Menu
What's New
By:
Filters

Question: Punishing the Pattern Recognition of Players

Msturlese said:
Now MA is a listed company. so need to fgive bigger disclosures to market participants.
let's wait for next balance account
and dont forget that on forum there was a reference on published paper from MA on how system is based.
Click to expand...
WHAT ...

Can you give me the link to that paper please ....
 
Katie Chalmers said:
I’m too smooth brained for this conversation.
Click to expand...

unga bunga me smash sabakuma first time, good ugh.

Me go smash snablesnot only now, ugh very good yay for while good, then ugh pattern makes loot bad me punish bad!

Is what i gather from it xD

And as a avid "never stick to one thing, then suddnely stick for a while on one thing" kind of player this seems more like placebo than actuality :p I've said one time before but with the amount of "how loot works" theories people churn around at one point we must also start to incorporate what Super computer we think Mindark employs to constantly run these algorithms. I personally think that even the Fugaku computer would struggle at this point.

Loot is very much simpler than people expect it to be, and the only "constant" variable we know of that some people also know how to utilize, is waves. I'm sure others also knowmore depth of how it works and how deadass it is, but if mindark realizes a lot of people knows it changes. If people know how to utilize it in the majority, they would lose money.


TLDR Loot is simple, people over complicate it a LOT, and if enough players figured out or were getting close to realizing how it actually works (as in a lot of people enough to make it common talk somewhat fast) it would be changed, and Mindark already has a plan B, and a plan C, and a plan D on how to do the loot.
 
Mika said:
unga bunga me smash sabakuma first time, good ugh.

Me go smash snablesnot only now, ugh very good yay for while good, then ugh pattern makes loot bad me punish bad!

Is what i gather from it xD

And as a avid "never stick to one thing, then suddnely stick for a while on one thing" kind of player this seems more like placebo than actuality :p I've said one time before but with the amount of "how loot works" theories people churn around at one point we must also start to incorporate what Super computer we think Mindark employs to constantly run these algorithms. I personally think that even the Fugaku computer would struggle at this point.

Loot is very much simpler than people expect it to be, and the only "constant" variable we know of that some people also know how to utilize, is waves. I'm sure others also knowmore depth of how it works and how deadass it is, but if mindark realizes a lot of people knows it changes. If people know how to utilize it in the majority, they would lose money.


TLDR Loot is simple, people over complicate it a LOT, and if enough players figured out or were getting close to realizing how it actually works (as in a lot of people enough to make it common talk somewhat fast) it would be changed, and Mindark already has a plan B, and a plan C, and a plan D on how to do the loot
Click to expand...
Mining is a bit different to the other professions.

We can select the exact location and depth that we wish to test for a resource., which makes it incredibly flexible regarding the tests we can perform.
 
R4tt3xx said:
Mining is a bit different to the other professions.

We can select the exact location and depth that we wish to test for a resource., which makes it incredibly flexible regarding the tests we can perform.
Click to expand...

Mining and crafting i have no clue about, at all, i litterally am the epitome of "HURR DUUR BIG AMP FOMA GOOOO" xD So take my reply as a hunting loot only i suppose :p
 
R4tt3xx said:
Mining is a bit different to the other professions.

We can select the exact location and depth that we wish to test for a resource., which makes it incredibly flexible regarding the tests we can perform.
Click to expand...
Don't forget the mapping of ores changes constantly you might one day find cave sap in one place the next day it's oil or something else. I don't even think that's wave based or if it is very short bursts which make it hard to counter.
 
R4tt3xx said:
Mining is a bit different to the other professions.
Click to expand...

slick27 said:
Don't forget the mapping of ores changes constantly you might one day find cave sap in one place the next day it's oil or something else. I don't even think that's wave based or if it is very short bursts which make it hard to counter.
Click to expand...

Or maybe, and bear with me here, it is exactly the same as other professions, the parameters just have different names?

Let us oversimplify stuff again, because it is fun, and simple

Perhaps an area in mining ia equivalent to a creature type in hunting. Perhaps the depth is "equivalent" to maturity of said creature type in hunting.

Some days, for whatever reason, you get statistically significant difference in resource composition in hunting, I would assume it is the same for mining.

Sometimes in hunting, hunting the same creature type with a different depth(maturity) allows for new resources to be collected from that creature type.

Fun rhoughts to play around with
 
Ferial said:
Or maybe, and bear with me here, it is exactly the same as other professions, the parameters just have different names?

Let us oversimplify stuff again, because it is fun, and simple

Perhaps an area in mining ia equivalent to a creature type in hunting. Perhaps the depth is "equivalent" to maturity of said creature type in hunting.

Some days, for whatever reason, you get statistically significant difference in resource composition in hunting, I would assume it is the same for mining.

Sometimes in hunting, hunting the same creature type with a different depth(maturity) allows for new resources to be collected from that creature type.

Fun rhoughts to play around with
Click to expand...
We need R4tt3xx to figure this out so I can 1 drop cave sap etc every time :sneaky:
 
slick27 said:
We need R4tt3xx to figure this out so I can 1 drop cave sap etc every time :sneaky:
Click to expand...
Found something interesting, going to try it tonight, spiral + depth...

And if this works... Something else in the guide would have to change ...
 
Last edited:
hint: look at craft system
parameter 1: Quality rate of the BP
parameter 2: Skill to use the BP
parameter 3: condition slider

cheapest test for distribution and randomizer... load 1k ped nanocube and click 50.000 EP1
come back in 2 days and you have a 50k loot event statistic

test with MU involved: load "basic filters" load 50k oil and 50k lysterium, afk 2 days cehck results
you will see also salvage outomes

test 3: load a qr 100 "lysterium power container", 1.2 ped per click, run 50k click (60k ped anc ome back in 2 days)

when i read in chat "today xxxx are bearing good loot" from someone after 100 kills for a daily i smile.
we can not student T entropia, the noise in first 500 loot events (loot nanny user know what i mean) prevent it
but an analysis of 5000 10.000 50.000 loot events is meaningful.
the real matter is that if a multiplier occurs on a 10 pec random bberycled after long sessions on rextelum
we "waste" a multiplier event.
but what to kill (ped per kill) part of risk control. if i kill only 1 ped monsters (whatever monster) i dont have multipliers swings.
and once every abotut 40.000 to 90.000 kiit events a multiplier over 1000x arrives
statistics..... perception... deception....
 
Ferial said:
Or maybe, and bear with me here, it is exactly the same as other professions, the parameters just have different names?

Let us oversimplify stuff again, because it is fun, and simple

Perhaps an area in mining ia equivalent to a creature type in hunting. Perhaps the depth is "equivalent" to maturity of said creature type in hunting.

Some days, for whatever reason, you get statistically significant difference in resource composition in hunting, I would assume it is the same for mining.

Sometimes in hunting, hunting the same creature type with a different depth(maturity) allows for new resources to be collected from that creature type.

Fun rhoughts to play around with
Click to expand...
Mining is 90 degrees hunting.
 
Msturlese said:
hint: look at craft system
parameter 1: Quality rate of the BP
parameter 2: Skill to use the BP
parameter 3: condition slider

cheapest test for distribution and randomizer... load 1k ped nanocube and click 50.000 EP1
come back in 2 days and you have a 50k loot event statistic

test with MU involved: load "basic filters" load 50k oil and 50k lysterium, afk 2 days cehck results
you will see also salvage outomes

test 3: load a qr 100 "lysterium power container", 1.2 ped per click, run 50k click (60k ped anc ome back in 2 days)

when i read in chat "today xxxx are bearing good loot" from someone after 100 kills for a daily i smile.
we can not student T entropia, the noise in first 500 loot events (loot nanny user know what i mean) prevent it
but an analysis of 5000 10.000 50.000 loot events is meaningful.
the real matter is that if a multiplier occurs on a 10 pec random bberycled after long sessions on rextelum
we "waste" a multiplier event.
but what to kill (ped per kill) part of risk control. if i kill only 1 ped monsters (whatever monster) i dont have multipliers swings.
and once every abotut 40.000 to 90.000 kiit events a multiplier over 1000x arrives
statistics..... perception... deception....
Click to expand...
did you know when to change the slider ?


That's the problem, you are not willing to interact with the systems in place in order to change the outcome....
 
Ferial said:
Or maybe, and bear with me here, it is exactly the same as other professions, the parameters just have different names?

Let us oversimplify stuff again, because it is fun, and simple

Perhaps an area in mining ia equivalent to a creature type in hunting. Perhaps the depth is "equivalent" to maturity of said creature type in hunting.

Some days, for whatever reason, you get statistically significant difference in resource composition in hunting, I would assume it is the same for mining.

Sometimes in hunting, hunting the same creature type with a different depth(maturity) allows for new resources to be collected from that creature type.

Fun rhoughts to play around with
Click to expand...
I mean isn't it literally this? Even mobs change what loot they drop after VUs
 
Mika said:
Loot is very much simpler than people expect it to be, and the only "constant" variable we know of that some people also know how to utilize, is waves. I'm sure others also knowmore depth of how it works and how deadass it is, but if mindark realizes a lot of people knows it changes. If people know how to utilize it in the majority, they would lose money.


TLDR Loot is simple, people over complicate it a LOT, and if enough players figured out or were getting close to realizing how it actually works (as in a lot of people enough to make it common talk somewhat fast) it would be changed, and Mindark already has a plan B, and a plan C, and a plan D on how to do the loot.
Click to expand...
It used to fascinate me to no end, how it appeared some people had somehow kind of figured out a method of taking advantage of this. I only have a small amount of time to be online now, so I get what I get. My old days I could spend a whole day on grinding one mob or hopping around. I noticed when I shoot the same thing over time, you can see how there are loot periods where most of the time was regular loot, and random periods of increased loots, and sometimes you get what they call multipliers if you're lucky.

Interesting for me to note that over time I had many people adding up on my friend list. In periods of increased loot events, I would see certain people come online, and in some cases they would have a significant loot event immediately or shortly after coming on - and then they would log off.

Now, I'm not smart enough to know the reason for all of this. It's just one of those human pattern recognition things that I observed. Somehow it appeared they knew when to come on, they knew what to hunt, and when to hunt, and would regularly have significant loot / item events. Now, that could just be luck, or random, or whatever, but it always made me wonder wtf was going on. I tried many things over the years and never had a significant item or loot event. So, I gave up trying to force it or figure it out. Now I just plink around randomly on low level codex mobs to get skills - whatever that's worth, and whatever happens happens.
 
Msturlese said:
hint: look at craft system
parameter 1: Quality rate of the BP
parameter 2: Skill to use the BP
parameter 3: condition slider

cheapest test for distribution and randomizer... load 1k ped nanocube and click 50.000 EP1
come back in 2 days and you have a 50k loot event statistic

test with MU involved: load "basic filters" load 50k oil and 50k lysterium, afk 2 days cehck results
you will see also salvage outomes

test 3: load a qr 100 "lysterium power container", 1.2 ped per click, run 50k click (60k ped anc ome back in 2 days)

when i read in chat "today xxxx are bearing good loot" from someone after 100 kills for a daily i smile.
we can not student T entropia, the noise in first 500 loot events (loot nanny user know what i mean) prevent it
but an analysis of 5000 10.000 50.000 loot events is meaningful.
the real matter is that if a multiplier occurs on a 10 pec random bberycled after long sessions on rextelum
we "waste" a multiplier event.
but what to kill (ped per kill) part of risk control. if i kill only 1 ped monsters (whatever monster) i dont have multipliers swings.
and once every abotut 40.000 to 90.000 kiit events a multiplier over 1000x arrives
statistics..... perception... deception....
Click to expand...
based on this assumption , if you hunted rex and lost 10k ,then did a daily chirpy and got a 1000x multi that the 10k would be lost .
it would mean that 90% of the playerbase would be down heaps of peds . MA stated that average players returns are around 90(or 95% cant remember ) .
so this theory does not add up
 
WildD3mon said:
MA stated that average players returns are around 90(or 95% cant remember ) .
so this theory does not add up
Click to expand...

MA made a statement on averages. It is impossible to look at an average and say anything about any particular data point that makes up that average.

For instance if the entire active consists of 1000 players and 999 of them have 50% returns, and one player has around 47420% returns the average would still be 95% (if they cycled the same amount)

There are other factors to consider when looking at those averages. Someone with poor results is more likely to quit, take a break or otherwise cycle less. Whereas the opposite is true for someone with better results.

In short, it is impossible to come to the conclusion you came to, accurately

I would also like to add that based on the statement that there is no personal loot pool or other compensation systems in place, it could not work in any other way than your example, since any other way (that I can think of right now) would require the system to use your current (and past) results as input for future results. This is the definition of a "compensation system"

Also, I recall a very old MA post(i think) that more or less said *Lost PED is lost*. I do not remember the wording but I think it was in reference to someone blowing a bunch of big mining amps and expecting the lost money to come back.
 
Last edited:
@Ferial you pretty much pointed to a 98% return random generator (in TT Terms).

We cannot know the exact formula but we can for sure examine loot series and assume that in the long term they converge to the reported 98% return stated in 2017

This leads to a single residual question: how much is the combined effect of looter and eff impact?
personally i think that it is some thing like 7% COMBINED and total impact is a number like 140 as a sum of looter+eff.
this would be added to a 92% base return to go from 92 to 99 with linear effect.
i know i am not on the "90 eff bandwagon) but a 70 eff and 70 looter i consider maxes the formula
as well as a 90 eff and a 50 looter
or 144 looter and a grindhouse (yes, the example is specific)
 
So here is an interesting observation... I just want to make sure that the I am correct with regards to this..

An Md1 has an average depth of 104.5. Most finders have a min / max depth of -+200m. Would I be correct in saying that the md1 is left skewed as it's recorded min depth is 5m ?
 
R4tt3xx said:
An Md1 has an average depth of 104.5. Most finders have a min / max depth of -+200m. Would I be correct in saying that the md1 is left skewed as it's recorded min depth is 5m ?
Click to expand...

Based on the assumption that +-200 is correct (it is not) the depth distribution with that partucular finder would be right skewed, not left skewed.

 
Msturlese said:
@Ferial you pretty much pointed to a 98% return random generator (in TT Terms).

We cannot know the exact formula but we can for sure examine loot series and assume that in the long term they converge to the reported 98% return stated in 2017

This leads to a single residual question: how much is the combined effect of looter and eff impact?
personally i think that it is some thing like 7% COMBINED and total impact is a number like 140 as a sum of looter+eff.
Click to expand...
This has been tested by Zho and looks to be 7% from looter and 7% from efficiency:
How Loot Returns Work as Function of Looter Level and Efficiency, Based on Actual Data, as of March 3rd, 2022
 
Ferial said:
Based on the assumption that +-200 is correct (it is not) the depth distribution with that partucular finder would be right skewed, not left skewed.

Click to expand...
ok so if the data to the left is missing, it is right skewed ... got it... Will take a look at the video again when I get some time..
A moving average does not change skewness correct ?
 
Last edited:
WildD3mon said:
based on this assumption , if you hunted rex and lost 10k ,then did a daily chirpy and got a 1000x multi that the 10k would be lost .
it would mean that 90% of the playerbase would be down heaps of peds . MA stated that average players returns are around 90(or 95% cant remember ) .
so this theory does not add up
Click to expand...
Kind of agree with this as I do little large stuff but how would I still end up around 90% for certain time periods it's like it's adding up input and output.
 
R4tt3xx said:
A moving average does not change skewness correct ?
Click to expand...

I don't understand the question. skewness is something you look at as a static analysis tool. An average is a single datapoint in time. You cannot look at skewness in a single datapoint. A moving average is essentially the difference between two data sets. You can compare the skewness between the datasets. But you can't do anything like that with the average

Averages are useful for a lot of things, but as mentioned before it cannot say anything about the data points that make up that average.

When talking of left skewness I am pretty sure that @Msturlese was talking about the fact that there are more "apparent multipliers" that are lower than your expected average than those that are higher. As such, multiplier distribution is left skewed.
 
What about a PRNG ?
 
Ferial said:
I don't understand the question. skewness is something you look at as a static analysis tool. An average is a single datapoint in time. You cannot look at skewness in a single datapoint. A moving average is essentially the difference between two data sets. You can compare the skewness between the datasets. But you can't do anything like that with the average

Averages are useful for a lot of things, but as mentioned before it cannot say anything about the data points that make up that average.

When talking of left skewness I am pretty sure that @Msturlese was talking about the fact that there are more "apparent multipliers" that are lower than your expected average than those that are higher. As such, multiplier distribution is left skewed.
Click to expand...
You guys did notice the moving average in both the reported loot classes and level breakdown in this game correct ?
 
R4tt3xx said:
What about a PRNG ?
Click to expand...
I don't think you are actually asking the questions you are interested in learning the answers for. The purpose of the prng is to output seemingly random numbers.

What then uses the output from the prng to produce results can the be tuned to produce whatever distribution you would like.

R4tt3xx said:
You guys did notice the moving average in both the reported loot classes and level breakdown in this game correct ?
Click to expand...
In the averages that was published after loot 2.0 inplementation there was no moving average. There were different averages for different data sets.

Go study a bit, come back later and we will be able to have a much more productive discussion.
 
Ferial said:
I don't think you are actually asking the questions you are interested in learning the answers for. The purpose of the prng is to output seemingly random numbers.

What then uses the output from the prng to produce results can the be tuned to produce whatever distribution you would like.


In the averages that was published after loot 2.0 inplementation there was no moving average. There were different averages for different data sets.

Go study a bit, come back later and we will be able to have a much more productive discussion.
Click to expand...
0110101101
miss hit hit miss hit... Results are predetermined when a tool is activated..... BUT I do see what you are saying about not just using the loot event as the PRNG trigger, so that would either be the count in a sequence or a timed response.. Those would be the most likely triggers.


Was loot 2.0 applied to mining ?
 
R4tt3xx said:
so that would either be the count in a sequence or a timed response.. Those would be the most likely triggers.
Click to expand...

The most likely "trigger" would be buckets. As that is one of very few ways to ensure that the outcomes align with expectations.

Get prng value : 34210
Lookup value: 34210
32410 is in the 25000-50000 bucket: return 0.75x cost spent.

Very crude example.

Loot 2.0 was hunting only, tweaks were made to volatility in mining and crafting shortly after. This is in the dev notes
 
Ferial said:
The most likely "trigger" would be buckets. As that is one of very few ways to ensure that the outcomes align with expectations.

Get prng value : 34210
Lookup value: 34210
32410 is in the 25000-50000 bucket: return 0.75x cost spent.

Very crude example.

Loot 2.0 was hunting only, tweaks were made to volatility in mining and crafting shortly after. This is in the dev notes
Click to expand...
Im going to like this forum... I have a lot to learn, so please proceed to teach and advise...

Pity I cant play the game for a while, gfx card is RIP... We will see later on today.
 
So I am already seeing something that I recognize.. LCG, So Ma wants me to try it, fail and deposit more peds into the system, that has been the cycle from day 1....

So off the top of my head, take the last digit of your x coordinate and head in that direction cos(2*pi()*(x/9))*55 sin(2*pi()*(y/9)*55... Pretty simple...

And it will last for a few attempts, then diminishing returns kick in and render your pattern moot.
 
Last edited:
You must log in or register to reply here.
Back
Top