Nightbird's Loot Distribution Theory

Status
Joined
Feb 8, 2006
Posts
3,841
Location
Chicago, USA
Society
Project-Asylum Corp.
Avatar Name
Revan The Nightbird Evenstar
__My theory is based on time and short "cycles". I won't reveal everything here but I want to leave some clues for fun and perhaps see if anyone else has caught the same patterns I have.

__My main assumption is that deposit locations are not pre-generated and stored in a database (when a bomb is droped, a search is done in the database to find if there are any deposits within the finder radius).

__First, what is a cycle? Well, a cycle is just a word that I thought was fitting to describe this phenomenon.

For those with a math background, what I think governs mining loot is a sine (or cosine, doesn't really matter which) function based on time. It goes from 1 to 0 to -1 to 0 then back to 1 every X seconds. If you drop a bomb/probe at a time Y, and sin(Y) is between 0.5 or 1 (I call this the loot window, more on this later), then you find a claim. Why is this a cycle then? Because sin(Y)=sin(Y+X)=sin(Y+2X)=sin(Y+3X) etc etc... so if you know X, and get a hit on Y, then you can get hit after hit after hit.​

To put this more simply, say the cycle is 12 seconds long.

1, 2, 3, 4, 5, 6, 7, 8, 9, [10, 11, 12], 1, 2, 3, 4, 5, 6... etc

__Then if you drop a bomb/probe on seconds 1-10, you don't hit anything. If you drop on seconds 10-12 (the loot window), you find something. Sounds good?

__The difficulty then lies in finding what the total length of the cycle is, (I used 12 seconds just as an example). Finding the length requires a very methodical method. To make it worse, the cycle length is always changing (albet slowly), so you have to adapt on the fly.

__Why do I think this is a reasonable method for MA to control loot? That is because of the loot window. The loot window size, like in the example above, determines the returns of a person dropping randomly. If the window was [6 to 12], then that person would hit alot and profit, if the window was [11 to 12], then that person is gonna cry if he persists too long. With the change of a single number, by making the window [0 to 1] or [.5 to 1] or [.75 to 1], MA can create good and bad periods with no formula changes or server load.

__To penalize overlapping, every drop is blacklisted by the server (it records the drop location, the radius from that location of the find, or the radius of the finder if nothing is found). Dropping too close to a blacklisted spot results in a penality (for example, a window of [0,1] becomes [0.75,1]), with the penality being more severe the closer you are to a blacklisted spot. Blacklisted spots expire after a period of time and there is a limit to the list of blacklisted spots - when full, a new one will bump the oldest one off.

__Finds that are generated by a drop are stored in a database, making it possible to find unclaimed resources.

last notes:

1. cycle length is not likely a whole number in seconds
2. I did not reveal everything such as the method here, and will not for obvious reasons
3. my method is hardly perfect anyways, it's very hard adapting to cycle changes but I can boast of 9 finds with no misses in between and 80+ finds with up to 2 misses in between.
4. I won't prove my method to anyone for the same reason that I won't share my method, as there is some information that I've refrained from divulging that would become obvious should someone witness my method.
5. Cycle lengths are short, under 1 minute for certain.
6. Finding out the system is fun, and I am still having fun refining my system.
7. This theory can also be applied to hunting, but cycles changes occur much faster there so hunt with teammates on the same mob to kill and loot very quickly might help reveal the cycles because it gives more data points
8. Accuracy to the 0.1 seconds is enough (and necessary)
9. I don't do this for profit, but for the challenge of solving a puzzle. I won't say no to profit though :)
 
Last edited:
Two things I wanted to add: I didn't state the sine formula correctly but anyone understanding of math should understand what I meant and I didn't want to type a messy formula, especially since I don't know the ascii code for pi.

Second of all, this theory is more likely than the rest due to occam's razor. It is the simpliest explanation possible for how EU loot works, the good and bad periods, the least server load, random, and yet depends on user choice.
 
Very well thought out, and well presented. +rep
 
Second of all, this theory is more likely than the rest due to occam's razor. It is the simpliest explanation possible for how EU loot works, the good and bad periods, the least server load, random, and yet depends on user choice.


Hmmm...true....perhaps we see more globals during ;periods when lots of people are on not simply because there is more loot in the pool, but with so many people hunting at the same time....more chances for a loot-window to "pop" at the right time.

An interesting premise and a soc mate and I did notice soemthing along these lines when hunting once. It seemd when a big global or HOF hit at the same time we popped a mob, loot was significantly better. The trouble is, this was always retrospectively observed...we didn't try to time it.

There was a thread not too long ago about soem crafting cycles and it showed a sinuous pattern over days. Perhaps HOF's are the "perfect storm" of a resonance when these big and little cycles meet...like biorythms. If they are based upon prime numbers a nice way to randomize that which can not technically be made random.

I wonder if they re-seed the waves from time?

Interesting idea, to be sure...
 
It would explain why I usually see no globals for half an hour then 3 come along at once.........just like buses ;)
 
It would explain why I usually see no globals for half an hour then 3 come along at once.........just like buses ;)

Ever see 4-5 people disconnect at the same time? Server only sends that log off update every x seconds and sometimes they get buched up, server to busy to send. Same thing can happen with globals.

Nightbird, I think your pretty close but like you pointed out, the cycles are not constant- perhaps like a PID with a constantly changing derivetive :) I'm not much of a miner but I agree with you, resources no longer apear to be 'in the ground' but generated at the time of the bombing.

Fuga thought of the day hehe
 
Right, I was about to mention that my theory doesn't cover globals. It does make rattax's spiral theory possible under special cases, though it not often.
 
So if your theory is correct, once you find out the time interval wouldn't you be able to just mine the same spot forever?
 
I have noticed that you often loot your "own" peds in globals. Lets say ur down 150ped on a hunt, at the end you recieve a 100 ped global just to compensate some of your losses.

I think there is a personal "loot pool" and a global "loot pool". Global is for building ubers and such, and personal is for compensating and loot.
 
So if your theory is correct, once you find out the time interval wouldn't you be able to just mine the same spot forever?

I hurried through this and made a mistake. All drops are blacklisted. For each location, the radius of the find if any or the max radius of the finder is recorded. The server can do a quick calculation using these two pieces of information to calculate penality

I have noticed that you often loot your "own" peds in globals. Lets say ur down 150ped on a hunt, at the end you recieve a 100 ped global just to compensate some of your losses.

I think there is a personal "loot pool" and a global "loot pool". Global is for building ubers and such, and personal is for compensating and loot.

I've noticed this too. The more hits I get, the smaller the amounts I get per deposit. My theory doesn't over loot amount though, as it is too hard to judge variables and no way to control them.
 
Doubt we are talking seconds here. If it is coded with timebased-randomness this cycle wouldnt be more than a second at most, containing many numbers. This loot thing probably has variables, one based on the mob, one based on lootpool, one based luck, one based on droprate, etc, etc.

It will never be possible to figure this out and use it, there are to many variables in it, and the calculations are probably done quicker than a human can react.

This randomness would also be affected by stuff like people online, your isp´s load on bandwidth, your connection, your pc´s current use of ram and cpu aswell as the actual servers and so on.

Impossible to gain anything from it even though we would have the facts in front of our eyes.

Though, interesting theory.
 
Doubt we are talking seconds here. If it is coded with timebased-randomness this cycle wouldnt be more than a second at most, containing many numbers. This loot thing probably has variables, one based on the mob, one based on lootpool, one based luck, one based on droprate, etc, etc.

It will never be possible to figure this out and use it, there are to many variables in it, and the calculations are probably done quicker than a human can react.

This randomness would also be affected by stuff like people online, your isp´s load on bandwidth, your connection, your pc´s current use of ram and cpu aswell as the actual servers and so on.

Impossible to gain anything from it even though we would have the facts in front of our eyes.

Though, interesting theory.

In order to make it possible for two people mining in the same area (but not overlapping eachother) get different loot results, the cycle must be slow so that the behavior pattern of one person might be better suited to that cycle length and thus two people mining at the same site can come out with different returns. If cycles were too fast, everyone would get the same average return and also it becomes very easy to avoid the bad looting periods since one can establish the average loot at the time very quickly.

Hence the difference between player behavior driven randomness and complete randomness.



P.S. To get the system I have now, I started with thinking how I would design the system so that it appears random, has good and bad periods, is deceptively simple, and yet difficult to notice patterns in. Since finder reload time and run speed limits the number of data points collected in a short period of time, I came to several possible systems and finally after extensive testing arrive to the conclusion that this one is correct.
 
Doubt we are talking seconds here. If it is coded with timebased-randomness this cycle wouldnt be more than a second at most, containing many numbers. This loot thing probably has variables, one based on the mob, one based on lootpool, one based luck, one based on droprate, etc, etc.

It will never be possible to figure this out and use it, there are to many variables in it, and the calculations are probably done quicker than a human can react.

This randomness would also be affected by stuff like people online, your isp´s load on bandwidth, your connection, your pc´s current use of ram and cpu aswell as the actual servers and so on.

Impossible to gain anything from it even though we would have the facts in front of our eyes.

Though, interesting theory.

Ahhh but it was possible at one time :)
 
I never tried it but has anyone ever probed/bombed the same spot every 5 or 10 minutes to find the respawn. If so how long it take.
 
Here is a support case answer I have recieved a long time ag. Might help on this cycle business, this is a response I got for having shitty loots for 3 months, losing like 60% on a hunt every time.

"19 Aug 2006 MindArk Support:
Hi,
We do understand your disappointment. However, the Entropia Universe is designed to be dynamic and this affects all areas, including the loot system. At some times the loot can be lower and at others higher. By killing a bigger mob with better health your chances to get better loot generally increase, but there is no guarantee. Make sure that your equipment matches your skills, since using equipment superior to your skills can cause higher costs. You may also consider asking other participants about good hunting areas and what mobs to hunt. Perhaps you should also consider trying out some of the other professions within the Entropia Universe?

As you probably understand, we cannot give you any more detailed information about the core dynamics of the Entropia Universe, but you will probably have more luck next time. However, since this seems to be an occurring issue for you, we will forward it to the development team. Should any irregularities be found the will be adjusted automatically as soon as possible. "
 
There is one thing that makes me wonder in mining. When i find a resource, I can usually tell as soon as the finder pops up with the Searching... message. I have dropped over 100k probes on this PC and i can tell a few milisecond "slugging" occurs whenever a resource is going to be found. This is hardly noticeable on Calypso but much easier to tell on CND (probably as the deposits are larger).

Now, during the bad runs, I can get several "false alarms" when I notice the delay but end up finding nothing. This serves me as a good way to recognize the lower periods of the loot cycle.

Although I believe most of the uberfinds occur during these "low periods" (often resulting in mining societies chat being spammed with "FFS we paid for that!"), I preffer not to, using a slot machine terminology, play the tight one and go after a jackpot. I rather play a nice loose one that gives me more or less my money back and let my skills talk markup.
 
So let's say everyone in the universe uses this theory 100% correctly , and lets say the loot pools exists then doesnt someone have to make a loss and someone else profit , but if everyone uses the theory 100% correctly , who makes profit and who makes loss , help me out here ;)
 
So let's say everyone in the universe uses this theory 100% correctly , and lets say the loot pools exists then doesnt someone have to make a loss and someone else profit , but if everyone uses the theory 100% correctly , who makes profit and who makes loss , help me out here ;)

MA makes profit
People who withdraw make profit
Rest makes loss

when you factor the time spent / opportunity costs

MA makes profit
Rest makes loss
 
As long as your having fun, your the winner :)
 
This is incredibly interesting. Nightbird seems very confident so I know there's more to it. It's a very smart idea. Could one look at the globals on the chat window to help find when the "high period" would be? It would be neat to record the times of globals then graph the results to see if it ended up like a sin or cos graph. + rep for such an excellent idea, thank you Nightbird. Please reveal more secrets to me or taking me out mining some time, lol ;)

Could you also tell me if you are making profit or at least breaking even with this new theory?
 
This is incredibly interesting. Nightbird seems very confident so I know there's more to it. It's a very smart idea. Could one look at the globals on the chat window to help find when the "high period" would be? It would be neat to record the times of globals then graph the results to see if it ended up like a sin or cos graph. + rep for such an excellent idea, thank you Nightbird. Please reveal more secrets to me or taking me out mining some time, lol ;)

Could you also tell me if you are making profit or at least breaking even with this new theory?

I profit 25% on average. So if I leave and drop 200peds of bombs, I return with 250peds of goods (including markup value). The problem with my system is that I can't avoid rebombing places where others have been to so there are noticable "lulls" when I can't find the cycle no matter how hard I try (someone probably carpet bombed there :p). Also, it might be due to the "personal loot" theory, but I just can't seem to global :/ I must constantly keep my personal loot depleted lol. The only times I global mining seem to be after I have blown 200peds away hunting warriors, and even then the global sizes only match my hunting losses. Still, this isn't conclusive and I'm still hoping for a mining uber.

Here is a support case answer I have recieved a long time ag. Might help on this cycle business, this is a response I got for having shitty loots for 3 months, losing like 60% on a hunt every time.

"19 Aug 2006 MindArk Support:
Hi,
We do understand your disappointment. However, the Entropia Universe is designed to be dynamic and this affects all areas, including the loot system. At some times the loot can be lower and at others higher. By killing a bigger mob with better health your chances to get better loot generally increase, but there is no guarantee. Make sure that your equipment matches your skills, since using equipment superior to your skills can cause higher costs. You may also consider asking other participants about good hunting areas and what mobs to hunt. Perhaps you should also consider trying out some of the other professions within the Entropia Universe?

As you probably understand, we cannot give you any more detailed information about the core dynamics of the Entropia Universe, but you will probably have more luck next time. However, since this seems to be an occurring issue for you, we will forward it to the development team. Should any irregularities be found the will be adjusted automatically as soon as possible. "

This cycle and my cycle are completely different. This is actually the loot "window" I was talking about. When the window is large, it's a good period and vice versa. Read first post again if you need to.


p.s. One uber, I won't reveal whom because I haven't asked for his permission, confided in me that he is using this very same system and to his knowledge we two are the only ones who know. We both agreed that the hints here don't give too much away, though it does give people a stepping stone if they want to try their hand at cracking the system. The actual purpose of this thread was actually to see if anyone knew my system already and have them PM me, revealing themselves :laugh: I am glad that it worked and to know I am not the only one :)
 
Last edited:
Second of all, this theory is more likely than the rest due to occam's razor.

sorry, but how can you say this and yet completly dismiss the possiblity that resources are spawned in the ground already? there is no evidence to prove that deposits are avatar generated, so its a pretty massive assupmtion to make.

Not that i think the theory is wrong necessarily, just a logical faux pas there. I suspect that loot values might work on this basis.

btw, thread title should include "mining"
 
sorry, but how can you say this and yet completly dismiss the possiblity that resources are spawned in the ground already? there is no evidence to prove that deposits are avatar generated, so its a pretty massive assupmtion to make.

Not that i think the theory is wrong necessarily, just a logical faux pas there. I suspect that loot values might work on this basis.

btw, thread title should include "mining"

I think that because it is extremely server intensive to keep track of 10s millions of deposits in the ground, just as difficult as giving every mob an ID tag and keeping track of a gigantic loot table with all the loot pregenerated.

This is all IMHO though :) just a theory.

If all loot is computer generated anyways, would you

1. pre-generate it and keep it stored in a gigantic database?

or 2. generate it when it is asked for and skip the database?

Hence Occam's Razor
 
your razor is blunted by the observable fact that there is already a massive databse tracking the location of all the mobs. why not deposits too?

generating vs tracking of loot values is a different matter.
 
Interesting stuff Nightbird.

How does this work if I don't claim a find? What I mean is sometimes I have dropped a bomb and found something. Then by mistake I have equipped something else, thus dismissing my finder before planting the claim rod. But if I re-equip my finder and drop another bomb I find the same sized claim the same distance and direction away from me.
 
your razor is blunted by the observable fact that there is already a massive databse tracking the location of all the mobs. why not deposits too?

generating vs tracking of loot values is a different matter.

Because it is necessary to track the location of mobs, as they have to be visible.

Interesting stuff Nightbird.

How does this work if I don't claim a find? What I mean is sometimes I have dropped a bomb and found something. Then by mistake I have equipped something else, thus dismissing my finder before planting the claim rod. But if I re-equip my finder and drop another bomb I find the same sized claim the same distance and direction away from me.


The theory says how the claims are generated, not how it's recorded but it is obviously :) in case people die before they make it to a claim.
 
Because it is necessary to track the location of mobs, as they have to be visible.

fair enough. heres a couple of other considerations.
Why develop a second system for deposit locations when you reuse and tweak the mob system?
Wouldn't generating a claim (or miss) every drop not impose a load on the server, most likely the same as looking up in a database?
 
fair enough. heres a couple of other considerations.
Why develop a second system for deposit locations when you reuse and tweak the mob system?
Wouldn't generating a claim (or miss) every drop not impose a load on the server, most likely the same as looking up in a database?

First of all, I remind everyone again that I don't know HOW it works precisely, I'm just working from my observations and finding a theory that best fits it.

Why develop a second system for mining? Actually it's also used for hunting loots. Ever notice that all loot on a mob suddenly stops right before or after an uber? If loot was recorded in a mob when it spawns, then loot wouldn't stop immediately, because you have to kill all the good looting mobs first and then the bad looting mobs spawn. I consider this definitive proof that loot is not pre-generated
 
Why develop a second system for mining? ...

Indeed. Or crafting for that matter. Big difference between crafting compared to hunting&mining is that it takes place at fixed point -- construction machine that is. Yet, they all are simple resource-in-resource-out routines. Why don't use set of get-whats-paid-curves?

About that "new loot pool" crap -- timers; would you build system like this on something that is not protected against statistical anomalities, players being "smart" enough, and (worst of all) everything can be simulated and/or leaked by ppl who developed it? To get ATH just set your clock to +20 years to tell the system you haven't got any loot in past 20 years!? (yes, i got your point. ignore this statement) Paleeze...

Hunting is the way to notice that time/cycle runs faster at places with many consumers. Any observative hunter tells you that. Hof/ath table tells you that.

Thinking ppl are more interesting that non thinking pp, tho :)
 
loots and mobs/deposits are two different things. unless you make the assumption that a deposit doesnt exist until its loot is generated. which maybe you are, and there's nothing to say that you are wrong, though i have reservations about that aspect of the theory. Loot value being determined at the time of looting is quite clear, as loot lag proves.

i was really highlighting the fact that a very large assumption is made at the very beginning, leading to a logical flaw.
 
Status
Back
Top