Quantifying Treasure Hunter
Darren Grey
not the Treasure Hunter talent was really worth it. I'm not aware of
any specific research done into this, so I decided to do some myself.
Results are potentially spoily though...
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Okay, so the Guidebook says 50% more items, but only as an estimate. I
wanted to check things out more thoroughly. Nothing too major, because
although I was bored I wasn't completely crazy, so perhaps people can
try testing this out with experimentation on a larger scale.
First experiment was with a hurthling paladin with Alert and Miser
talents. Went straight to ID, got to the cusp of level 3, then saved.
Backed up the save and then with one copy took Treasure Hunter at level
3, with the other took Long Stride. So, both were identical beyond the
Treasure Hutner talent. With each of them I went to the down staircase
on ID 1 and kept going up and down, recording how many items were found
in the room each time (without killing monsters). Gold was not counted
as an item find (since I wasn't sure if Treasure Hunter influenced this
instead of Miser). Both characters ended up at level 6 by the end.
Now there is some room for error in this - different room sizes and
such, as well as random error - but I figured that averaged out over a
large data size it would be fairly clear what the general effect of
Treasure Hunter was. Also, the copy who took the TH talent had two
early reductions in Pe due to room effects - the other had no
reductions.
Results then: The character with Treasure Hunter found 94 items in 530
rooms - 17.74%.
The character without the talent found 118 items in 620 rooms - 19.03%.
It should be obvious that Treasure Hunter has no significant effect on
items generated when you enter a new dungeon level - if it was on the
order of 50% this would have shown more clearly. The fact that the
char with the TH talent found less might be due to the Pe reductions or
just random differences. The results show no obvious change in the
rate of item finds as time (and levels) progressed, so I didn't think
to monitor this factor in my next experiment.
I was somewhat surprised by this result, so I determined to then check
drops by monsters when you kill them. I got another hurthling paladin
to level 2 in the ID, and repeated the same split. This time each copy
was going up and down between ID1 and 2, killing any monsters in the
rooms and recording their drops. I categorised drops into Items,
Corpses and Gold for the character with Treasure Hunter. After doing
this I figured with the non-TH character that I should seperate out
"standard drops" - things like short bows dropped by kobolds and crude
knives dropped by orcs - in case these weren't changed by Treasure
Hunter and made a big impact on the numbers. As it turned out there
seemed to be significantly fewer of these in the second round, implying
that Treaasure Hunter was equally at work on these as on normal items.
Also, the Treasure Hunter character had an early drop of 3 Pe whilst
the other saw no change - this may mean the Treasure Hunter should have
found more in equal conditions.
For each I took a sample set of 400 kills. Each character ended up at
level 9.
Treasure Hunter: 98 items, 58 corpses, 20 piles of gold
Non-TH: 57 items, 59 corpses, 6 piles of gold
Instantly striking is how close the corpse counts were - obviously
Treasure Hunter has no impact there. Also obvious is that gold is
affected by it (both characters started with Miser), probably just
rated as another type of item drop. If we group together the figures
then we see that Treasure Hunter had 29.50% drops, Non-TH had 15.75%.
It seems that the effects of Treasure Hunter on item drops by monsters
double what you'd normally get, making it more powerful than previously
thought. I think there can certainly be now doubt that this is the
most valuable talent.
Of course like I said there is room for error here. I would like to
see this repeated by someone with a gremlin bomb and a lot of patience
- figures on the order of thousands would give much more definite
results. Also, this was restricted to a low level character on low
danger levels. From what I saw it seemed that item finds definitely
increased significantly as the character progressed in levels (unlike
with the previous sampling). This could affect how powerful Treasure
Hunter is at later levels.
Some other interesting little observations:
-No monster ever dropped more than one item, unless the second item was
a corpse or a "standard item". When 30% of monsters drop items you'd
expect some of them to be dropping 2 or even 3. It seems that the
number of potential items is possibly restricted by monster type (we
know full well that dragons can drop more).
-For both the killer characters I increased their food preservation by
about 10 points around halfway through, and it quickly produced very
noticable changes. More research needs to be done on this perhaps, but
it seems that even small changes (50 to 60) can have a significant
effect.
-What does Miser do? Increase the amount of gold you receive when a
pile is generated? But that's limited by danger level anyway. Seems
pretty useless (though we knew that anyway I guess).
Anyway, I'm done with that boring stuff now - hopefully someone finds
this of interest.
--
Darren Grey
Soira
Darren Grey raše:
> Someone recently questioned on the Hall of Fame forum as to whether or
> not the Treasure Hunter talent was really worth it. I
>
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
> .
>
>
> Non-TH: 57 items, 59 corpses, 6 piles of gold
>
>
e
x
t
r
e
m
e
s
p
o
i
l
e
r
s
p
a
c
e
my archeological dig into this group's threads revieled old thread
about code diving.
it was stated what TH talent gives you 12.5 extra drops then monsters
are killed and do not drop items.
taking your results:
400 kills - non TH 63 items including gold -- 15.75 procent
TH 118 items including gold - 29.5 procent.
fits theory just fine.
of course chances of item drop depends in monters in question. also
excluding drops and ones who droped more then one would be nice. but
that's is a bit too much for me.
I had plans of killing 1k of gremlins with and without treasure hunter,
but i really don't see a point to do so.
as i said in Hall of Fame froum: "it depends on how you like to count
things up."
soira
lochok
>
> Darren Grey raše:
>> Someone recently questioned on the Hall of Fame forum as to whether or
>> not the Treasure Hunter talent was really worth it. I
>>
<SNIP>
> my archeological dig into this group's threads revieled old thread
> about code diving.
> it was stated what TH talent gives you 12.5 extra drops then monsters
> are killed and do not drop items.
<SNIP>
> soira
Results: Good
Using the 'CD' word: Bad
A brief reminder that the Creator promised to smite the Divers!
Out of respect for TB, who has said that he doesn't wan't CDing going on,
please don't really use those results or code dive...
Lochok
Darren Grey
> Results: Good
> Using the 'CD' word: Bad
>
> A brief reminder that the Creator promised to smite the Divers!
>
> Out of respect for TB, who has said that he doesn't wan't CDing going on,
> please don't really use those results or code dive...
It's generally accepted that code-diving results from before TB said to
stop are considered open game material. Much of Andy's Guidebook
contains this sort of information, as well as a lot of "general
knowledge" of the game. Also, to be honest, quite a lot of code-dived
info is very hard to interpret correctly. I mean what is "12.5 extra
drops" supposed to mean out of context?
--
Darren Grey
The Wanderer
> lochok wrote:
>
>> Results: Good
>> Using the 'CD' word: Bad
>>
>> A brief reminder that the Creator promised to smite the Divers!
>>
>> Out of respect for TB, who has said that he doesn't wan't CDing
>> going on, please don't really use those results or code dive...
>
> It's generally accepted that code-diving results from before TB said
> to stop are considered open game material.
I wouldn't be so sure about that.
> Much of Andy's Guidebook contains this sort of information, as well
> as a lot of "general knowledge" of the game. Also, to be honest,
> quite a lot of code-dived info is very hard to interpret correctly. I
> mean what is "12.5 extra drops" supposed to mean out of context?
IIRC it was expressed more clearly than that at the time of the original
discussion. I read it as meaning something like "if the regular check
for whether an item is dropped fails, there is a 12.5% chance that an
item will be dropped anyway".
--
The Wanderer
Warning: Simply because I argue an issue does not mean I agree with any
side of it.
Secrecy is the beginning of tyranny.