Random Drops

You've voiced something that is an on-going, never-ending issue here, with lots of posts over the years. "Could, should, would, can, shall, will" generally fall on deaf ears. Your frustration is shared.
Try picturing polite, dinner conversation, while the Titanic is taking on water. The crew follows their orders, not the passengers.
Players deal with it in different ways. Good Hunting!

 

my forum is messing up,i just seen this comment on speakers corner,"random drops"posted seconds ago.
or is it just me?
but i dont see his comment on this post,only in speakers corner,,anyone one else notice this?
yes i know,should be in tech,sorry

matami said: ↑
Sure they wont reveal how luck is generated in this game but maybe if we complain enough they will look into it could there be something that have messed it up when they have updated the game. At least i dont buy that if we have totally random luck and throw coin 100 times and get 95 times heads like it seems to be for some players now. I know that was bad example but you understand my point

 

I guess the mods got tired of the off topic posting, but saw it as constructive so they didn't just delete it like they sometimes do. Rather, they took the offending posts and put them all in a new thread.

 

I have created this thread and moved your posts here.
It is better solution than deleting two pages of off topic comments.

 

ok,i thought i was going crazy.

 

Sorry Mal, and thank you. It happens sometimes that the inevitable meandering moves us in an actually productive direction.

On topic: I built a reward process(loot table) a few decades ago using DB3 and BASIC, and I see many similar elements in DSO. First comes the kill which triggered 4 Rand 100 rolls that were stored in variables. The first variable looks at the 1st table and determines which subsequent table out of many is accessed. The second table/variable determined the item, the third determined the enchantments, and the 4th any special characteristics. It was for DnD and was a way of randomizing chest rewards and random drops. It was later modified for a couple of different table top games. A few extra RAND's and you have a matching system for DSO. In a newer language it would access a series of subroutines instead.

It is hard for me to take those glasses off when I look at loots in DSO. Mobs have a certain number of possible drops and a certain multiple, each is rolled separately with limiters. It's why destructor drops white items and a level2 archer in swerfield can drop a legendary and you get a pink every time you kill a HOD knight. Just change the Rand range and number of times it repeats.

It is also why I gaaafaaaa at some of the things people come up with to try and get a grip on why they have good or bad luck.

 

For people who are IT illiterate, PRGs are entirely unfathomable. I believe it was Arthur C. C. who said that any sufficiently advanced technology is indistinguishable from magic. To a caveman, a lighter would be pure magic because there is no way for him to understand how it works. Same goes for PRGs and people who haven't got the foggiest about programming 8)

 

"It's fricken magic" is as good an explanation as any when it comes to DSO. Even to those with knowledge.

 

It's interesting that you mentioned 100 rolls. Thanks for bringing that up.

Sometimes the drop pattern remind me of a 0-9 digit, 100-entry, random number table I memorized decades ago. (It was very useful in poker.)
The pattern would hold until the map cycled to a new instance, probably another tier of 20+ possibilities in the Parallel Worlds. DSO might be using slightly different base numbers (idea may not apply at all, or apply differently). The concept could have spawned from simpler models for the first maps of the game. One can almost track the gears changing over a 2 week period.

It would be nice to have number overlays embedded in my mind, but I never was that detailed in math applications. It took me 40 years to test basic Calculus. LOL

Any PhD's in math out there?

 

Here's what I envision for a standard mob. This is of course just a foundational example needing a few more levels to get a completed item, and the number ranges are of course not accurate, just spit-balling. A random 100 roll takes you to a new table. The Rand rolls are then modified by items that can affect drop rate. Example Random 100+3; for peeps wearing the cape.
00-40: no drop go no further.
41-70: coin drop, subsequent rand 100 for amount
71-85: Goto White item table for character level
86-90: Goto other item table, realm frag, gem, etc
91-95: Goto green item table for character level
96-97: Goto blue item table for character level
98: Goto Purple item table for character level
>=99: Goto Rare item table for character level

Example Item table for 99 roll: Roll second Random 100
00-87: No drop go no further
88-98: A different Leg item assigned to each number, follow on/goto Enchantment table for that item
>=99: Goto unique table

Example Enchantment table for 1 handed weapon:
Each enchant has a range for possible assignment like other tables. Repeat Random 100 as needed for Item level, 1 time for green, 4 times for purple. Math formulas after selection place enchant values in the correct possible range.
00-30:Crit rate: Random 100 + 150
31-50: Crit dmg: Random (100/.1) +3.5
55-60: Wpn Dmg %: Random (100/.2) + 15
etc
etc

This is of course me just spit-balling as I said. But it seems to match the things I see in the game. Dmg lines with a decimal places, long streaks of absolutely no drop. The ability to add a cape to the game that increases drop chance for reg mobs by 3%. That's simple just add 3 to the first roll. The fact that getting 4 enchant lines or even 3 the same is so hard because you would have to roll within the same range of numbers 3 or 4 times. Increasing Leg drops in PW's is 200% higher? OK no problem just change the range in the second table for leg items, make it 68-98. An entire server starts getting drops from Khally or Gorga? Yep, the variable from the first table or second is not being reset each time because of a programming mistake, or the roll has a modifier that was mistyped. Truth is IDK, but this is how I would write the code.

 

How many numbers there are between 0 and 1? Infinite. Therefore and the possibilities are infinite.
But what was I talking about is not about BP's programming.

In addition to this post by Baragain consider the fact that hardware like CPU for an example is using machine language (code) ... and as far as I know every individual machine language is processor specific depending on the architecture ... and it is defined by it's manufacturer. Even though the compiler should output the same machine code for a given architecture ... regardless of what architecture the compiler itself is running on ... these series of processes could affect the outcome.

 

Not totally true in a 32 or 64 bit computer. In a 64 bit machine using double floating point, there are "only" about 10^16 to 10^17 numbers between 0 and 1 if you are keeping the decimal point fixed... there are quite a few more if you let the decimal float, but it is still finite. For a 32 bit machine, there are "only" about 10^6 to 10^9 (again, more if you want to let the decimal point float.) Not trying to be pedantic or anything, just nerdy

 

I wasn't talking about programming ... it was basic math. Otherwise I would have typed it: "between 0.0 and 1.0"


When some people are talking about drops being random ... I always say: there is no "random" in computing.
When we generate random results in real life ... we often do so by flipping a coin, rolling a dice ... etc. Now... those events involve many physical variables like: momentum, gravity, air resistance ... etc... etc ... and they are almost impossible to predict and control ... thus they are random ... for all intents and purposes.
Computers can't take advantage of physical variables ... they know only about false or true and nothing in between ... therefore computers are incapable of generating random results ... instead ... they simulate randomness most often by using PRNG.
In other words computers are only doing what they have been told to do.

 

Trak is right, well current closest true-randomness is electrical noise in the atmosphere feeded to an operational amplifier (random dot org and wikipedia can explain it in details), so actually computer know what random is
but games just use time as seed to rng and a simple formulae like r = 1 +sqrt(previous_r)

as per drakensang few rules 's been established since the beta

monsters are represented in classes
each class got a list of item type that can be dropped with healing sphere at top and uniques at bottom

a monster must be marked to move in the list or you get 50% penalty

mark lasts 30 seconds since player took 25% off moster life

drop is affected by malus depending on how long in minutes is currently playing (anti addiction)
with minimum malus for just logged in, and maximum for playtime longer than 5 hours

 

Stellar Post! Incredibly high informative value. Thanks.

 

One theory is that the DSO changes the level of drop as needed. I'm inclined to agree with that.

Just today I wrote about the terrible drop in the crypts where it is impossible to get a good item but andermants it raining. Only decent dropa I get in hour after midnight (just before new day). I remember the story of how Mortis was too hard to kill and few week after and mediocre players was able to kill him. For last dragan event a player is able to pull a few hundred HPots first few days .... and then the drop level abruptly returned to normal. Whole January I personally could not get 19 frags from mortis priest .... and then I got 4 in one day.

I'm pretty sure that the DSO will never explain what determine the drop because thay, certainly not on a daily basis, very often (when required) change the parameters in order to maintain players interest in the game.