Developer-Notes--3

[wizzszz]

Indeed, and that talks against "personal" pool. It's rather validation of a per-mob payback (which would be the simples implementation, which the well-known razor would be in favor of).

Hmm, i disagree, despite Occam.

Let's assume that every mob you loot has 90% of the ammo spend in loot (average)
This is independent from who spent the ammo.

Now what happens when the mob has less than the average loot?
It gets tracked somewhere and repaid later in a mob that has more than average loot.

This is "loot tracking", or more commonly known as "personal loot pool".
It does not matter who spends the PEDs, the not-paid-out loot goes somewhere and will be (has to be?!!) paid out later, to achieve the 90% average loot.

It doesn't have to be tied to YOUR losses directly, but to "how much loot was missing from that mob from the 90% of it's kill costs.

That a mobs loot increases when you blast more PEDs into them already proves that they (short-term) track your spendings - and this is exactly as expected, as they have to do that anyway, to determine who is allowed to loot a mob.

Gun with half dmg/sec: Average global size= 86,3 ped
Gun with doouble dmg/sec: Average global size= 193,4 ped (wich is quite high, had one HOF @ 1134 peds, without HOF its 143 peds average)

If bigger (loss compensating) payouts are time based (something that has been reported ountless times already), it would explain your observations: The double dmg/sec gun allows you to lose far more within the same period of time, so the compensating globals have to be bigger, too.

Would explain, too, why small mobs hardly global - you simply don't lose enough between "pay-days".

 

[narfi]

Here are some of my thoughts from my experiences and from reading others here on the forum (I can only ever really 100% trust my own observations, but everything should be acknowledged as a possibility)

Before these dev notes, I somewhat believed in a personal loot pool.
My 'loosely' believed theory was that:

Each individual mob tracked tt costs into it regardless of who does the dmg. (Ammo + weapon decay + amp decay)
It could be teams or individuals etc... doesn't matter the mob knew how much tt went into it. (but didn't care about armor/fap)

When looted the mob then spit out a loot based on that cost. (this is easier to observe on mobs that dont give no-looters)

For example absorbs 40ped tt of weapon decay and ammo from reasonably eco weapon/amp combos. (ignoring extreme possibilities of super eco or super uneco tt output weapons)
The average loot would be about 20ped with globals of around 300 ped every once in a while. (once in a while very low or very high multipliers, but not common)

With the belief in a personal loot pool and my observations I believed that the looter was the one who's loot pool was credited or debited depending on if it was above or below the 90% mark.

At the time this seemed to make sense to me and each time I fell below 2-3k ped under the 85% mark (in a 5% taxed la) I would get an uber or a series of 2-700ped hofs which caught me back up.

Then these notes came out and I am forced to believe there is no personal loot pool.
So I have changed my thoughts a little while keeping the core.

Now instead of a personal loot pool, I believe it is a Mob (I don't know if its maturity specific or not) based loot pool.

I have always tried to target mobs that few other people are hunting, and have always had acceptable averages that way.

There have been only a handful of campers hunting Things over the last few months while I was there and likely no more than 4 at any given time. I believe this eliminates alot of the 'chance' for individual loots with a mob based loot pool theory, and might explain why I have observed what looks like a personal loot pool while hunting them.

I toyed with extreme over-amping and had poor results, so could not do a large enough test to completely eliminate it as a possibility from my mind.
Assault Rifle(L) from the Hunt The Thing tt is a blp rifle which does 2dmg.
L7 Things are high regen 200hp mobs.

I used Assault Rifle(L)+beast to do 3dmg for a stupid cost till I had pumped 20ped of ammo+gun+amp into an L7 Thing and then killed it fast with a larger weapon. I did this 5 times and each time I received less than 2ped of loot. (one no looter, one with fragments, one less than 1ped of loot and 2 with 1-2ped of loot)
(from a normal mob I would expect 200hp to cost 70pec of decay at 2.85dpp and return around 1/2 of that(35pec) as a typical loot size)

I am not familiar enough with no loot mobs to know if the 1-2ped loots are common from reasonable kill costs or not, but they seemed higher than the 200hp would suggest they should loot.

 

[wizzszz]

Obviously the system will not compensate your for losses that are a result of bad choices, like shooting in the air, wasting amp damage and such.
I think this was the main point in the dev notes, along with no compensation for money wasted on (L) weapons with a huge markup.


However, if you tend to think now that the loot is in the mob, it is possible that someone gets all the biggies and you are left with all the "below average" mobs.

This is clearly random and only the law of large numbers makes your loot (somewhat) converge to the 90% payout rate - how would that work with the "it's not gambling" statement? Roulette works the same way (and still converges to a fixed payout rate over time, even more than 90%).

 

[Uvas]

Hmm, i disagree, despite Occam.

Let's assume that every mob you loot has 90% of the ammo spend in loot (average)
This is independent from who spent the ammo.

Now what happens when the mob has less than the average loot?
It gets tracked somewhere and repaid later in a mob that has more than average loot.

This is "loot tracking", or more commonly known as "personal loot pool".
It does not matter who spends the PEDs, the not-paid-out loot goes somewhere and will be (has to be?!!) paid out later, to achieve the 90% average loot.

It doesn't have to be tied to YOUR losses directly, but to "how much loot was missing from that mob from the 90% of it's kill costs.

That a mobs loot increases when you blast more PEDs into them already proves that they (short-term) track your spendings - and this is exactly as expected, as they have to do that anyway, to determine who is allowed to loot a mob.




If bigger (loss compensating) payouts are time based (something that has been reported ountless times already), it would explain your observations: The double dmg/sec gun allows you to lose far more within the same period of time, so the compensating globals have to be bigger, too.

Would explain, too, why small mobs hardly global - you simply don't lose enough between "pay-days".

There is no personal loot pool.

The 90% return is just an overall average return that seems to be the median point.

There is no tracking of when someone goes over or under the average. The average return of all players seems to be around 90%.

*individual returns may vary*

If you are currently far below average, odds are that future returns will be higher than your current average, so returning you closer to average return over the long run. Reverse, if you are doing very well.

No need to track anything beyond what went into killing the mob.

 

[wizzszz]

There is no personal loot pool.

Thanks for that insight. Without your post i would have been searching on forever.

There is a little flaw left: You can't prove your point, nor can you disprove mine.

The 90% return is just an overall average return that seems to be the median point.

You stopped thinking too early:
It is not a coincidence that return rates converge to 90%, and there is a reason why.


Now try to come up with a reason why return rates converge to 90% that does NOT rely on random distribution in combination with the law of large numbers.

And, as you do not like that theory, a reason that does NOT require tracking of personal losses.

Because, and you simply ignore that inconvenient fact, it is not that easy to find an explanation that's left after you rule out these two things.


I guess all we will ever hear from you again on that topic is...

*crickets*

 

[Uvas]

The reason the return seems to be 90% is because the loot equation(s) are designed to give that average.

And wow....magic!

The average return is 90%

As to theories I don't like, they are the conspiratorial kind that offer not the slightest shred of evidence to back them up.

 

[wizzszz]

The reason the return seems to be 90% is because the loot pool equation(s) are designed to give that average.

And wow....magic!

The average return is 90%

As to theories I don't like, they are the conspiratorial kind that offer not the slightest shred of evidence to back them up.

You have completely ignored my objections, good job.

 

[Uvas]

You have completely ignored my objections, good job.

Your objections are pointless. That is why I didn't reply to them.

The loot equations control the loot outcome pure and simple. There is no need for anything more.

There is no personal loot pool. Proof?

Just to clear up some apparent confusion and misunderstandings related to the Personal Lootpool portion of Developer Notes #2-- Entropia Universe has never had any kind of so called personal lootpool...

You have the burden of disproving this.

So far I have not seen ANY data from you to even start to disprove this statement.

 

[wizzszz]

Find an explanation that matches the prerequisites outlined above, everything else is irrelevant and only serves the purpose to distract from you being unable to find one.

 

[Uvas]

Find an explanation that matches the prerequisites outlined above, everything else is irrelevant and only serves the purpose to distract from you being unable to find one.

Find some data that refutes the statements of the game developers. All other things are just you trying to distract from the fact that you have nothing.

 

[wizzszz]

Find some data that refutes the statements of the game developers. All other things are just you trying to distract from the fact that you have nothing.

What a lame excuse - i challenged you first you, owe me a reply before you try to challenge me.
Everything else is just distraction from the fact that you can't.

But who am i trying to fool here, you won't provide a meaningful answer anyway.
Back on ignore i'd say.

 

[Uvas]

In other words, what should I logically think is more correct?

A statement made by the game developers, with which there are no internal ingame discrepancies.

Or...

Believe in a forum troll who puts forward a conspiratorial theory without a shred of evidence to back his claims.


Not a very hard decision.

I think I am done with you.

 

[Skaffen]

Sigh Fi

There is a little flaw left: You can't prove your point, nor can you disprove mine.





You stopped thinking too early:

Science demonstrates that there is no facts - only theories of increasing approximation (if useful). The 'reason' for this derives from the observation that underneath the Classical is the Quantum, and it is probabilistic. I suggest you start thinking in a wider context and adopt a flexible approach to what you think you understand.

This 'new' way of thinking is ~100yrs old, it is time you caught up.

 

[aridash]

There is no personal loot pool.

The 90% return is just an overall average return that seems to be the median point.

There is no tracking of when someone goes over or under the average. The average return of all players seems to be around 90%.

*individual returns may vary*

If you are currently far below average, odds are that future returns will be higher than your current average, so returning you closer to average return over the long run. Reverse, if you are doing very well.

No need to track anything beyond what went into killing the mob.

i think this is obvious conclusion from the evidence. there's little point clinging on to "personal loot pool" idea when it simply isn't necessary. some might like to say that means its random... guess what, it most certainly is! obviously with boundries, blah, etc. just because its officially not considered gambling, it doesnt mean there aren't random elements which escaped the authorities attention or concern. and why is one MA statement to be denied and another MA statment held as sacrosanct anyway?

everything i see in the data and adacdotal evidence above, especially tamlin's comments (assuming i understand correctly that non-support hunting is "normal"), seems to reinforce this outline hypothesis.

 

[mrproper]

You have the burden of disproving this.

So far I have not seen ANY data from you to even start to disprove this statement.

Empty affirmations can be dis-proven with.... empty affirmations!

So it is equally acceptable to have an opposing opinion of same value and quality to a statement made by an official.

If you want ANY data on the counter-argument, I am going to ask YOU, where is the data associated with the quoted statement?

 

[Uvas]

Empty affirmations can be dis-proven with.... empty affirmations!

So it is equally acceptable to have an opposing opinion of same value and quality to a statement made by an official.

If you want ANY data on the counter-argument, I am going to ask YOU, where is the data associated with the quoted statement?

An official statement made by a game designer stating that something doesn't exist.

A forum troll with a conspiracy theory.

Hmmm, these two statements have equal weight in your mind?

 

[wizzszz]

This thread has become a farce - people are using whatever pops up as half-proven fact to support their position and to trash everything else - no matter how illogical.

Not everyone, but well, you should know who you are.

But the worst is the "wimp out when asked to give an alternative explanation that matches the few facts"... and wrap their posts up in pseudo-logics and silly attacks to disguise that they simply can't provide a better explanation.

Waste of time - whoever has a legitmate idea can PM me (if this should ever happen) - until then i'll leave you to the crickets...


*unsubscribes*

 

[Uvas]

This thread has become a farce - people are using whatever pops up as half-proven fact to support their position and to trash everything else - no matter how illogical.

Peeking in the mirror, are you?

 

[Joat]

i think this is obvious conclusion from the evidence. there's little point clinging on to "personal loot pool" idea when it simply isn't necessary. some might like to say that means its random... guess what, it most certainly is! obviously with boundries, blah, etc. just because its officially not considered gambling, it doesnt mean there aren't random elements which escaped the authorities attention or concern. and why is one MA statement to be denied and another MA statment held as sacrosanct anyway?

everything i see in the data and adacdotal evidence above, especially tamlin's comments (assuming i understand correctly that non-support hunting is "normal"), seems to reinforce this outline hypothesis.

You are allowed to have random elements in features, as long as the general impression
of the game isn't seen as if it is based on randomness.
This is balanced with features based on a ingame performance done by the user.
It's this balance that will decide if it's gambling or not, except that the random part
must always be before the performance, otherwise it will be judged as gambling.

 

[Uvas]

Empty affirmations can be dis-proven with.... empty affirmations!

So it is equally acceptable to have an opposing opinion of same value and quality to a statement made by an official.

If you want ANY data on the counter-argument, I am going to ask YOU, where is the data associated with the quoted statement?

Wizzizzz or whatever is disagreeing with official statement of game designers.

He has to provide some sort of evidence to make his disagreement something other than a trollish remark.

I am just pointing out that he has no evidence that the game developers are lying.

 

[wizzszz]

Wizzizzz or whatever is disagreeing with official statement of game designers.

He has to provide some sort of evidence to make his disagreement something other than a trollish remark.

I am just pointing out that he has no evidence that the game developers are lying.

And you are agreeing with the official statement of the game developers.

But you can neither prove that they are not lying nor can you prove that we do not misinterprete what they said.

Makes your remarks as much "trolling" as mine, and the burden of proof is on you as much as it is on me.
When are you going to provide any evidence that your interpretation of the dev notes is correct?




There is a significant difference between my posts and yours:

I made an effort to explain how i came to all my conclusions.
You, in contrast, just claim that you are right.


I want to find out HOW it works.
You, in contrast, just want to be the one who's right.



Don't bother replying, there's nothing more to add from my side.

 

[falkao]

Nice analysis. Surprising result to me, but does certainly look ...

Have this to quote again. I've checked the analysis of Qwackers data.

He was testing LR53(L)+A204 vs ImpMKII in about 580 kills per weapon. Ammo per dmg of LR53 setup is 17.4 whereas ImpMKII has 20.8. This leads to a statistically significantly difference in mean ammo cost per kill of 3.51 PED with ImpMKII and 2.88 with LR53 (p<.001).

With the new results ImpMKII should therefore lead to higher loot but it seems as there is no significant difference. Mean hp dmg was the same and hence results are not consistent.

Btw. Mean total weapon cost was lower with ImpMKII (4.05 vs 4.41 with LR53 p<.001) and therefore LR53 leads to lower TT%.

It could be that DLxE is only an exception. Hence, further studies are necessary.

 

[Uvas]

And you are agreeing with the official statement of the game developers.

But you can neither prove that they are not lying nor can you prove that we do not misinterprete what they said.

Makes your remarks as much "trolling" as mine, and the burden of proof is on you as much as it is on me.
When are you going to provide any evidence that your interpretation of the dev notes is correct?




There is a significant difference between my posts and yours:

I made an effort to explain how i came to all my conclusions.
You, in contrast, just claim that you are right.


I want to find out HOW it works.
You, in contrast, just want to be the one who's right.



Don't bother replying, there's nothing more to add from my side.

There is no personal loot pool. The game designers have stated this.

You have offered nothing of substance to make me believe otherwise. The people who created this game says there is no such thing. You offer sophistry.

You say they must be lying because....because....because....you want to believe in a personal loot pool. Fine.

You can believe in whatever you want, but if you want to convince anyone else, you have to come up with a little bit of evidence to support your theory.

And cutting down other people, making unsubstantiated slanders, does not constitute evidence for a personal loot pool.

 

[Wollongong]

Wow... this is STILL going on...?

 

[FallenAngel]

It means you should be sceptic, too, in case something has indeed changed.
This is no fallacy, but what you do is close to one - disregarding possibilities because you simply do not like them is a cocksure way to go wrong.

You might say that i am just being paranoid, but then you are calling Xen (and everyone who wants to be on the safe side) paranoid, too...

"Better dafe than sorry" is a known concept to you?

Scientific data, especially empirical data, cannot be confirmed enough.

Xen only qualified his data with the statement that he has know idea if the data is still relevant, i do not know if the sun will rise tomorrow but im not betting money on the fact that it may not rise?

Possible only if you think MA lies and the swedish authorities do so, too.

While this is indeed a possibility, we should start somewhere and take some prerequisites as granted.
If you don't like them, please elaborate on why you think the swedish authorities might lie and we might take it into consideration. (i.e. bribery?)

That coefficient is just another way to express that loot is based on your past losses - another term for the same mechanism is "personal loot pool".
You didn't even realize that you are argueing in favour of "MA tracks long-term loots and compensates players", did you?

you know that the swedish authorities only made a limited examination of entropia, i am not the one claiming MA is lying, it is still possible for there to be random elements to the process without it being gambling, you can play blackjack or poker as if it is gambling or you can use the laws of averages and large numbers.
If you are looking for work and turn up for interviews unprepared and unwashed in dirty old clothes and you keep not getting a job it is easy to think its all just a gamble, if you turn up prepared and are qualified even if you dont get the job you know you did everything right and it wont feel as if its a gamble knowing that the next interview could be the one where you succeed.

maybe its a language thing, for me a personal loot pool implies everyone is sitting in their own puddle containing only their deposits and peds gained in game (trading etc.) and you can never gain more than 90% (or whatever) back from the peds you put into the game. A hybrid personal loot pool (the bastard son of the personal loot pool? ) would be where some of the losses you make go into a general pool from which it is possible, if rarely so, to gain more than your original input, minus MA's costs.

Anything else is everyone swimming in the same pool imho.

It is very well supported by facts i have seen.
And i cannot comment on whether it has changed recently or not, no time to play atm - but i said that in my last post already.

In the end you offer nothing but sophistry, you have seen facts you refuse to share, you have no empirical data to prove your point, you have no time to play so have no idea regarding loot/returns at this current time yet you insist something is different now.
In the end what should guide me, the unsupported claims of a sophist, or the analysis of data that remains unchallenged and confirms quite nicely with my own observations in the field, and my current observations are based upon logging the TT returns of the 10K drops i do a month?

i think this discussion has run its course dont you?
and since Falkao is doing interesting work with real data i think my time would be better spent trying to understand what that means instead of "debating" with you?

 

[rhogenbe]

Small first run of data on Hogglo Young 2000HP (tax 4%)

Profession levels:
BLP Pistoleer 82.82
Ranged BLP 77.77
BLP sniper 71.67

Maddox IV + Beast. Cost 14.866 pec a shot

mob id;# hits;# misses;total damage done;loot
1;46;4;2069.9;6.91
2;46;6;2068.1;1.31
3;42;3;2072.3;5.28
4;45;6;2048.1;2.02
5;47;1;2109.1;7.04
6;43;4;2086.0;7.88
7;44;6;2110.6;4.81
8;41;6;2083.8;3.17
9;39;3;2072.0;1.30
10;46;5;2072.0;4.13

Swine Deluxe + Evil. Cost 53.950 pec a shot

mob id;# hits;# misses;total damage done;loot
11;18;5;2067.3;1.57
12;18;1;2055.6;7.72
13;17;1;2091.8;7.93
14;18;4;2099.4;35.69
15;19;1;2171.8;2.83
16;18;0;2037.7;2.64
17;18;4;2083.6;3.87
18;17;3;2115.3;9.29
19;16;1;2061.2;8.26
20;17;3;2073.9;8.23

I will do some more kills this evening and then post that data also.

Cheers
Siam

 

[wizzszz]

you know that the swedish authorities only made a limited examination of entropia, i am not the one claiming MA is lying, it is still possible for there to be random elements to the process without it being gambling, you can play blackjack or poker as if it is gambling or you can use the laws of averages and large numbers.

Now it's geting really absurd - did you really claim that the authorities didn't do a proper job only to support your theory?

Look i have already said that, it's not a rule-of-thumb thingie where the gambling commission decide based on how they feel that day.


If your theory requires that the swedish authorities are dumb or inept and my theory requires that we misinterprete MAs dev notes... i will stick to my theory, tyvm.


And btw - if you need the help of the law of large numbers to explain the observed loot pattern it IS gambling.

In a random loot distribution there would be ups and down all around the 90% percentile.
In a tracked personal loot you see different patterns: the ups and down are all situated below or directly on the 90% percentile.


I do not need to provide any data, because the forum is full of graphics that show this, you are just too lazy to search - I have read these threads back then, you obviously didn't.

 

[tamlin]

Hmm, i disagree, despite Occam.

Let's assume that every mob you loot has 90% of the ammo spend in loot (average)
This is independent from who spent the ammo.

Now what happens when the mob has less than the average loot?
It gets tracked somewhere and repaid later in a mob that has more than average loot.

It seems we are in agreement thus far. I think we have different views on the following.

This is "loot tracking", or more commonly known as "personal loot pool".
It does not matter who spends the PEDs, the not-paid-out loot goes somewhere and will be (has to be?!!) paid out later, to achieve the 90% average loot.

It doesn't have to be tied to YOUR losses directly, but to "how much loot was missing from that mob from the 90% of it's kill costs.

I agree on the latter part, but that does not imply the following:

That a mobs loot increases when you blast more PEDs into them already proves that they (short-term) track your spendings - and this is exactly as expected, as they have to do that anyway, to determine who is allowed to loot a mob.

This is not proven.

It could rather be as simple as MA instead keeping track of money spent on a specific mob in a specific spawn, maybe even during a specific timeframe (say 24h for simplicity).

That implementation would not only be way easier (on the system), but could also result in effectively exactly the same behavior if you hunted alone (no other ava's affecting that spawn).

I'm definitely not claiming this is the way it is, but this is a (very) simple way to implement it, needing very little tracking, and still allow for the observed behavior to closely match a potential "personal pool".

Coming to think of it, I'm sure we can test this. What about doing a small-ish run on some smaller mob, simultaneously on a spawn "out of harms way". Would be awesome if we could switch weapon (-types) midway too, to something with same-ish dps but different ammo/decay ratio. Drop me a PM.

 

[wizzszz]

This is not proven.

It looks like you have missed Xens test:

Here's my test:

100 kills with each weapon, alternating each kill.

Pretty clear pattern imo.

It is short, i know, too small to really prove it, but it's not the only test, i.e. gravedigger made a similar one years ago, and everything points in the same direction: Loot is based on PEDs spent to kill it.

It could rather be as simple as MA instead keeping track of money spent on a specific mob in a specific spawn, maybe even during a specific timeframe (say 24h for simplicity).

Even if Xens test shows something different, let's just think about your suggestion for a minute:
How would you distribute loot then?
Random?

That implementation would not only be way easier (on the system), but could also result in effectively exactly the same behavior if you hunted alone (no other ava's affecting that spawn).

I'm definitely not claiming this is the way it is, but this is a (very) simple way to implement it, needing very little tracking, and still allow for the observed behavior to closely match a potential "personal pool".

The "too hard to implement personal tracking" or, as you word it "way easier to do it differently" assumption pops up every now and then, to give one of the theories more weight by further assuming that MA didn't want to or was unable to implement a complex system.

This is WRONG.
I cannot say it more clearly.

There is nothing complicated in coding any of these systems, neither a mob based nor an avatar based poll.

However, as long as you do not comment on how the accumulated loot is distributed i am missing a vital piece of the puzzle - even though i see no sense in having a mob based loot pool and distribute it (assuming oyu meant that) randomly (a.k.a. gambling)

Coming to think of it, I'm sure we can test this. What about doing a small-ish run on some smaller mob, simultaneously on a spawn "out of harms way". Would be awesome if we could switch weapon (-types) midway too, to something with same-ish dps but different ammo/decay ratio. Drop me a PM.

Not sure if i understood that correctly - hunting side by side to eleminate server and time based effects, the switching weapons in between (per kill) with same dmg/sec but different distribution of costs (ammo vs. decay)?

To prove what exactly? That only ammo determines loot and decay is ignored?
Any idea which weapons we could use? There is a plasma handgun which does not even decay at all, this would be perfect for the test, but i don't know who owns one.


But yeah, lets do that, but you should post beforehand what the expected results would be to prove a hypothesis A and how would the result look to be a valid counterproof (proof for hypothesis B)?

 

[falkao]

Small first run of data on Hogglo Young 2000HP (tax 4%)

Maddox IV + Beast. Cost 14.866 pec a shot
vs
Swine Deluxe + Evil. Cost 53.950 pec a shot

atm it looks like Swine will lead to higher loot. The mini in Swine does not influence that much.

Swine vs Maddox IV (p=.123, ns)