Loot Abundance broke our books.

[MAD World Server]

I run a public server and since updating to b173, Loot abundance is broke. I set it it to 300% because i dont want people constantly looting and actually creating some PvP scenarios. Raiding each other bases. but setting it to 300% means now that in book cases if a book spawns it spawns 3 of them. I don't really quite know how to fix this issue. I dont want every group in my server to be able to find the book they want and then be able to give the copies to friends... Would kinda ruin levelling. 

 

Thanks for any help in advance.  

[SylenThunder]

This is not a bug or an issue. This is precisely because you set the loot to 300%.

 

Increasing the loot % does not increase you chance of finding items. It increases the quantity of items in stacks.  So at 100% loot, if there is a chance for a book to drop, you get 1 book. Set to 200% you get 2, and 300% you get 3.

 

To fix the "issue" set your loot back at 100%.

[MAD World Server]

In alpha 18 This wasnt the case though... For books, 1 book would spawn... for everything else it increased the stack size of lootables that made for a more enjoyable game. 

[katarynna]

2 hours ago, MAD World Server said:

In alpha 18 This wasnt the case though... For books, 1 book would spawn... for everything else it increased the stack size of lootables that made for a more enjoyable game. 

not everything else. in a18 it was possible  (likely even) to get 1 egg from a nest with 200% loot abundance, which would mean (if it was working correctly) that at 100% i would have gotten 1/2 an egg.

 

17 worked the same way it does now in 19. 18 was just a bit broken in that regards.

[Liesel Weppen]

On 7/29/2020 at 1:09 AM, katarynna said:

not everything else. in a18 it was possible  (likely even) to get 1 egg from a nest with 200% loot abundance, which would mean (if it was working correctly) that at 100% i would have gotten 1/2 an egg.

Just my thoughts. If loot abundance doesn't affect the chance of items, but their amount, i wonder what you could loot with 75% loot abundance... No not-stackable items anymore, because you can't drop a 0.75 item?

But i can confirm, with 75% oder even 50% you can still loot books, mods, tools and so on. There don't drop half books.

[meganoth]

@Liesel Weppen This is a necroed thread

 

[Liesel Weppen]

I knew, but still an interesting question and probably with a wrong answer in the end.

[katarynna]

4 hours ago, Liesel Weppen said:

Just my thoughts. If loot abundance doesn't affect the chance of items, but their amount, i wonder what you could loot with 75% loot abundance... No not-stackable items anymore, because you can't drop a 0.75 item?

But i can confirm, with 75% oder even 50% you can still loot books, mods, tools and so on. There don't drop half books.

 

Settings above 100% do indeed work like that. I am not sure how settings below 100% work.

 

There have been some threads reporting never finding workstation schematics in loot at 25% either here, on steam, or both. Perhaps the game rounds up if it rolls 50% or more? 

 

I personally have no clue.

 

To test 200%, 300% etc on books would be simple though, if you are not sure that part is accurate. I can confirm in my experience 200% loot ALWAYS gives 2 copies of each book, schematic, and recipe.

[Liesel Weppen]

28 minutes ago, katarynna said:

There have been some threads reporting never finding workstation schematics in loot at 25% either here, on steam, or both. Perhaps the game rounds up if it rolls 50% or more?

Correct, i've seen dozens of those reports too. Also infact bug reports here in the forums. However the lowest loot abundance i played with was 66%.

25% is pretty low and since some blueprints are rare anyway and since loot is still a dice roll, "i haven't found yet"-reports are relatively non-saying. I also had playthroughs with 100% where i haven't found neither a crucible nor the receip up to day 50.

 

And that is exactly the point, that makes that question very relevant. Affecting just the amount, and furthermore if it rounds up or down or just truncates the result imho is just wrong and should be fixed.

 

Maybe someone can do the math? If an item has 1% chance to drop 1 item and a loot abundance of 25% does NOT reduce it to .25% by the roll itself but instead rolls as usual with 1% chance but reduces the outcome amount afterwards by 0.25... The chance of getting that item is indeed exactly zero?

With loot above 50% and IF it rounds up, the chance onb 1-drop items in the end is still 1%, so it does nothing on items that are supposed to drop only 1 amount?

 

Amount and chance are two essentially completely different things.

Edited April 8, 2021 by Liesel Weppen (see edit history)

[Boidster]

I can try to do some testing, but I think the consensus (?) view that Loot Abundance affects quantity, not chance is consistent with all reported results. Based on previous testing related mostly to chance, but not quantity, I think this is the likely process upon opening a new loot container:

 

1) Run the loot tables using player's loot game stage (chance)

2) Award N items for each item selected in 1, also according to loot tables (raw quantity)

3) Apply Loot Abundance modifier to N, and round the result (adjusted quantity)

 

If that is correct, then any loot item with a quantity value of either "1" or "0,1" in the loot tables will never show up in a container if Loot Abundance is less than 50%. I believe books and schematics fall into that category, usually. However there are exceptions:

 

<item group="perkBooks" count="1,2" prob="0.015" />

 

That particular loot entry could possibly choose twice from the perkBooks list, and if it landed twice on the same book, it is possible (I have not tested this; not even sure how to besides brute force dozens of searches) that raw quantity=2 and even after a 50% or more reduction you'd get 1 book. At a 75% reduction you'd be back to 0 books always.

 

On the other side, with say 300% Loot Abundance, any time a Book is selected in step 1 you'd be guaranteed to get 3x that book (possibly 6x if the same book were selected twice).

 

Edited April 8, 2021 by Boidster (see edit history)

[meganoth]

I think a dev once explained that a 50% loot chance would be handled correctly in that an internal dice roll would decide if an item drops or not.

 

Whether that is true, false or bugged only a test can show

 

[Boidster]

18 hours ago, meganoth said:

I think a dev once explained that a 50% loot chance would be handled correctly in that an internal dice roll would decide if an item drops or not.

 

Strictly speaking about the chance of an item being selected as loot, I think that is correct. My own testing has shown the loot probabilities work as expected. What I think is happening with the Loot Abundance setting, specifically, is that it applies after all of the "chance" calculations are done, and it only applies to the quantity of items dropped. In other words, it's not a probability/dice roll thing, it's strictly "multiply the final results by X% and round".

 

I will do some testing to see if an exact 50% loot abundance allows qty=1 items to drop. I'll try 49% too, to confirm the rounding behavior.

 

--- an hour later ---

 

Well, I'm not sure how I'm going to set up the test, because the "count" attribute of a loot item doesn't seem to be working the way I'd expect. I modified the Bookcase loot conatiner like this:

 

<!-- bookcase -->
<lootcontainer id="49" count="1" size="6,2">
    <item name="resourcePaper" count="1" prob="1" />
</lootcontainer>

 

This was just a test case so I could confirm basic behavior. The expected behavior is to find exactly 1 piece of paper in each bookshelf. Instead I found about 5-8 pieces of paper in each bookshelf. It was always only paper (never books), so I know the mod did correctly restrict the results to only paper, but for some reason the "count" attribute is not working the way the XML documentation says it should. I even tried just removing "count" altogether:

 

<!-- bookcase -->
<lootcontainer id="49" size="6,2">
    <item name="resourcePaper" prob="1" />
</lootcontainer>

 

Still getting 5-8 paper in every bookshelf. I can't really test Loot Abundance if I can't reliably limit a loot container to give exactly 1 item every time. I think this might even count as a bug, but first I'm going to go dig into resourcePaper a bit. Maybe there's something special about it.

 

--- 30 minutes later ---

 

Well there were a couple of other bookshelf-related loot settings (apparently for top half/bottom half bookshelves) and I changed all of them to use the above loot container. Still getting 5-8 paper every time. <shrug> Giving up now.

Edited April 9, 2021 by Boidster (see edit history)

[Naz]

I did some testing a while ago, recording the results of looting hundreds of containers, book piles and bird nests, 100%,25%, 300% etc Loot abundance. 

It's something they're still working on, is the good news. The bad news is the system has been Wonky for years at this point. It never behaves how you would expect and is fundamentally broken under 100%, so I can definitely relate to everyone's frustration with the Loot at times. It's been reported alpha after alpha but the game is early access so all we can really do now is hope the pimps eventually come through and give it a final polishing that remedies the issues. 

[Boidster]

I haven't read closely through your results, but what I saw on the 25% abundance actually lines up with what I'd expect, if the system works as I described above. That is, Loot Abundance only affects the quantity of items you get and only after all other checks (basic loot prob + perks) have been made.

 

So for example, a bird's nest has a very small chance (about 6% 3%) of producing 2 eggs under normal conditions. Since bird's nests aren't scaled with gamestage, loot-boosting perks don't affect it.

 

Since Loot Abundance only affects the final quantity of items after all other checks have been made, when it is set to 25% the only way to get an egg in a nest is for the RNG ("all the other checks") to land on the "2 eggs" result. The Loot Abundance then multiplies that by 25% resulting in 0.5 eggs which rounds up to 1 egg.

 

I would expect that to happen approximately 6% 3% of the time, and for 100 trials getting 4 eggs (at 25% Loot Abundance) makes sense to me.

 

(Edit - fixed chance of getting 2 eggs; 3%, not 6%)

Edited April 12, 2021 by Boidster (see edit history)

[meganoth]

There already seems to be a routine in there that handles fractions of loot results. If you are right and rounding is used there the fix would be to simply change that to use that routine instead.

 

[Naz]

11 hours ago, Boidster said:

Since Loot Abundance only affects the final quantity of items after all other checks have been made, when it is set to 25% the only way to get an egg in a nest is for the RNG ("all the other checks") to land on the "2 eggs" result. The Loot Abundance then multiplies that by 25% resulting in 0.5 eggs which rounds up to 1 egg.

 

I would expect that to happen approximately 6% of the time, and for 100 trials getting 4 eggs (at 24% Loot Abundance) makes sense to me.

Only the devs know for sure, but if you're theory is on point. The system may indeed be 'functioning' as intended but produces results that are easily mistaken as bugged/broken. 

 

When it comes to the quantity of Loot, is it shippable to have a Loot setting that scales completely differently depending on which side of 100% you go (the end results that is) . 

 

If you set Loot to 125% and get roughly 25% more eggs than 100% which is what I think is reasonable to expect. Why should the system scale the results differently on the other end of the scale. If I set 25% am I mad to expect 75% less Loot? Instead of 90-95% less? 

 

Same with perk boosted Loot modifiers, if I'm playing 25% naturally I would go for those perks to max out my chances, but because of the way the Loot is calculated, all those perks and buffs end up doing nothing.

 

So at that point I think it's not whether or not the system functions correctly, it's is it good design to have systems that dupe players into wasting their time on gameplay that actually does nothing for them? 

[Boidster]

On 4/11/2021 at 8:08 AM, Naz said:

If you set Loot to 125% and get roughly 25% more eggs than 100% which is what I think is reasonable to expect.

 

You would not get 25% more eggs at Loot Abundance of 125%, for the same reasons outlined above. Eggs can only be given in integer amounts, with a minimum of 1 and a maximum of 2 in a single nest (vanilla settings). With 125% LA, the only time you would get an extra egg is when the RNG lands on "2 eggs" (about 6% 3% of the time) in which case you'd actually get 3 eggs. The rest of the time, you'd get 0 eggs or 1 egg as usual (because there is no way to take the "1 egg" result and apply a 125% result to it; it just rounds down to 1 egg).

 

It sounds like what you and the other "duped" players are looking for is some sort of system that tracks global loot acquisition across all containers and smoothly adjusts the total items you will get across your player's lifetime. So if you might normally get 100 eggs after playing for 10 in-game days, you want to see that at 75 eggs if Loot Abundance is set to 75%. Unfortunately there is no easy way to do that, since loot in this game is not fixed - the containers are not pre-filled with stuff which can be adjusted behind the scenes before you open any of them. Each container is its own self-contained set of RNG results, which are then modified by LA. For relatively high-quantity-per-container items like paper or feathers, you could probably get pretty close, though as the LA restriction gets smaller (25% say) it gets harder to provide an accurate (to the player's mind) result.

 

Note that the same is true for LA values above 100%, as explained above for eggs @ 125%. For large-quantity-per-container items, the LA should track pretty well at least down to 50% and as high as you want. For low-quantity items (eggs, books, guns) it's not going to track very well at all, except at 200%, 300%, etc.

 

(Edit: fixed % chance of getting two eggs in a nest - it's about 3%, not 6%)

Edited April 12, 2021 by Boidster (see edit history)

[meganoth]

9 hours ago, Boidster said:

 

You would not get 25% more eggs at Loot Abundance of 125%, for the same reasons outlined above. Eggs can only be given in integer amounts, with a minimum of 1 and a maximum of 2 in a single nest (vanilla settings). With 125% LA, the only time you would get an extra egg is when the RNG lands on "2 eggs" (about 6% 3% of the time) in which case you'd actually get 3 eggs. The rest of the time, you'd get 0 eggs or 1 egg as usual (because there is no way to take the "1 egg" result and apply a 125% result to it; it just rounds down to 1 egg).

 

It sounds like what you and the other "duped" players are looking for is some sort of system that tracks global loot acquisition across all containers and smoothly adjusts the total items you will get across your player's lifetime. So if you might normally get 100 eggs after playing for 10 in-game days, you want to see that at 75 eggs if Loot Abundance is set to 75%. Unfortunately there is no easy way to do that

 

Wrong. The 75 eggs would automatically happen if any egg found would only be found with a chance of 75%. In other words a system like you described above but instead of rounding a percentage chance is used for determining if something is found in loot or not

 

9 hours ago, Boidster said:

 

since loot in this game is not fixed - the containers are not pre-filled with stuff which can be adjusted behind the scenes before you open any of them. Each container is its own self-contained set of RNG results, which are then modified by LA. For relatively high-quantity-per-container items like paper or feathers, you could probably get pretty close, though as the LA restriction gets smaller (25% say) it gets harder to provide an accurate (to the player's mind) result.

 

Note that the same is true for LA values above 100%, as explained above for eggs @ 125%. For large-quantity-per-container items, the LA should track pretty well at least down to 50% and as high as you want. For low-quantity items (eggs, books, guns) it's not going to track very well at all, except at 200%, 300%, etc.

 

(Edit: fixed % chance of getting two eggs in a nest - it's about 3%, not 6%)

 

 

 

[Boidster]

14 hours ago, meganoth said:

Wrong. The 75 eggs would automatically happen if any egg found would only be found with a chance of 75%. In other words a system like you described above but instead of rounding a percentage chance is used for determining if something is found in loot or not

 

I said there wasn't an easy way to do it, but it seems you have an easy way, which is great. Can you tell me where, exactly, in this process you apply your 75% chance, and how it works for 125%? Obviously you'd want to eliminate step 3 entirely; I just put it here to show the current loot algorithm as I understand it.

 

Search Bird's Nest

1. groupBirdNestNestMain

  1a. Add 1 feather to loot

  1b. Process groupBirdNest

2. groupBirdNest

  2a. RNG check

    2a.1. <= 0.25: add Egg to loot

    2a.2. >0.25: add 1-5 Feathers to loot

  2b. RNG check

    2b.1. <=0.50: Go to step 3

    2b.2. >0.50: Repeat 2a only, then go to step 3

3. Multiply Feather count and Egg count by Loot Abundance

4. Put loot, if any, into container

 

If your easy fix works here, I'd like to try with one of the book or weapon groups since they are more complex. Maybe we can just hand TFP a simple solution and get, I dunno, royalties or something. They'll smack their heads for not thinking of it, for sure.

Edited April 13, 2021 by Boidster (see edit history)

[meganoth]

I cite your previous statement, do you still stand by it?
 

Quote

 

1) Run the loot tables using player's loot game stage (chance)

2) Award N items for each item selected in 1, also according to loot tables (raw quantity)

3) Apply Loot Abundance modifier to N, and round the result (adjusted quantity)

 

 

If yes, replace step 3 (which is also step 3 in your detailed example if I understand that correctly) with

 

for each item {
result=0;
while (abundance>100%) {  result=result+1; abundance=abundance-100% }
if ( rnd(abundance)==1 ) { result=result+1 }

// rnd(k) returns 1 with a probability of k, otherwise 0.

 

This is just the easy to understand pseudocode for single items. This can be easily generalized to stacks of items, so there is no need to do that calculation for every single bullet of a stack of bullets.

 

Note that this algorithm makes sure that at 100% nothing would change. And at 200% exactly 2 items instead of 1 would drop. And at 150% there still would be the one item guaranteed, but another item would drop with 50% chance.

 

For a value of 25% abundance each item that would drop at 100% would now have only a 25% chance to drop. Exactly what you would expect from such a setting

Edited April 13, 2021 by meganoth (see edit history)

[Boidster]

1 hour ago, meganoth said:

I cite your previous statement, do you still stand by it?

 

I mean...I'm not employed by TFP so it's not like the One Known Truth or anything, but that is my understanding of how LA works, yes.

 

1 hour ago, meganoth said:

If yes, replace step 3 (which is also step 3 in your detailed example if I understand that correctly) with

 

for each item {
  result=0;
  while (abundance>100%) {  
    result=result+1; abundance=abundance-100% }

  if ( rnd(abundance)==1 ) { result=result+1 } // rnd(k) returns 1 with a probability of k, otherwise 0.
}

 

 

Ah, okay, I'm sorry I didn't grok your suggestion before. I see what you are doing here and I'm going to go test it with a couple of different loot containers.

[Boidster]

Well I tested 10,000 iterations and the actual loot quantity from your algorithm (slightly modified) matches very well with expected results. Eggs especially are a lot better at LA < 100%.

 

 

I used Excel just for ease of what-iffing a bunch of scenarios and parsing out different bits of the calculation. The expression I ended up with, which works with both low-quantity items (eggs) and high-quantity items (feathers) is:

 

(INT(LootAbundance) * ItemCount) + IF(RAND() <= MOD(LootAbundance,1), ItemCount, 0)

//ItemCount is the total # of items 'dropped' by the loot tables after RNG checks
//RAND() returns a decimal between 0 and 1

 

I expect that I've only re-invented your wheel here, but my head is completely wrapped around the idea now, so thanks. If they (TFP) do actually have a "Step 3" like I showed above, just replacing it with your algorithm or the above calc would fix it it seems.

Edited April 13, 2021 by Boidster (see edit history)