Hunting Loot Theory: Interconnectedness of all Things

Joined
Jul 3, 2008
Posts
4,439
Location
Sweden
Society
Death Unlimited
Avatar Name
Johan Roadkill Deadmeat
I'm not as active anymore (and have not been for the past few years) but I've long had an idea on the basic concepts of loot and how to better predict it. The way loot has worked has largely remained the same over the years, at least fundamentally.

I have though to test this for a while but I don't play enough to gather relevant data so I will just post my thesis and how I imagine testing it would work and if anyone feels up for it go nuts.

Disclaimer:
Few of the things I theorized have already been largely proven and few might have been disprove. I obviously can't say much of how loot looked prior starting in 2007 and my memory is not accurate of exactly how loot looked and how it's changed since then so take it with a grain of salt. This is also being presented as a simple case, the system is most likely more complex.

I will state my believes more as if they were facts to improve readability and brevity

Loot the fundamentals:
Loot is finite and every component has a set amount of resources​
Loot amount is a constant, ideally this constant would be a base value multiplied by active player but a perfect market does not exist and there is no good way of determining the level of participation as an uber does not cycle as much as a noob. This could either be a complex algorithm or a manually tweaked value high enough that it's never (or almost never)0.

The reason the value could be 0 is that loot is constantly being cycled (sold to TT, crafted, refined) so unless all activity stopped we would never feel it other than as a diminished return of that resource when those resources were being horded before being cycled.

Resources are divided by random, probability and by set parameters​
Loot is divided by planet (+ space) and the amount is decided by random. The remainder is later distributed by mob at random and by multiplying parameters. These parameters could also be random but a big probability is that HP is always a multiplying factor as a large hp mobs needs more allocated resources. So far neither outcome is predictable.

Creature loot​
If a planet is distributed a large amount of say thyroid it would mean that any creature with a high thyroid yield has a better probability of good loot however this may or may not be true for all mobs that normally drop thyroid, we'll see why.

Example:
[mobA's loot is on average for 1 sizable run: (45% shrapnel) (15% thyroid) (40% other)
("1 sizable run" is what we assume is enough total tt spent for a sample)]


Case 1: Each mob has a set range [1% - 100%] for how much their loot consists of e.g:
Loot range for a set component has an ideal value that is influenced by other parameters:
This means that a mob will without any multipliers and a "full" pool deliver e.g 15% thyroid. If the pool is not full it obviously can't but multipliers such as bigger loots, globals and hofs still might deliver a higher yield, but is at that point less likely. The reason for this system could be that it allows a player to farm mobs for resources that are know for a higher base yield.

to test this you would need to record loot thoroughly over time for at least 2 mobs having similar loot that isn't shrapnel at the same point in time, the second highest loot % after shrapnel would probably be best.

We would see a low dynamic range in mobs (at different days, locations and players at the same time) but a higher dynamic range between mobs with similar loot, e.g mobs B could have a 90% thyroid drop average and mob A 15% so B would on average always drop more.

Case 2: loot range is decided on whatever is available in the creature pool
This system would be more dynamic relative to the entire universe as a whole market. It could mean that a high planet thyroid yield could give amazing loot to one mob and shitty to another but no on average over time. The way loot is later distributed could be any number of factors. The same is also true in case 1 but with a base value it would be less dependent on these factors to produce a result. E.g if we assume location(planet scope) makes a difference, case 1 would still have it's base value so the influence on loot could be less than in case 2 where location could be a bigger deciding factor.

using the same test as case 1 you would see a higher dynamic range in the same mob (at different days, locations and players, at the same time) but less dynamic range between different mobs with similar loot, e.g Mob A might have a high thyroid allocation 1 week and then less the next making the average difference between A and B over time less.

Seeing the difference could be very hard depending on other deciding factors, an example would be popularity.
In case 1 popularity could skew an attempt at deciding base return as it would almost always produce more than base value x1 and in case 2 it could have a much higher chance of getting a sizable allocation compared to other thyroid mobs.

What does this mean?​

It means that observing the universe is much more important than grinding if you wish to improve your chances.
If we disregard most of the probable 1000 or so influencing factors that creates this living market(with more testing the bigger factors could be guessed) I'll give a you a simplified example:

Crafting day:
This is a term i coined probably back in 2008 already and it is the seed for this theory. It means that by observing crafters and markup you could make an accurate guess of what materials were being horded and prepared to be used in crafting. So if you saw a lot of globals or auction activity for say simple I conductors then the pool could be drained of muscle or eye oil because a lot of it has been sitting in storage ready to be cycled back.
This day would not have been a good day to hunt mobs that drop a lot of muscle or if you saw an increase in crafting over all then perhaps not a good day for hunting anything.
Generally the next day some resources would drop a bit in value, crafting globals would go down and hunting globals would go up (by my casual observations).

Another thing:
If I remember correctly at some point MA made loot less dynamic between mobs and they dropped much of the same stuff (they don't anymore but correct me if I'm wrong), this could have been a poor attempt to make loot a bit better overall.
Imagine case 2: if you were hunting a mob with low resource allocation where 20% of the loot would have been thyroid, but at the same time a big part of the thyroid were being tied up, you would probably see terrible returns.
I believe this was later balanced with more ammo, components and now shrapnel so it's no longer such a concern

Loot 2.0​
As I mentioned I don't play enough lately to make any observations on how much of this has changed but as for shrapnel I think one of it's uses is to balance loot so that when you are hunting "the wrong mob" you are not as penalized and probability might still save you, obviously hunting the right mob will always have a higher yield probability


How does this help​

If case 1 is true then an average loot drop % for each component could be calculated and by looking at trends you could quite accurately predict which mobs have a better chance of loot.
if case 2 is true then it would be harder to predict which mobs but it would lessen your options, if thyroid is a trend then you simply have to try different thyroid mobs.
ofc which each deciding factor that you can find out or guess the more accurate it would become.

Factors​
Few factors I believe influence your loot(definitely not a complete list and an absolute guess)
Obviously: planet and mob,
mob type,(mutant, robot etc),
spawn location,
cost to kill,
time(meaning resource allocation between mobs would switch and recalculated at random time intervals),
popularity((mobs killed,tt spent etc)/time both global popularity and local (and also planet if cross planet species)),
looter professions,
weapon(weapon type, range, dps),
damage taken
etc etc

Since I've never written all of this down I've probably forgotten a lot of good thoughts but I leave all of this for you guys to add to and do with what you wish, cheers.

If any of you wish to test this (would need to be multiple people and a lot of peds cycled on a lot of mobs) I might be up for tailoring an app to automate as much of it as possible, to suite your plan of action(but no promises).
 
Back
Top