 9999:
chance(n) = p * chance(n + 56) + (1 - p) * chance(n - 1)
If you start with a low number of potions of exchange and repeat dipping
until you either have 0 or at least 10,000 potions, then the highest
possible number at which you can stop is 10,055 (9,999 + 56).
The formulas above then define a system of 10,056 linear equations with
an equal number of variables, namely chance(0), chance(1), ...,
chance(10,055). This system of linear equations has a unique solution
[2]. You can find it using standard procedures such as Gaussian
elimination, although you need to make sure that you
a) have a program suited to sparse matrices, because otherwise it
will likely choke on the 100,000,000 matrix entries and
b) avoid floating point arithmetic because the results are likely
not to be mathematically stable, unless you know how to avoid
such problems.
Using standard methods, the floating point gave me too much trouble and
I ended up writing my own solver using rational numbers which exploits
the chararacteristica of this particular system of equations.
Let me recall what I wrote above:
>>For a better than fifty-fifty chance, you need at least 27 potions.
>>For 90%, you need 87 potions.
>>For 99%, you need 174 potions.
With the correct values (see footnote), these should actually be:
For a chance better than 50%, you need 27 potions.
For a chance better than 90%, you need 88 potions.
For a chance better than 99%, you need 176 potions.
You might wander what happens if we set the cut-off point lower than
10,000, i.e. what happens if we say we've got enough potions earlier,
say at 1,000 potions. That makes very little difference, because the
chances of going broke once you've got 1,000 potions are extremely small.
In fact, for the three thresholds above, I get *exactly* the same
probabilities for both cases:
27 => 0.506751 vs. 0.506751
88 => 0.900087 vs. 0.900087
176 => 0.990017 vs. 0.990017
Since these values are calculated exactly for the given level of
precision, this means that the difference between the two real values is
at most 0.000001. In other words, once you've got 1,000 potions, the
chance that you fall back to 0 rather than go up to 10,000 is less than
one-in-a-million.
Malte
[1] In my previous calculations, I actually calculated what happens if
your number of potions increases by *57* each time you're successful,
not 56. So these were slightly too optimistic.
[2] If you think about it from the right angle, this is kind of obvious,
but of course there's also a mathematical justification for that. If you
are interested, read up on "markov processes" and "stochastical matrices".
DavidByron
> > Where do you get enough holy water?
You wish for water and then use holy water to bless the water
The cost of a cycle (in wishes) are:
rings - none - you recycle the 19 brass rings
!oEx - about 10 wishes (3 per wish; need about 30 potions)
water - about 3-4 wishes
You need to bless:
1 set of 19 rings of djinn
about 7 stacks = 28 !oEx (about 3 of the wished for
!oEx will be initially blessed and you can deduce which)
about 2-3 stacks of potions of water need to be blessed
Of course you can save that last one if you have an altar
knocking around. So I suggested about 6 usable wishes per
iteration and about 60 turns are required per iteration.
> > I disagree. 57 potions of exchange don't grow on trees.
> > I think this is a nice and inventive idea; it should remain possible.
It now looks like you can get away with about 40 !oEx on
_average_ plus maybe another couple to help you get the
initial 19 rings.
> Scumming the ID would bring
> in the startup gear, but you could also scum for almost everything you can
> wish for. I guess it's fine as is. I think I'm going to see if I can get
> it started by just heading to the ID...
Gremlin bomb in the WDL might be faster. You don't have to
waste actions on anything but pick-pocketing. Rings would
be easiest from a ring shop or the casino.
Malte Helmert
If I recall correctly, potions of exchange *never* work on items unless
blessed. I do not know if they affect the B/U/C status of the items you
dip into them. The people who have tried to get the engine going will
know about that. If the status is not affected, then at least you don't
need additional holy water for the rings.
Malte
Malte Helmert
>>S Sawchuk wrote:
>>
>>>>What news client are you using, and on what operating system? I cannot
>>>>tell from your headers...
>>>
>>>The Netscape newsreader, comes with Sympatico, on Windows 98. Any ideas?
>
> I'll take a stab at 40tude, someone also suggested the new Mozilla/Firebird, thanks
> for the aid!
Firebird (which recently changed its name to Firefox, by the way) is a
pure browser, not a newsclient. From the Mozilla Suite, you could (and
IMO should) try either Mozilla, or Mozilla Thunderbird.
If you're currently using Netscape for browsing, then use Mozilla. The
browsers are based on the same technology, but Mozilla is much more
advanced. The same is true for their e-mail and newsclients.
On the other hand, if you're using another browser and don't want to
change, then Mozilla Thunderbird is a mail and news client without a
browser attached. In this case, though, I probably wouldn't bother.
Malte
Malte Helmert
> blessed, or no wish. I'm not sure how many you can bless of each per holy
> water, but my guess would be 19. Also, you have to bless the water you wish
> for, as it will usually not be blessed. The holy water is not
> insignifigant, though I think 1 wish for water in the 19 will cover it.
In my experience, the amount of stuff you can bless at one time is
calculated like this:
* If it weighs 20s or more, you can bless one.
* Otherwise, you can bless 19s / item weight at once, rounding down.
Thus, you can bless with one potion:
* 19 items weighing 1s.
* nine items weighing 2s.
* six items weighing 3s.
* four items weighing 4s.
* three items weighing 5s or 6s.
* two items weighings 7s, 8s, or 9s, or
* one item weighing more.
So you can bless 19 rings of djinni summoning, or four potions of
exchange, using one potion of holy water.
Malte
Malte Helmert
> "John S" <gtg...@prism.gatech.edu> wrote in message news:<c42kvr$ljh$1...@news-int.gatech.edu>...
>
>>"Malte Helmert" <hel...@informatik.uni-freiburg.de> wrote in message
>>news:4064d3e8$1...@olaf.komtel.net...
>>
>>>
>>>
>>>>
>>>>
>>>>s
>>>>p
>>>>o
>>>>i
>>>>l
>>>>e
>>>>r
>>>>
>>>>s
>>>>p
>>>>a
>>>>c
>>>>e
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>.
>>>>
>
>
>>>Where do you get enough holy water?
>
> You wish for water and then use holy water to bless the water
No, you don't do that before you've really got the engine going.
If you do it that way, you waste about 25% of your water. You wish for
water then drop it on an altar. But that still means losing a
significant fraction of your wishes.
> The cost of a cycle (in wishes) are:
> rings - none - you recycle the 19 brass rings
> !oEx - about 10 wishes (3 per wish; need about 30 potions)
> water - about 3-4 wishes
>
> You need to bless:
> 1 set of 19 rings of djinn
> about 7 stacks = 28 !oEx (about 3 of the wished for
> !oEx will be initially blessed and you can deduce which)
> about 2-3 stacks of potions of water need to be blessed
>
> Of course you can save that last one if you have an altar
> knocking around. So I suggested about 6 usable wishes per
> iteration and about 60 turns are required per iteration.
Let me see what this does to the success formula. Whenever you get your
19 wishes, you spend 15 wishes on potions of exchange and 4 wishes on
water. That gives you 45 potions of exchange and 12 potions of (soon to
be holy) water. We assume that four potions of exchange are already,
blessed; the rest can be blessed with 11 potions of holy water, the
remaining potion is used for blessing the rings when you get =oDS for
the next time.
So I need to plug 45 into the program where before I had 57. (Or 44,
where before I had 56.)
>>>I disagree. 57 potions of exchange don't grow on trees.
>>>I think this is a nice and inventive idea; it should remain possible.
>
> It now looks like you can get away with about 40 !oEx on
> _average_ plus maybe another couple to help you get the
> initial 19 rings.
I wouldn't use this technique if I only had a 50% chance of success.
Remember, "failure" means no usable wishes at all, and all potions of
exchange wasted. Using the new values, I get:
For a 50% success rate: 35 potions.
For a 75% success rate: 70 potions.
For a 90% success rate: 116 potions.
For a 99% success rate: 231 potions.
Potions of exchange are rare. In my experience, they are about as rare
as potions of gain attributes. If you can live with a 25% chance of
failure, then you need 70 potions of exchange. By the time you've found
them by pickpocketing, you will have found about 70 potions of gain
attributes, too, which takes most of the fun out of wishing. ;-)
Of course, you can speed up the process if you find wishing items, but
still it seems very tedious and somewhat risky.
Malte
John S
The optimal place to do this is CoC 8 due to the item generation, so a quick
jaunt to Dwarftown will handle the water.
DavidByron
technical
> Since these values are calculated exactly for the given level of
> precision, this means that the difference between the two real values is
> at most 0.000001. In other words, once you've got 1,000 potions, the
> chance that you fall back to 0 rather than go up to 10,000 is less than
> one-in-a-million.
>
> Malte
But every potion is another 3-4 turns or so the decison to take the
stock up to 1000 from 300 would be another couple of thousand actions.
A little tedius... What are the probabilities of failure for say 250
potions?
DavidByron
Except for some reason you can only bless 1 potion of gain attributes
at a time despite it weighing 4s. This has always seemed very odd to
me.
Andy Williams
> > >>>>
> > >>>>
> > >>>>s
> > >>>>p
> > >>>>o
> > >>>>i
> > >>>>l
> > >>>>e
> > >>>>r
> > >>>>
> > >>>>s
> > >>>>p
> > >>>>a
> > >>>>c
> > >>>>e
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>.
> > >>>>
> Except for some reason you can only bless 1 potion of gain attributes
> at a time despite it weighing 4s. This has always seemed very odd to
> me.
Game balance.
--
Andy Williams
ADOM Guidebook: http://www.andywlms.com/adom/
Mirror: http://users.rcn.com/andy.williams/adom/
Malte Helmert
> Malte Helmert <hel...@informatik.uni-freiburg.de> wrote in message news:<4065c018$1...@olaf.komtel.net>...
>
s
p
o
i
l
y
s
t
u
f
f
s
p
o
i
l
y
s
t
u
f
f
>>Since these values are calculated exactly for the given level of
>>precision, this means that the difference between the two real values is
>>at most 0.000001. In other words, once you've got 1,000 potions, the
>>chance that you fall back to 0 rather than go up to 10,000 is less than
>>one-in-a-million.
>
> But every potion is another 3-4 turns or so the decison to take the
> stock up to 1000 from 300 would be another couple of thousand actions.
> A little tedius... What are the probabilities of failure for say 250
> potions?
If you completely stop trying to get new potions once you reach 250 and
instead only wish for new potions whenever you fall below that value,
then you'll lose your potions sooner or later. This is true for any
fixed cut-off value.
Instead, I evaluated the following strategy:
Whenever you've got 250 or more potions, you don't spend all your wishes
on potions of exchange and water. Instead, you spend one wish on
something else, and only 18 wishes on the potions.
That way, the chance that you'll ever fall back to zero potions after
you've once had 250 is 0.7520%.
Malte
Dr. Michael Ulm
<hel...@informatik.uni-freiburg.de> wrote:
> Brian Eriksen wrote:
>>
>>
>>s
>>p
>>o
>>i
>>l
>>e
>>r
>>
>>s
>>p
>>a
>>c
>>e
>>.
>>.
>>.
>>.
>>.
>>.
>>.
>>.
>>.
>>.
>>
Aaargh! Treating this simple difference equation as a system of
linear equations is quite inefficient. Here is how to compute the
values for chance(n) with much reduced time and space requirements:
define chance0(n) and chance1(n) with the same recursion formula
as chance(n):
chance?(n) = p * chance?(n + 56) + (1 - p) * chance?(n-1),
where I used the abbreviation chance? to mean either
chance0 or chance1. As for chance(n), set
chance?(10000 + k) = 1 for k>=0.
Finally, set chance0(9999) = 0 and chance1(9999) = 1. This
allows you to recursively compute chance?(9998) then chance?(9997)
down to chance?(0). Finally, compute q such that
q * chance0(0) + (1 - q) * chance1(0) = 0,
then
chance(n) = q * chance0(n) + (1 - q) * chance1(n).
HTH,
Michael.
--
&&&&&&&&&&&&&&&&#@#&&&&&&&&&&&&&&&&
Dr. Michael Ulm
FB Mathematik, Universitaet Rostock
micha...@mathematik.uni-rostock.de
Przemyslaw Brojewski
: Hmmm... not sure what you're calculating here. If you are unlucky enough
: your early wishes on rings.
Oh come on. That's what ID is for. Or one can gamble and buy out the
whole casino until there's more then enough rings. Potions of exchange
are more of a problem so they are what I would spend early wish on.
brojek.
Przemyslaw Brojewski
: "John S" <gtg...@prism.gatech.edu> wrote in message news:<c42kvr$ljh$1...@news-int.gatech.edu>...
Noooo! You will loose the precious water. Since you are most likely
to do it on CoC:8 why not go to Dwarftown every several hundred
turns and use the altar?
Przemyslaw Brojewski
: Carl Chisholm wrote:
: Malte
Alas, the status of exchanged rings is random. Perheaps in the same
way as status of random drops.
brojek.
Malte Helmert
[snip formulas describing the probability of getting such a thing]
>>The formulas above then define a system of 10,056 linear equations with
>>an equal number of variables, namely chance(0), chance(1), ...,
>>chance(10,055).
[...]
> Aaargh! Treating this simple difference equation as a system of
> linear equations is quite inefficient. Here is how to compute the
> values for chance(n) with much reduced time and space requirements:
[snip]
Actually, that's exactly how I solved it. I didn't want to go too deep
into technical detail -- most people understand what a system of linear
equations is, at least.
Malte
Brian Eriksen
Sorry to jump back into the discussion at this advanced stage, but I think
we still need to face the practical application. Empirical experimentation
suggests that 45 blessed PoX is a reasonable yield from 19 wishes.
If we treat the system briefly as a binomial distribution with 45 trials
and 0 to 45 success' we get the number of wishes we can expect to get back
on average as
SUM(0:45)[Bin(n,45)*19] = 22,38
So there is an average yield of 17,8% from each cycle. I note that
1) It is quite a lot once the machine is running
2) It takes a lot of cycles (hopefully with decreasing probability of
failure) until you get clear and have enough PoX to make the system secure
so to speak.
I hope you understand what I am aiming at. Perhaps it is (yet again)
because I do not fully conprehend the previous sections, which I should
add I have read before they were snipped, but isn't it important to
discuss how to apply this wish engine? How to start it and when you should
stop stockpiling PoX? I must insist that we assume no more than 45 PoX,
since we have to stockpile a *lot* and hence must assume we have to use
wishes on blessing etc.
If I manage to start it I would continue to stockpile until I reach a set
goal of PoX and then use wishes on what ever I need when I am in excess of
that number and start stockpiling again when I drop below this arbitrary
PoX threshold.
-- maddog
Malte Helmert
> On Mon, 29 Mar 2004, Malte Helmert wrote:
>
s
p
o
i
l
s
t
u
f
f
s
p
o
i
l
y
s
t
u
f
f
s
p
o
i
l
y
s
t
u
f
f
> Sorry to jump back into the discussion at this advanced stage, but I think
> we still need to face the practical application. Empirical experimentation
> suggests that 45 blessed PoX is a reasonable yield from 19 wishes.
> If we treat the system briefly as a binomial distribution with 45 trials
> and 0 to 45 success' we get the number of wishes we can expect to get back
> on average as
> SUM(0:45)[Bin(n,45)*19] = 22,38
You can calculate this more easily by noticing that the expected value
of a binomially distributed variables is linear in the parameter n.
In this case, the expected number of wishes gained from 45 potions is 45
times the expected number of wishes gained from one potion.
> So there is an average yield of 17,8% from each cycle. I note that
> 1) It is quite a lot once the machine is running
> 2) It takes a lot of cycles (hopefully with decreasing probability of
> failure) until you get clear and have enough PoX to make the system secure
> so to speak.
> I hope you understand what I am aiming at. Perhaps it is (yet again)
> because I do not fully conprehend the previous sections, which I should
> add I have read before they were snipped, but isn't it important to
> discuss how to apply this wish engine? How to start it and when you should
> stop stockpiling PoX? I must insist that we assume no more than 45 PoX,
> since we have to stockpile a *lot* and hence must assume we have to use
> wishes on blessing etc.
I hope I understand what you're getting at, but let me rephrase the
issue in my own words so that you can see whether I'm answering your
question or not.
So far, we've mainly talked about how likely you are to go "bankrupt" if
you spend all your wishes on potions of exchange (and water, as
necessary). In practice, you will also want to spend some wishes on
other things, or else the "wish engine" is pointless.
One possible strategy you could use is the following:
A. Always wish for potions of exchange (and water) until you have
obtained at least N potions. Once you have N potions, spend all your
wishes on other things, until you fall below N. When you do, wish
for potions again until you have recovered, and so on.
The problem with strategy A is that it is guaranteed to eventually run
out of potions of exchange, no matter how high you choose the value of
N. For large enough values of N, you will probably not run out of wishes
before the heat death of the universe, but this state of affairs is
still not pleasing.
So I propose the following strategy instead:
B. If you have fewer potions than some threshold N, spend all your
wishes on potions. If you have at least N potions, use *one* wish
on something else each time you get 19 wishes, and use the remaining
18 wishes on potions of exchange/water.
In another posting, I wrote that for N = 250, this strategy will yield
an infinite amount of wishes (i.e. you'll never fall back to 0) with
99.248% probability, provided you ever reach the point where you have
250 potions. So maybe this strategy is too conservative, and you should
spend two or more wishes on something else.
The probability of first reaching 250 potions while spending all wishes
on feeding the engine is basically identical to the probability of
reaching 10,000 potions which I mentioned earlier. If I recall
correctly, it is about 50% if you start with 35 potions.
> If I manage to start it I would continue to stockpile until I reach a set
> goal of PoX and then use wishes on what ever I need when I am in excess of
> that number and start stockpiling again when I drop below this arbitrary
> PoX threshold.
Ok, that's strategy A from above. Using this strategy, you will lose all
potions sooner or later, but it's typically later. If you wish, I can
try to calculate the expected number of "net wishes" that you'll get
using this strategy before the engine runs out of steam.
I cannot do this now though -- I'm sitting at the wrong PC.
Malte
Brian Eriksen
> [SNIP: Calculating an average yield of ~18% as 1-(0.026176*45)]
Okay I think I understand. Say we start the engine with 80 holy!, 10 PoX
and 6 wishes. Not unattainable by taking a hard-hitting pet to the BDC.
Now with this system we have a reasonable risk of failure, so to avoid
being left without anything we initially apply one wish pr. cycle to our
needs. If the system breaks down after 4 cycles we would still have a 2
wish loss, but not a complete loss. If/once the system have spawned enough
PoX to give a (false) sense of security we start to exclusively build PoX
until a threshold and thereafter to harvest. If the system reverts
downward we switch strategy. At low levels we start to extract wishes
again. So ..
Low levels of PoX: Extract one wish pr. cycle.
Medium levels of PoX: Build up PoX stock.
High levels of PoX: Run the engine and extract an average of 3 wishes pr.
cycle.
Perhaps if one has extra ring piles a stockpile of RoDS should also be
build to make sure that wishes does not need to be spend right away.
Carl Chisholm
wish engine]
> Okay I think I understand. Say we start the engine with 80 holy!, 10
> PoX and 6 wishes.
Bit of a sidetrack here, but this makes me wonder - why? The theory being
worked with so far has been to create an infinite number of wishes. But in
practice, unless you're trying for an Archmage ending or a similarly wildly
extreme challenge, a much smaller number of wishes is usually *quite*
sufficent, right? Just getting the system to succeed once translates to,
what? 19 PoGA or AoLS? 38 SoCR? 65+ pairs of 7LB? 4000+ speed points? Aside
from the challenge of creating an infinite wish engine, what real practical
need is there (aside from the challenge of creating one)? Especally since,
to get the required startup equipment, you're needing an end-game quality
character *anyhow*. Perhaps a better set of calculations is to try and find
ways to maximize the success chances for the first iteneration, since
that's the most common scenario of need.
DavidByron
> Brian Eriksen wrote:
> > On Mon, 29 Mar 2004, Malte Helmert wrote:
> >
> >> Brian Eriksen wrote:
> >> > On Mon, 29 Mar 2004, Malte Helmert wrote:
> >> >
> >> s
> >> p
> >> o
> >> i
> >> l
> >> y
> >>
> >> s
> >> t
> >> u
> >> f
> >> f
> >>
> >> s
> >> p
> >> o
> >> i
> >> l
> >> y
> >>
> Aside
> from the challenge of creating an infinite wish engine, what real practical
> need is there (aside from the challenge of creating one)?
Heretic! A wish engine is an essential good!
You are not of the true faith. :)
Well of course the challenge was a lot of it, but first of all it was
because I wanted a wish engine for non-wizards, and secondly because
certain challenges such as getting all stats to 99 or having all
skills at 100% or having killed one of each kind of creature, or
limiting yourself to sitting in the SMC forever, either require a huge
amount hard work or a moderate amount of wishes - more than just half
a dozen that is.
> Especally since,
> to get the required startup equipment, you're needing an end-game quality
> character *anyhow*.
Nah. My scientists are currently working on how to get 40 !oEx the
most easy way. Of course it's going to be easiest with a wizard
character.... easy repeatable theft from same (ring) shop, possible
alchemy use for creation of booze, I'm not sure yet.
Honestly if I find a way that's TOO easy I may not post it here since
I wouldn't want to post a trick that's a game breaker. That doesn't
look likely though.
On a related note it begins to look like pick pocketed items' levels
are NOT related to the danger level of the dungeon level you are in at
the time...
> Perhaps a better set of calculations is to try and find
> ways to maximize the success chances for the first iteneration, since
> that's the most common scenario of need.
Someone could try to see if Luck / Fate smiles or the other kind of
luck helps.
Malte Helmert
> Carl Chisholm <temp...@flarg.sympatico.ca> wrote in message news:<T86ac.15620$1A6.7...@news20.bellglobal.com>...
(not spoily)
>>Aside from the challenge of creating an infinite wish engine, what real practical
>>need is there (aside from the challenge of creating one)?
Seems like I missed this posting.
> Heretic! A wish engine is an essential good!
> You are not of the true faith. :)
Yes, that's basically it. A wish engine is simply a nice achievement in
its own right. Becoming an archmage is much harder than simply winning
the game, but still several people were sufficiently motivated to go
through all the tedium.
Malte
Brian Eriksen
Also a lot of challenge games would benefit from this. At least a titanium
man that cannot gain HP and PP from leveling and must work on Mana
Battery and !oUH to do so. Furthermore a titanium man must have a super
pet making BDC harvesting easier (since you won't get über dragons). Not
easy, but easier none the less.
It could also mean the difference between a strong regular ending
character and an ultra ending candidate.
Finally I'd say that setting it up is very diffficult. Most likely you
need a lot of external factors working for you, such as potion shops etc.
> Malte
>
>
Malte Helmert
>>>goal of PoX and then use wishes on what ever I need when I am in excess of
>>>that number and start stockpiling again when I drop below this arbitrary
>>>PoX threshold.
[snip other strategy: extract few wishes at low level of PoEx, no wishes
at intermediate levels, more wishes at high levels]
I wouldn't extract any wishes at low levels of PoEx, because these are
the critical values. In fact, thinking it about it again, I'll go back
to your original suggestion, which I called strategy A in my other
posting. I know I said that it will eventually run out of steam, but
thinking about this again, "eventually" can be a very high number.
If you pick 250 as your magic number (i.e. whenever you have at most 250
potions, spend all wishes on potions of exchange and water, whenever you
have more, spend all wishes on something else), the expected number of
wishes you will get, provided that you ever obtain 251 potions, is on
the order of 1,000.
If you pick 500 as your magic number, the number of wishes is on the
order of 100,000, which should be sufficient for all intents and
purposes. The number of wishes grows exponentially with the threshold,
so there aren't really and practical limits.
> Say we start the engine with 80 holy!, 10 PoX and 6 wishes.
It's somewhat cumbersome to integrate the holy water into the
mathematics, so I hope it's okay if I ignore it for now. If we convert
the six wishes into 18 potions of exchange, that means we start of with
28 potions of exchange. From that niveau, the chance of ever reaching
the > 250 niveau is 43.0079%. The chance of ever reaching the > 500
niveau is 42.7926%.
That basically means that from this starting point the chance of getting
the engine running is about one half, and if you do get it running, you
get all the wishes you want (at least if you pick 500, not 250 as your
threshold).
> Perhaps if one has extra ring piles a stockpile of RoDS should also be
> build to make sure that wishes does not need to be spend right away.
That is an interesting idea. This is something an Archmage cannot easily do.
Malte
Przemyslaw Brojewski
:> Perhaps if one has extra ring piles a stockpile of RoDS should also be
:> build to make sure that wishes does not need to be spend right away.
: That is an interesting idea. This is something an Archmage cannot easily do.
Bzzt! wrong. I have an archmage who wished for potions
of exchange till he had 140 RoDS in his backpack. (rings came from
buying out the casino few times) Just scumish, not hard.
And since elemental gauntlets halve the negative effects of
the mana orb, that means one needs to wish for SoCR only every second
use of it. And some numbers:
Whish cost: 3000pp. From book with great book caster: 4500. But,
with mana battery corruption, it's 20% cheaper: 3600. In a magically
charged room, another 20% off: 2880. I had crowned neutral archmage,
so worked with that value. A silvernight/darknight would make it even
easier. On the other hand, my wishing session lasted three days.
A two hundred wands, clothes of power and Trident of the Red Rooster
allowed me to start. Soon had enough PP to cast two wishes and later 3
per one use of mana orb. In the end I had 11700pp and could cast four
wishes per one use of mana orb. That means 6 free wishes, one for SoCR
and one for holy water, total 8. Nothing is hard for an archmage.
brojek.
Malte Helmert
I knew someone would bring this up. There's always one. ;-)
Of course, with an unlimited number of wishes you can do basically
anything, including wishing for brass rings and potions of exchange.
Just like if you've got the ring search engine going, you can switch to
casting Wish from the book by getting the necessary things from wishes
(lots of power points, a few castings of Wish, a spellbook of Wish). The
difference is that the ring engine gives you that feature for free. I'm
not saying that's an enormously important feature; it's just nice.
Malte
Przemyslaw Brojewski
Yeah, that's why I bothered with my archmage to get those rings.
After inserting mana orb into it's anomaly, casting wish (from book
or not) becomes impractical -- you just cannot get PP back to max
in a reasonable time.
brojek.
DavidByron
I think he simply meant that getting wishes from a spell usually requires
considerable set-up and being on a special day / special room. Stuff to
handle getting the PPs needed. Therefore the wishes must be used as they
are made, and are not storable.
> I have an archmage who wished for potions of exchange
But you can't wish for =oDS directly so you have to essentially use the
non-Wixard engine
> Whish cost: 3000pp. From book with great book caster: 4500. But,
> with mana battery corruption, it's 20% cheaper: 3600. In a magically
> charged room, another 20% off: 2880. I had crowned neutral archmage,
> so worked with that value. A silvernight/darknight would make it even
> easier. On the other hand, my wishing session lasted three days.
> A two hundred wands, clothes of power and Trident of the Red Rooster
> allowed me to start. Soon had enough PP to cast two wishes and later 3
> per one use of mana orb. In the end I had 11700pp and could cast four
> wishes per one use of mana orb. That means 6 free wishes, one for SoCR
> and one for holy water, total 8. Nothing is hard for an archmage.
Why use the corrupting orb when you could just dump 100 million
on an altar and pray to replenish your PPs? Also why did it take
2 days just to cast Wish a few times? I think using the wands and the
mana battery corruption seems more hard work than you need too.
Why not just cast from health? Then you only need about 600 hp and
1000 pp which is really easy to get with a few boost potions as
long as you are level 50 already.
Carl Chisholm
> Bzzt! wrong. I have an archmage who wished for potions
> of exchange till he had 140 RoDS in his backpack. (rings came from
> buying out the casino few times) Just scumish, not hard.
> And since elemental gauntlets halve the negative effects of
> the mana orb,
They do? I don't remember seeing that documented anywhere. Does it for all
the Orbs, or what?
Igor D. WonderLlama
I thought that you could not sucessfully wish for spellbooks of Wish.
Malte Helmert
>>I knew someone would bring this up. There's always one. ;-)
>>
>>Of course, with an unlimited number of wishes you can do basically
>>anything, including wishing for brass rings and potions of exchange.
>>Just like if you've got the ring search engine going, you can switch to
>>casting Wish from the book by getting the necessary things from wishes
>>(lots of power points, a few castings of Wish, a spellbook of Wish). The
>>difference is that the ring engine gives you that feature for free. I'm
>>not saying that's an enormously important feature; it's just nice.
>
> I thought that you could not sucessfully wish for spellbooks of Wish.
If that's the case, just wish for spellbook of Create Item. Even the
black tome of you-know-who would be good enough in this case. I agree
that this is cumbersome. What I'm saying is that, once you've got a wish
engine buzzing, you can do basically everything you want to do. That's
the whole point of a wish engine after all.
Malte
Arturus Magi
Date: 2 Apr 2004 07:53:46 -0800
>I thought that you could not sucessfully wish for spellbooks of Wish.
You can, however, wish for spellbookss until you get a spellbook of Wish. You
just can't wish specifically for the spellbook of Wish, like you can with the
other spellbooks.
DavidByron
....
....
nho...@aol.com (Arturus Magi) wrote in message news:<20040404235933...@mb-m15.aol.com>...
Wouldn't that take forever pretty much? I'm assuming the Wish
spellbook is available even at level 1 (just unlikely) but this may
not be true. It would have to be 1,000:1 against or more.
Malte Helmert
> spoiler space
>
>
>
>
>
> ....
>
>
>
>
>
> ....
>
>
>
>
>
>>You can, however, wish for spellbookss until you get a spellbook of Wish. You
>>just can't wish specifically for the spellbook of Wish, like you can with the
>>other spellbooks.
>
> Wouldn't that take forever pretty much? I'm assuming the Wish
> spellbook is available even at level 1 (just unlikely) but this may
> not be true. It would have to be 1,000:1 against or more.
I don't think it'd be 1000:1, judging from past experience.
Cannot test this at the moment, though.
Malte