Curse

Talarian · Mar 28, 2008, 12:00 AM // 00:00

I suppose that is what is important, yes. I guess the test whether the system differentiates between AoE and non-AoE damage is another thing entirely...

joshuarodger · Mar 28, 2008, 12:03 AM // 00:03

i have been doing a lot of smite farming lately with my two mains. one a dervish and the other an ele. i always check the cash in my inventory before doing a run and again after i merch everything i get during the run, just so i can tell how profitable the run is given that i have to pay 1k just to farm there. There is not a noticeable difference in the number of ectos i collect, usually between 0 and 2 per run (usually closer to 0 than 2). Ectos are exempt from loot scaling so this makes sense. sure the derv uses AoE attacks with his scythe but he can only hit 3 targets at a time and there are usually 5 smites in a group and they usually end up dying one at a time. My ele uses AoE attacks which kill 5-15 smites at a time. Invariably when i get done merching everything my derv ends up with more money than when i started the run. My ele almost never does. The farm is quicker with my ele and since i'm mainly farming for ectos and finding just 1 every 3-4 runs more than pays for itself i still use the ele. but if i'm short on cash and wanna still do smite runs, i take my derv even though it takes longer. this is just my experience with it, but there is no arguing the point in my mind. AoE mass killing is far less profitable than single foe killing.

FoxBat · Mar 28, 2008, 12:10 AM // 00:10

I can't believe people are still arguing this. Based on their "extensive experience of farming" rather than some controlled tests.

Go to Elona Reach NM and kill the minotaurs in the following ways (while solo):

1) AoE insta-kill, I like sandstorm echo with aftershock to finish.

2) Fast single-target kill, sliver armor.

3) Painfully slow single-target kill, eg. -10 degen.

Everyone seems to be aware that 2) yields more drops than 1). But guess what, 3) yields more drops than 2). Take a break from your super-efficient-killer builds and go try it out. Every single mino should give you something if you kill them solely through degen one at a time. That won't be true with sliver armor (assuming you have at least 12 earth and a large group)

It really has nothing to do with "killing them at the same time" or "it senses AoE", it is simply a timeframe in which deaths happen. When you go over X deaths in Y seconds loot scaling increasingly applies to you. It's intentional and that simple.

Kill 3 at the same time and your drops should be decent, because your X isn't very high. Kill 30 and they won't.

Conversely, stagger your deaths with AoE damage (Ie they don't all die at the same time), and your drops will be slightly less crap. The more staggered, the more drops.

manitoba1073 · Mar 28, 2008, 12:27 AM // 00:27

Quote:

Originally Posted by Antithesis

Sorry manitoba, your tests and results prove nothing because you're comparing AoE with AoE, not AoE with Single. Your results look no different to any of my average Raptor runs.

Try that same run again the same number of times with Sliver Armor then post those results. Others have posted theirs and without fail, Sliver nets more drops than AoE.

I think you're on your own on this one. Maybe you're neither intuitive nor a loon...perhaps sh*t-stirrer is more appropriate

Actually Im done explaining it. I was getting an increase in the numbers of drops directly agaisnt what everyone else is saying. Cant blame you for not seeing the obvious.

Quote:

Originally Posted by teotuf

here is something I noticed, if I were to kill 7+ enemies at the same time, then kill another 7+ within 10 secs after the first group died, i would NEVER get ANY drops from the 2nd group (i'm talking about groups with 2 different kinds of mobs, one with higher hp and one with lower hp). I think it is more reasonable to assume that drops are assigned based on time instead of individual drops. (somewhat similar to manitoba's reasoning, but I really dont think there is a bug)

this then assumes that it would be impossible to kill a group with aoe so they die at the same time (i still don't think SoJ counts as aoe, because they don't die at exactly the same time) and obtain drops assigned to different people.

a easy way to test this (which I never got around to), is to have a group of 8, but only have the solo farmer use aoe to kill. if the drops were not reduced, but just assigned to another member, then manitoba would be right. if however, we kill a group of say 24 in a raptor farm, and NO ONE in the part of 8 gets ANYTHING, then we can conclude that AoE is affecting the drop rates, instead of something to do with the assigning of the drops.

and to disprove his theory, if in a mob is killed all at the same time, and it assigns loot to more than one person, then manitoba would be wrong. (and i guess so would i >_>)

It dont hang or hiccup long enough for a large mob to assign all the drops to one person.

Talarian · Mar 28, 2008, 01:21 AM // 01:21

Quote:

Originally Posted by manitoba1073

Actually Im done explaining it. I was getting an increase in the numbers of drops directly agaisnt what everyone else is saying. Cant blame you for not seeing the obvious.

No, you really aren't. In all the screenshots within the raptor cave, you killed the mobs with AoE. That does not prove anything.

The reason I don't like your image of the invisible henchmen is because it makes no sense, and yet is the only thing that is supporting your theory. I accept that it is possible in a full group (either it be other players or H/H) that all the drops be assigned to one person from a big AoE kill. I have already speculated about that above. What I do not accept is that there are any other parties involved for the system to bug to when solo farming. I fail to see why you are not getting this. YOU ARE THE ONLY PERSON IN THE INSTANCE, ALL DROPS ARE ALWAYS ASSIGNED TO YOU, THE LOOTSCALING CODE MEANS THAT WHEN YOU ARE SOLOING 1/8TH OF THE MOBS WILL ACTUALLY DROP LOOTSCALED ITEMS. There is nowhere for it to hang and select someone else - there is no-one else!!!!!

And FoxBat, if you had actually read through the thread, you would know that I have only been talking about incidents where there are 10+ mobs. You would also see that I have speculated on an explanation as to why SA yields fewer drops than killing foes 1 by 1 but more than all at the same time. You would in fact be aware that I stated that the system may assign death by AoE if the mob dies within, say, half a second of the last mob. Please don't just read the OP and look at the number of replies and assume we don't know what we are talking about.

joshuarodger · Mar 28, 2008, 01:29 AM // 01:29

Quote:

Originally Posted by FoxBat

I can't believe people are still arguing this. Based on their "extensive experience of farming" rather than some controlled tests.

Go to Elona Reach NM and kill the minotaurs in the following ways (while solo):

1) AoE insta-kill, I like sandstorm echo with aftershock to finish.

2) Fast single-target kill, sliver armor.

3) Painfully slow single-target kill, eg. -10 degen.

Everyone seems to be aware that 2) yields more drops than 1). But guess what, 3) yields more drops than 2). Take a break from your super-efficient-killer builds and go try it out. Every single mino should give you something if you kill them solely through degen one at a time. That won't be true with sliver armor (assuming you have at least 12 earth and a large group)

It really has nothing to do with "killing them at the same time" or "it senses AoE", it is simply a timeframe in which deaths happen. When you go over X deaths in Y seconds loot scaling increasingly applies to you. It's intentional and that simple.

Kill 3 at the same time and your drops should be decent, because your X isn't very high. Kill 30 and they won't.

Conversely, stagger your deaths with AoE damage (Ie they don't all die at the same time), and your drops will be slightly less crap. The more staggered, the more drops.

this actually makes the most sense and sounds more reasonable than anything i've read so far.

Talarian · Mar 28, 2008, 01:40 AM // 01:40

Quote:

Originally Posted by joshuarodger

this actually makes the most sense and sounds more reasonable than anything i've read so far.

*sigh* We agreed that these observations are true on the first page. What we are trying to ascertain is why.

manitoba1073 · Mar 28, 2008, 03:10 AM // 03:10

Quote:

Originally Posted by Talarian

No, you really aren't. In all the screenshots within the raptor cave, you killed the mobs with AoE. That does not prove anything.

The reason I don't like your image of the invisible henchmen is because it makes no sense, and yet is the only thing that is supporting your theory. I accept that it is possible in a full group (either it be other players or H/H) that all the drops be assigned to one person from a big AoE kill. I have already speculated about that above. What I do not accept is that there are any other parties involved for the system to bug to when solo farming. I fail to see why you are not getting this. YOU ARE THE ONLY PERSON IN THE INSTANCE, ALL DROPS ARE ALWAYS ASSIGNED TO YOU, THE LOOTSCALING CODE MEANS THAT WHEN YOU ARE SOLOING 1/8TH OF THE MOBS WILL ACTUALLY DROP LOOTSCALED ITEMS. There is nowhere for it to hang and select someone else - there is no-one else!!!!!

And FoxBat, if you had actually read through the thread, you would know that I have only been talking about incidents where there are 10+ mobs. You would also see that I have speculated on an explanation as to why SA yields fewer drops than killing foes 1 by 1 but more than all at the same time. You would in fact be aware that I stated that the system may assign death by AoE if the mob dies within, say, half a second of the last mob. Please don't just read the OP and look at the number of replies and assume we don't know what we are talking about.

No its you who is not getting. Look at what the drops were. most of the drops were LS items. Just cause you cant grasp the simple idea that it is the same as if you went with real people or heros/henchies ie the invisible henchmen. Still dont believe me take a group of hero/henchies out move them away till they are blacked out. You're drops will be the same.

FoxBat · Mar 28, 2008, 05:52 AM // 05:52

Quote:

Originally Posted by Talarian

death by AoE if the mob dies within, say, half a second of the last mob.

You will see less drops at 20s between deaths
You will see even less drops at 10s between deaths.
Even less at 5s between deaths.
This has nothing to do with AoE. The goal is just to cap the rate at which you can get drops, and they just look at kills per second to do that.
You might not even be disagreeing with me, but it's misleading to bring AoE into the discussion at all, because the game doesn't care what you kill them with.
It just so happens that killing everyone at once, is a pretty good way to max out your scaling. But it's not the only way to get it up there.

Quote:

What we are trying to ascertain is why.

I put not only an observation but an explanation. (though could be debated) Capping the amount of loot you can get via deaths per time is a very elegant and simple solution for the programmers to have chosen and directly coded. Then it doesn't care what farming build you develop or how efficiently you kill them with or without AoE, because once you cap out your kills/second you get hit with full loot scaling. If I had more patience I could probably figure out the exact time window but a few others have already done that.

If mantinoba is not having everything die at exactly the same time with those weaker earth/water builds, then those SSs fit perfectly with my explanation. I don't think it's a bug at all, this is how the system is supposed to work. 24 guys at once is a pretty big kill/second, 12 followed by 12 a few seconds later is much less.

manitoba1073 · Mar 28, 2008, 08:42 AM // 08:42

Quote:

Originally Posted by Antithesis

Sorry manitoba, your tests and results prove nothing because you're comparing AoE with AoE, not AoE with Single. Your results look no different to any of my average Raptor runs.

Try that same run again the same number of times with Sliver Armor then post those results. Others have posted theirs and without fail, Sliver nets more drops than AoE.

I think you're on your own on this one. Maybe you're neither intuitive nor a loon...perhaps sh*t-stirrer is more appropriate

I did a good number of runs with sliver before and some tonight still highest is 4 drops with it for 20-24 foes.

Talarian · Mar 28, 2008, 03:08 PM // 15:08

Quote:

Originally Posted by manitoba1073

No its you who is not getting. Look at what the drops were. most of the drops were LS items. Just cause you cant grasp the simple idea that it is the same as if you went with real people or heros/henchies ie the invisible henchmen. Still dont believe me take a group of hero/henchies out move them away till they are blacked out. You're drops will be the same.

You still don't get it. I never said you wouldn't get any lootscaled items, but that you would receive fewer. Which you have. And I agree that you will pretty much receive the same drops as you would with a whole party - that's the point of lootscaling.

Foxbat, while I agree with you in theory, I think you are oversimplifying things. It would be nonsensical for it to be the way you say it, else there would be no point in the lootscaling code. People would simply go into UW, kill one smite, lose aggro, kill one smite, lose aggro, etc. and by your reckoning would receive much more loot than a terra tanking ele. I strongly believe there are at most 3 different "time of death" measurements, one being deaths per 0.5 secs, the other perhaps 0.5 to 2 secs, with the third being over 2 secs. In this situation, I merely use the term AoE to describe a situation where there is less than 0.5 secs between deaths. As stated above, the question over whether it actually differentiates between AoE and other forms of damage is another thing entirely.

FoxBat · Mar 28, 2008, 06:13 PM // 18:13

Quote:

Originally Posted by Talarian

Foxbat, while I agree with you in theory, I think you are oversimplifying things. It would be nonsensical for it to be the way you say it, else there would be no point in the lootscaling code. People would simply go into UW, kill one smite, lose aggro, kill one smite, lose aggro, etc. and by your reckoning would receive much more loot than a terra tanking ele.

Yes, but it would take them ages. Once farming becomes that slow, it's not that much better than going out and killing things with a group. That was the point of the change.

Although it is hard to test and therefore I haven't personally, I have tested that loot scaling only applies when you have a good kill/time ratio. So I therefore infer that this *is* loot scaling. So when anet claims that rare materials like ecto are exempt from the scaling, then they won't be affected by kills/time. Thus sandstorm is still the fastest route to ecto, you are just missing out on alot of white crap drops. Some have sortof confirmed this by still turning up decent purple/gold rates from rediculous mobs like the gate of kryta HM farm.

On the wiki talk pages, Gaile mentioned there is "code for filtering elite farmers" I'm certain this is it. This is so when I go out with my lame burning arrow ranger and solo dune burrowers for their jaws, I'm not significantly hit by LS. Sure enough I get a jaw most of the time. I have a warrior that has been soloing through prophecies and am drowning in white collectables for similar reasons.

Your "hard time window" theory may merit investigation but it doesn't fit with what I've seen. A poor mesmer at 12 earth takes 4-5 seconds to kill a minotaur and still recieves far less drops than taking 40 seconds per minotaur. A smooth scaling system is both easier to program and makes more sense for accomplishing their goals. Remember they've been fiddling with AoE and monster changes for a while now to make farming harder, this system is fairly abuse-proof by cutting it off at the root, at the drops received.

Ultima pyromancer · Mar 28, 2008, 06:52 PM // 18:52

By this are you saying you cannot get 2 drops one after another?

I find this hard to believe

Maybe a % chance of drop is done by time of death but not a complete block of something droppping immediately afterwards

Talarian · Mar 28, 2008, 07:24 PM // 19:24

So the two schools of thought here are that lootscaling either kicks in as soon as you enter an area (so there is a ceiling on the amount of drops you can get regardless of how slowly you kill mobs) or that it depends on the speed at which you kill mobs (or more precisely, the time difference between each mob death). I don't know how we can test this beyond what Skinny Corpse has been doing, unless we were to do the same simultaneous entry trick, with one person rampaging through the area, and the other killing only 1 mob per minute (would have to be done in NM so they could lose aggro). That would at least go some way to proving or disproving FoxBat's suggestion, and if Skinny Corpse's tests of nuker vs SV necro come back identical then we're all mad and imagining things.

FoxBat · Mar 28, 2008, 07:37 PM // 19:37

Quote:

Originally Posted by Talarian

So the two schools of thought here are that lootscaling either kicks in as soon as you enter an area (so there is a ceiling on the amount of drops you can get regardless of how slowly you kill mobs)

The fact that you can observe what I said with the very same character doing the same area with different builds makes that hypothesis pretty unlikely. Not to mention that drops are assigned to players/henchies dynamically based on presence (move player/henchie off the radar and they aren't eligible for drop assignments. Yet loot scaling with times/death is in force when you kill with henchie off the radar.) But the trick along the lines of Skinny's tests are more robust so I'd love to see those results.

There have been rumors of remenants of the "repeated entry farm code" still lingering around, which would relate to school #1. I have vague suspicions but haven't done enough to really establish evidence. At only a few entires you definitely see behavior along the lines of school #2 though.

Ultima pyromancer · Mar 28, 2008, 07:41 PM // 19:41

How i see it is, if this whole timer theory is true, then how about 8 man party's? or 6 man?

They would have to set a time for how many are in the group or how many "Invisible henchmen" there are which seems a bit too complex.

One thing i have noticed is, Whilst doing Raptor runs - thats killing about 25 i think, the average drop rate is 3 out of all of them, hard mode and normal mode seem to be the same apart from the quality of drop.

So maybe... This may be a bit over the top, The Loot scaling only allows ONE drop per group you kill, and then theres % chance drop and hard mode/normal mode to take into consideration.

My Lipgloss is Cool · Mar 28, 2008, 08:21 PM // 20:21

Ok, here goes this;

Three players, all doing the same run. We left Farahnur, killed the groups to the big skale group, then killed them. We all killed the same groups, and the only difference was the builds we used.

Here are our drops for the runs;

Me, with SS (The final skale group was all killed at once)

A Guildy, with SV

And Noodle, with a degen build.

If the kill/time thing effects drops, then Noodle's build should have many more drops than the SV build, as degen is a very slow way to kill...Anything...

Well, as expected both builds got more drops than the AoE build. Shocking.

Now it becomes interesting. SV got more drops then the degen build. The degen build got more gold then the SV build.

So...Yeah...I'm going to stick with drops are determined on if you killed several things at the same time/not at the same time, not the time between kills.

MLiC

Antithesis · Mar 28, 2008, 11:47 PM // 23:47

Try that again with time and mob synchonicity...that's the point of Skinny's thread. IF both time and mobs are sycned, 99% of drops are the same regardless of kill method.

There's been a new post in the thread comparing Sliver Armor to an Earth AoE build - whilst there was scatter that prevented a wider AoE effect, multiple kills were made instantaneously and, you guessed it, drops were the same. This needs further testing as no more than 3 were killed at the same time.

Based on Skinny's tests, it's possible the server generates all loot for that zone and assigns those drops to players x, y, z etc on instance creation based on a time-seed on the server. Creation of a zone is the only time we get a significant data load from the server - we've already got the bulk of a zone in the gwdata file on the client, the only things that need to load from the server (roughly) are the enemies, chests, quests, party members and loot for that instance.

Generating loot on instance creation would alleviate server load as there is less client > server interaction required than generating drops on death, which could help overcome variable client connection speeds and lag. This could be complete bollocks as some zones take hours to complete and The First City may be an anomoly, so maybe the results can't be applied to the rest of the game. But it would save bandwidth and CPU power for Anet and that's a savvy business decision.

Kill monster x and player y gets non-lootscaled drop IF that player is in radar range of the party leader. There's also a % chance that player y will snag a pre-assigned chalk drop for monster x. If the kill is made by AoE, there's less of a chance you'll get the chalk drop. It's a bit iffy but how else do we explain the consistent results of Skinny's thread? There seems to be very little randomness about who get's what when farming solo with the exception of some whites. It's a 4-man zone so maybe it's just easier to see here than in an 8-man zone. Testing with additional party members is kinda irrelevant (because we're farming) unless you want to find out to what degree does randomness come into distributing those drops to party members within radar.

This is doing my head in...I liked it better when I thought drops were random.

My Lipgloss is Cool · Mar 28, 2008, 11:55 PM // 23:55

Something I've been thinking about for a while.

What if every mob is given an item on zone creation, and when it's killed, the game checks against how many others were killed at the same time, and adjusts drop chance % accordingly?

This supports Skinny's findings, where all drops were pregenerated and explains how AoE killing lowers droprate.

MLiC

Antithesis · Mar 29, 2008, 12:14 AM // 00:14

AoE must play some role - it's easy to see that it does when doing Sliver runs.
It makes absolute sense that it does - lootscaling was implemented to combat AoE bot farmers.
To what degree it plays a role quite likely does come down to how many were killed over a unit of time.

As FoxBat suggests, maybe kills/time = lootscaling. As MLiC suggests, drops that otherwise would have been assigned to you are taken away....we're not rewarded for speedy kills, we're punished. Kinda flies in the face of the whole premise of the game in that greater skill = greater reward.

This supports almost every theory in this thread including manitoba's (i think, i can never tell what he's on about). But what drops does AoE effect?
At the very least we've created additional tests for Skinny's playground, results should be interesting.