Posting Freak
Posts: 6,070
Threads: 229
Joined: 2008-07
2009-04-19, 11:35 AM
(This post was last modified: 2009-04-19, 11:39 AM by KajitiSouls.)
Personally I think this might be a waste of time, although I have a few thoughts to contribute.
One way to get a good sense of how many monsters you killed at a training session, especially in places where you killed mix spawns such as Skeles, is to collect all the etc. items and count them up at the end. Since monsters drop their etc. items at a reliable rate, we can use this number to estimate how many kills we've made during a particular training session.
If we find an item, great! If not, then move on. To get an accurate picture of a rare item drop would require massive amounts of kills. You could probably achieve this with stuff such as ores and whatnot, but with equips and scrolls, hell no. You seem to understand this though =P
Oh yeah, at one point I was taking boss drop statistics. The only data of good promise is the one from Papulatus. Papulatus, on average, drops an equipment every kill, and drops a skill book every 6 kills. There may or may not have been a slight bias in the equipment drop rates, but I do suspect that it likes to drop Red Kattes.
I've included a log of the drops. No date has been assigned, although it began a few patches before Pirates, and ended a patch or two after Pirates.
Posting Freak
Posts: 3,185
Threads: 76
Joined: 2008-12
Gender: Male
Sexual Orientation: Bi
IGN: y0y0y0y0shi0
Server: Mardia
Level: 98
Job: Priest
Guild: Fateless
I'll start helping too, record everything I can, and will help with killing the designated monster. It won't be much, but it's better than nothing, right? 
Posting Freak
Posts: 8,478
Threads: 128
Joined: 2008-07
Gender: Neuter
Country Flag: canada
IGN: Oooh
Server: Bera
Job: Empress
Farm: Stereo
2009-04-19, 08:09 PM
(This post was last modified: 2009-04-19, 08:12 PM by Stereo.)
Cyanne Wrote:Chance of equipment dropping is generally based on the relation between the equipment's level req and the monster's level, with weapons, stars, and scrolls dropping much less frequently of course.
I'd tend to agree with this in general but I'm not sure if that's just how they happened to set the droprates or if it's automatic.
For example, Pan Lids drop from Star Pixies - but in my experience they're no more common than any of the other drops. Black Bandanas drop at Wraiths (well, used to) and they were rarer than Whips. Where Blue and Yellow Umbrellas drop (I think yellow come from Death Teddies) they're no more common than the regular gear.
On the other hand, if you kill low level mobs, you find way more equips. At one point early in Bera I killed about 10,000 Slimes on 2x drops, and got on the order of 70-80 equips. More recently I killed about... 100,000 Red Wyverns. And I found about 40 equips.
In order to get a good idea of drop rates you'd probably want to find a total of around 15-20 minimum of an item. At that point the individual luck is dropping off and it's more reliable. For some bosses (Papu) that is not so bad cause he drops equips pretty frequently and can spawn proportional to the number of people hunting him, for regular mobs it could mean 1,000,000+ kills.
Posting Freak
Posts: 1,069
Threads: 14
Joined: 2009-03
Gender: Male
Country Flag: canada
Server: Khaini
Job: Various
I have an idea about drop rates. Since the core routines of server software have to run say a million times a second, the drop rate variables are probably based on 16-bit integers. So rates would all be multiples of 1/(2^16) or 1/65536 or maybe 1/32768. So the rarest of the rare, chaos scrolls, nependeath honey and so on would be 1/65536, somewhat less rare scrolls 2/65536 or 3/65536, and so on.
Also, true random number approximations are computationally intensive. So more likely they use lookup tables of pseudo random sequences. Or maybe not so random. Many times I've not seen some item for several training sessions, then get a few of the item within a few minutes. I'm well aware that random doesn't always appear random, but still it makes you wonder.
Harrisonized
Unregistered
Yesterday, while hunting snowmen, somehow I noticed that the more damage I take from the snowman, the more drops I get (even though it's mostly useless ect). For example, if I just teleport into its body and stay there for the fight, a long row of items comes out, rather than standing from a distance and telecasting to avoid the damage.
Can someone else confirm this?
Posting Freak
Posts: 1,235
Threads: 69
Joined: 2008-07
Harrisonized Wrote:Yesterday, while hunting snowmen, somehow I noticed that the more damage I take from the snowman, the more drops I get (even though it's mostly useless ect). For example, if I just teleport into its body and stay there for the fight, a long row of items comes out, rather than standing from a distance and telecasting to avoid the damage.
Can someone else confirm this?
I can try!
...
But for the PCH thing...
We see a rarity difference between scrolls/equip types.
And I've heard forever that the lower the level of equip the higher the prob of dropping.
Posting Freak
Posts: 7,905
Threads: 472
Joined: 2008-07
Chompy Wrote:I can try!
...
But for the PCH thing...
We see a rarity difference between scrolls/equip types.
And I've heard forever that the lower the level of equip the higher the prob of dropping.
I've seen more Eagle Crows than anything else from Zombie Mushrooms, and those are fairly high level equips for that level of a monster.
Administrator
Posts: 17,191
Threads: 2,153
Joined: 2008-09
Gender: Male
Sexual Orientation: Gay
IGN: Aesilyn
Server: Mardia
Level: 200
Job: I/L ArchMage
Guild: Animus
MasPan Wrote:I would need someone who is pretty good at statistics to aid in compiling the information and obtaining a reasonable drop rate with a standard deviation dedicated to compiling everything once a week. Maybe Eos to help me create the database itself, as I don't have much skill in coding, but I can put it together by hand if necessary.
While I appreciate the intention behind this, I just don't find it feasible.
Aside from Nexon's ability to change the drop table at will, and the sheer amount of accidental human error that can bias the data, I just don't think you can accurately gain a decent enough static to noise ratio due to difference in situation (single drop, double drop, and now 1.5, triple and quad drop due to family bonuses).
You'd never be able to get a sample size large enough to normalize all the potential biases into a sense of normalcy - It's counting grains of sand with tweezers, and ultimately how useful is it going to be? Does it matter if what you want drops 1 in 30,000 or 1 in 50,000? If it's what you want/need, you're going to continue killing until you get it regardless of how long it takes.
It's not having what you want - It's wanting what you've got.
|
