FYI: How Loot Returns Work as Function of Looter Level and Efficiency, Based on Actual Data, as of March 3rd, 2022

Would they though? A noob even at 100% TT isn’t going to generate anything substantial. Muleing is pretty easy to detect unless you want to put a whole lot of effort into disguising it, and still wouldn’t be worth the risk in breaking TOS and giving MA a reason to suspend a withdrawal.
People used to do it for sweat.
 
Looking at baseline multipliers, we can extrapolate to the larger multipliers.

No, you can't do that, and you're one of the last persons on this forum who I'd expect to state such thing. The most basic counterargument is that, by necessity, the larger multipliers will be lower populations than lower multipliers and that means different treatment to ensure control. Because MA must ensure control from a point forward (or, rather, from a TT qty upwards).
 
No, you can't do that, and you're one of the last persons on this forum who I'd expect to state such thing. The most basic counterargument is that, by necessity, the larger multipliers will be lower populations than lower multipliers and that means different treatment to ensure control. Because MA must ensure control from a point forward (or, rather, from a TT qty upwards).
The probability distribution function doesn't change. The expected value is scaled simply by looter and efficiency, so MA maintains full control.

Every player has the same chance of hitting a particular multiplier on a particular mob. But one player may get that multiplier times 93%, another may get that multiplier times 98%, depending on efficiency/looter.
 
I started playing last week and I just say I don't have much experience like you guys. But it strikes me that "too identical Globals and Minis" do not correspond to the definition of Multiplier at all. That is, no matter what weapons I use, I always get the same type of "Multipliers".

How do you define this behavior in your formula?

To the multiplier of 0% or 7% in efficiency, my final Global value should be much larger, not identical to the previous ones.
I don't know, I'm not a mathematician, but maybe something still eludes me...
 
The probability distribution function doesn't change. The expected value is scaled simply by looter and efficiency, so MA maintains full control.

Every player has the same chance of hitting a particular multiplier on a particular mob. But one player may get that multiplier times 93%, another may get that multiplier times 98%, depending on efficiency/looter.
Agreed, this is the simplest implementation. No need to make it more complicated than this.
 
I started playing last week and I just say I don't have much experience like you guys. But it strikes me that "too identical Globals and Minis" do not correspond to the definition of Multiplier at all. That is, no matter what weapons I use, I always get the same type of "Multipliers".

How do you define this behavior in your formula?

To the multiplier of 0% or 7% in efficiency, my final Global value should be much larger, not identical to the previous ones.
I don't know, I'm not a mathematician, but maybe something still eludes me...
1) You would have to control for cost to kill, which depends on weapon dpp.
2) You would need many instances of globals within a particular multiplier group to get a decent average value without much uncertainty because there is some distribution in a particular multiplier group.
3) After all of that, the difference between efficiency like you said is 0-7%. Unless you are actually comparing 0% eff and 100% eff, the difference is probably smaller. Again, you'd need a large number of multipliers to get a decent enough mean value...this is why looking at the lower end multipliers make it much easier because you can get so many more loot instances for them.
 
1) You would have to control for cost to kill, which depends on weapon dpp.
2) You would need many instances of globals within a particular multiplier group to get a decent average value without much uncertainty because there is some distribution in a particular multiplier group.
3) After all of that, the difference between efficiency like you said is 0-7%. Unless you are actually comparing 0% eff and 100% eff, the difference is probably smaller. Again, you'd need a large number of multipliers to get a decent enough mean value...this is why looking at the lower end multipliers make it much easier because you can get so many more loot instances for them.

In your opinion, is it possible for a Yuka Globals(for example of 334.52 ped), to be a derivative of several bonuses including the standard TT from the kill itself. For example 4.52 + 30 + 300.

I mean if the weapon's efficiency has an impact of 0-7% on each shot that counts towards the base TT return of 4.52, but not the bonus values. Then its calculation, even on the smallest bonus, is not feasible.

So what I know is that Efficiency is an additional value to the basic Cost of the weapon, and that's why the weapon description says that it has an effect on dealing damage.

And at the same time, I get the same 10 ped globals from my favorite mob Kerbs whether I kill him with 2 hits or 30 hits. I want to say that if the efficiency multiplier has an effect on the globals, it will manifest itself in the larger ones weapons with a higher cost... but at least for me these Globals are identical. Therefore I think Efficiency doesn't have any impact on Multiplies, only on single interaction(hit).

Ok i dont say im right. Im just looking for answers. But if definition Multiplayer is right, then killing Kerbs by Gravis, my Globals should be 5 times biger than kill them by Barbarella.

How did you estimate that Efficiency has an impact on multiplications?
 
Thanks a bunch for taking your time with the tests and for sharing them. I will refrain from debating, don't want to come across as negative, the work you did here wouldn't deserve that.
 
nice math exercise. it remembers me the "normal" distribution and Student T shortcut tho.

I worked in volatility trading for too many years.
entropia is not normalli distrivuted. it is leptocurtic and etheroskedastic just like real world.
this model fits the sample and its limit is that is overfitted on the lower side imho (you used smaller samples)

there are also 4 sigma events (ATHs) and Swine/Grindhouse/Toy Snowballer effects of low efficiency that lead to pumping multipliers.

everytime i look at game mechanics i see crafting machine that hints me the real engine behind.. i dont think that the 3 profession have different engines... would be too hard to maintain. the randomizer function most probably is the same...

1. condition slider
2. Blueprint quality
3. Skill level

I think that the condition slider is related to efficiency and blueprint quality to looter (skill level to use an item is just that)

AND.... a rebalancing mechanism to realign to the mean reversion of returns (we all got a nice multiplier and then went dry for more loot)

i dont discuss the fact that your model DESCRIBES the dataset
but i leave a door open to hypotesis that this data set can be described in different ways

Great work, but this would say that puny have insane multipliers inside to allow people skill with very modest stats without destroying resources...
 
Thanks a bunch for taking your time with the tests and for sharing them. I will refrain from debating, don't want to come across as negative, the work you did here wouldn't deserve that.

But isn't that exactly the point of the dialogue, why do you give up your position before you have presented it and compared it against criticism? The thread is in the Loot Theory section, which means it's something insecure to work on and improve.

In my opinion, it is more disrespectful to tell someone they are wrong without giving an argument. I'm interested to know why you think Efficiency has no impact on larger multipliers.



The tower test clearly shows that the system always returns no less than 70%, which includes the smallest multiples on which the study is built.

This tower test can also be considered reliable that the Looter profession has an impact "only when searched in the corpse", because it only counts the Weapon's Efficiency, not the Looter level.

When Cyrene adopted the Codex, I hunted a mob that had no other hunters. And for 3 days I was getting ~75% TT return.

By comparison, from my casual experience, I think it is obvious that the system returns a minimum of 70%, and the others are distributed in some other way.

Before I can figure out how that 30% margin is distributed, I need specific values for Efficiency and Looter.

For the Looter profession test, I have several times offered PED to other players of different Looter level to join in a quick test, but my attempts to gather ~5 people were unsuccessful.

Similar to the tower test, I wanted several people to shoot at a large MOB with the same weapons *10 hits each, but with weapons with 0% Efficiency, with a new player without any Looter skills.

Once everyone has taken 10 hits (ie equal bet value)... much more accurate deductions can be made as to how it affects Looter... such as if each takes a different number of shrapnel, it will be equivalent to Looter having impact on your personal profession. But if everyone takes the same value, it will mean that the impact is entirely on the one picking the loot. From which you expect more shooting for the test until the full circle in the group is turned.

I personally haven't been able to get even a small group together to do a simple test on this. Good luck if you make it...

And how do you calculate Looter in your loot, on what basis?
0.07% did you check the credibility yourself before going ahead with some deeper theory.. and how?

DPP/Low bet per shot.
Although I can see the impact of the lower cost in the ratio.. in the end.. in the result. I can't call up exact data until I establish the influence of Efficiency and Looter. Simply because the above tests show that they increase the base by some amount in the bottom TT line. While the DPP analogy is decrease, not increase.
That is, opposite actions involved in the final result. So I cannot include them is my personal research, before I have reliable data from Looter and Efficiency, which seem easier to establish.

Open up a bit, anyway in the new version, there will be changes... so I don't see the point of secrecy in this case.

In my opinion, the multipliers use similar like the Critical Hits pattern.

Normal hit + the weapon max hit.

But in multipliers it's just with more frames that the game gives when certain values are reached...

Normal Loot + Bonus + Bonus + Bonus...

That's why I think Looter and Efficiency only have an impact in the main loot... those normal we took from the most mobs.

I am interested in understanding your arguments.
I am aware that they are built on increasing max losses.... 90% + Effectiveness + Luthor + DPP + Crit Hit + Crit DMG... that is, a vague equation with vague parameters that gives a better result. Do you have any specific values or are the calculations based on "longer run = better".
 
The new updata that you proposed has catched me in. i never focus on TT return but i will run a test myself in this term
Next Island Desert Crawler 5000 loot event each of
1. 56 looter with MF chips at 58% eff
2. 56 looter with EP38 and 78% eff
According with your tables i am expecting 95.6% TT Returns
at an expected average of 800 hp it is 8.000.000 Hp test or 24.000 PEd i hope that the sample is big enough to be meaningful or i add more cycle (just 14 codex there)
(i expect to return over 97% btw)
i will repost after the test.
 
The new updata that you proposed has catched me in. i never focus on TT return but i will run a test myself in this term
Next Island Desert Crawler 5000 loot event each of
1. 56 looter with MF chips at 58% eff
2. 56 looter with EP38 and 78% eff
According with your tables i am expecting 95.6% TT Returns
at an expected average of 800 hp it is 8.000.000 Hp test or 24.000 PEd i hope that the sample is big enough to be meaningful or i add more cycle (just 14 codex there)
(i expect to return over 97% btw)
i will repost after the test.
5000 not big enough for experiment I don’t think. Add a zero and you’ll be within 1-2% of expected.
 
I'd like to add from personal experience using a 90%+ eff gun as early as level 20 looter and now I'm currently 57 looter, it's very apparent and noticeable that efficiency and looter stacks and work together. Most of my hunts aligned on the expected averages similar to Zho's table and that was with around 10-30k PED cycle each hunt.
 
5000 not big enough for experiment I don’t think. Add a zero and you’ll be within 1-2% of expected.
damn 50k Katie if yuou are challenging me ... let's do with a monster that has some MU at least... Chimera? has same HP pool
 
But isn't that exactly the point of the dialogue,

Not necessarly, things get too personal too fast. My main issue is with the core asumption, respectively "multipliers". The way this notion is used on the forum has generated a whole myhtology portraying it as an accepted and validated truth, but in reality it has as much substance as believing in Lootius. "Multplier" is an empiric tool used by players to observe/quantify diferent loot events. That has exactly zero weight of proof in the direction that "multipliers" would be used in the organic process of loot creation/loot event calculation. The second layer of confusion, which is alot more important, is that all tiers of loot are equal in regards to the "democracy" of their generation, which is just a fallacy.

My view of the loot is somehow akin to collecting water from a piped spring. You have control of the valve, namely potential volume per time, and you have control of the can, namely efficiency of collecting. But in regards to wether the water is actually springing, you have zero control.

Inline with your Cyrene observation, in regards to hunting, I think there's a base loot function which is strictly governed by weapon (DPP, DPS, EFF) and skills (looter). Were an avatar played as you tried, this will yield a set return based on two factors: a base return (say 75%) which is then adjusted by weapon details, within a margin of few %, respectively the place of the weapon&others vs the overall scale (EFF 0 to 100, looter prof say 0 to 100 etc). This is set by MA in both regards (the "75%" and the "few%") and is the bread and butter of "volatility". This would be "regular multipliers" (god I hate the concept).

The rest, the "bigger multipliers" I see comparable to a dividend return system in which the player is rewarded first and foremost for activity from a "general lootpool". In this part, the "eco" (encompassing all above factors) is alot less relevant, with a primacy of volume/dps. The lower the access to this divident fund gets, then kicks in a mercy system (which you can find in great detail in gotcha games), such as nobody to get to a return as atrocious as "0 access to bigger hofs". Obviously a very low eff gun with a very high dps will then generate lots of hoffs with a greater penalty in base loot, such as to not milk the system by mindless clicking.

So then a super duper active and well kitted player (say Eve) will see best results ttwise in most crowded places/mobs/planets (based on what exactly fundament is MA using for their "loot pool") and worst on some irrelevant big mob in the butt of the universe. This is strictly ttwise, fortunately MU and access to MU can influence alot the considerations.

The beauty of this system, in my view as I was mentioning on another thread, is that allows multiple approaches in same time. Very large size samples will fall in rhytm with the base factors mentioned such as eff and looter because the bigger loots will melt in the overall return. Very short samples can have their inherent extreme volatility alleviated by MU. Returning to very high samples it won't be then alleviated too much by MU because the volume is inherently controlled by market and then weapon stats and skills get, the bigger the volume, the more importance of them, whereas for smaller samples ArMatrix is more than enough if correct MU is targeted.

I just wrote loads of platitudes which don't necessarly contradict the common "meta", with a substantial (in my view) difference: not everything scales (directly proportional) in Entropia.
 
well there definitely are some "hidden rules" in loot distribution, otherwise if it was strictly as simple as a roll for item, and a roll for TT value, people would be looting 10k blazar fragments from exasaurs as easy as they loot a few ped worth of muscle oil. but that is patently not the case, the game "knows" that the item has some kind of value that isn't represented by its TT value, or its roll is somehow based on quantity rather than TT value. how many hidden rules for loot like that are there? impossible to know.
 
Great update. Nice to see even more data points. Can't wait to see the next one.
 
well there definitely are some "hidden rules" in loot distribution, otherwise if it was strictly as simple as a roll for item, and a roll for TT value, people would be looting 10k blazar fragments from exasaurs as easy as they loot a few ped worth of muscle oil. but that is patently not the case, the game "knows" that the item has some kind of value that isn't represented by its TT value, or its roll is somehow based on quantity rather than TT value. how many hidden rules for loot like that are there? impossible to know.
That's not a hidden rule, that is wave loot
 
That's not a hidden rule, that is wave loot
Wow, the irony! I'm surprised you believe in something MA hasn't officially stated exists :ROFLMAO:.
 
you forgot the /sarc and laugh emoji, some noob like me might not realize you're joking! :)
I don't think it was a joke. This thread is about TT returns, not rare loot and markup.
There's no proof of waves in TT returns but rare loot is released on a timer. This is what most refer to as "waves".
From what I have seen there's 30 to 90 minutes between each wave.
Some items have almost 100% drop rate as soon as they are released on wave while others are still rare.

Example on items that you get immediately as they are released are rare crystals (like Purpurite) on Next Island, Calypso Bone Sample, and more that I will not mention :)

I don't think blazar are on wave though. You get it all the time, slowly.

the game "knows" that the item has some kind of value that isn't represented by its TT value
I think it's the other way around. MA tries to balance the loot to maintain markup on items to make the game more interesting.
By limiting the supply of items that are in demand (like tier comps) the MU is maintained.
Just as an example the daily hard cap for T4 comps might be 500 or something. You compete with all other players to loot this limited supply.
If every avatar switches to Maffoids it doesn't mean more T4 will drop, you just compete for the same limited supply of 500.

This is not just based on guesses. There are mobs where this behavior is much easier to observe.
 
I don't think it was a joke. This thread is about TT returns, not rare loot and markup.
There's no proof of waves in TT returns but rare loot is released on a timer. This is what most refer to as "waves".
From what I have seen there's 30 to 90 minutes between each wave.
Some items have almost 100% drop rate as soon as they are released on wave while others are still rare.

Example on items that you get immediately as they are released are rare crystals (like Purpurite) on Next Island, Calypso Bone Sample, and more that I will not mention :)

I don't think blazar are on wave though. You get it all the time, slowly.


I think it's the other way around. MA tries to balance the loot to maintain markup on items to make the game more interesting.
By limiting the supply of items that are in demand (like tier comps) the MU is maintained.
Just as an example the daily hard cap for T4 comps might be 500 or something. You compete with all other players to loot this limited supply.
If every avatar switches to Maffoids it doesn't mean more T4 will drop, you just compete for the same limited supply of 500.

This is not just based on guesses. There are mobs where this behavior is much easier to observe.
wave theory to me seems like it would be indistinguishable from a normal distribution unless you knew exactly when those waves were, and could show that you got some statistically significant amount of uncommon/rare loot for a given ped spend. i mean, you could watch globals, but by the time you see the global and have time to capitalize on it, the wave would be over, would it not? and if the time is wildly variable (as you said, 30-90 minutes), how could you target a "rich window" in which to make your play? to me it seems like an impossible theory to test
 
I think it's the other way around. MA tries to balance the loot to maintain markup on items to make the game more interesting.
By limiting the supply of items that are in demand (like tier comps) the MU is maintained.
Just as an example the daily hard cap for T4 comps might be 500 or something. You compete with all other players to loot this limited supply.
If every avatar switches to Maffoids it doesn't mean more T4 will drop, you just compete for the same limited supply of 500.

MU as far as I can tell is based on the same supply and demand strictures of any economy. Some things drop independently of TT return calculations (or in lieu of them). How MA calculates that is hard to say. There are some things that are common enough, but because their demand is high, they have decent MU. I'm seeing that on Cyrene. Every mob there drops stuff that has MU, because there's a robust set of crafting recipes or upgrade missions for nearly everything that drops.
 
wave theory to me seems like it would be indistinguishable from a normal distribution unless you knew exactly when those waves were, and could show that you got some statistically significant amount of uncommon/rare loot for a given ped spend. i mean, you could watch globals, but by the time you see the global and have time to capitalize on it, the wave would be over, would it not? and if the time is wildly variable (as you said, 30-90 minutes), how could you target a "rich window" in which to make your play? to me it seems like an impossible theory to test
Not impossible to test at all and rare loot can be "exploited" for sure. I'm hesitant to give to many details since MU is one of the few competitive factors in this game. But one hint is to use soc chat :)
But it's pretty annoying to camp for wave loot. You can get very high avg. MU with these tactics though. And once learn the methods you see it in a lot of places.
 
Not impossible to test at all and rare loot can be "exploited" for sure. I'm hesitant to give to many details since MU is one of the few competitive factors in this game. But one hint is to use soc chat :)
But it's pretty annoying to camp for wave loot. You can get very high avg. MU with these tactics though. And once learn the methods you see it in a lot of places.
This is by far one of the most important and hard to grasp concepts of the game

people are constantly reminded but never quite get it

this is a pvp game even though many don’t like it
 
wave theory to me seems like it would be indistinguishable from a normal distribution unless you knew exactly when those waves were, and could show that you got some statistically significant amount of uncommon/rare loot for a given ped spend. i mean, you could watch globals, but by the time you see the global and have time to capitalize on it, the wave would be over, would it not? and if the time is wildly variable (as you said, 30-90 minutes), how could you target a "rich window" in which to make your play? to me it seems like an impossible theory to test
There are certain places where the waves are very clear ( item waves not tt). I'll give an example ( which ended up getting changed, maybe as it was so evident but was a nice way to prove they are there :) ) Minotour horns were for a long time on a very specific timer- to the extent that the same crowdof people would turn up exactly every x number of minutes, farm the horns, then disappear. If you hunted out with this time and after they had all be depleted, then you wouldnt get any until it was back to that time again. It doesnt really need a lot of testing to prove it is there, its very evident - but of course as said above, most wont talk about it for obvious reasons.
 
Back
Top