Drop Rates SoJ vs. SV (Image heavy!)
hooded player
i dont give a **** what you guys say to give me "proof" that this isnt real but it think it is.
Chico
Quote:
Originally Posted by Nechtan Thaumaturge
Well, I tried it out the other day, in a way. I used two accounts and set up both with Ogden Stonehealer and 600/smite-ish builds.
I ran two trials and found essentially the same thing in both cases. Discounting coin drops, slightly less than half of the total drops were shared. This is what you would expect to happen if two identical sets of drops were each randomly split two ways. Therefore, I'm going to go out on a limb and say that if you could look in Ogden's inventory, you'd find the majority of the missing drops. Needless to say, loot distribution is not fixed in the same way that drops are. The governing algorithm re-seeds the random number generator or otherwise injects another source of "randomness" into the process.
Also, I wanted to mention that on the second run the chest spawns did not correspond, although the creature spawns were identical. That's never happened to me before on a successful sync, but I remember someone asked about the possibility of this happening. That's interesting. Based on just that one test I think it is quite safe to assume that -what- is going to be dropped is decided at instancing but -who- is going to get the drop is decided on the fly (thus the random distribution between your runs) or at least it doesn't depend on the instancing seed/time.
I can't read the whole thread again, but has this been used to test the time of death bug? This could throw very interesting results and maybe even confirm the existence of the so called 'Time of Death Bug'.
I know the title of the thread is SoV vs SV but I have no idea if those trigger the ToD bug or if you would need something like a nuker or SS to test this bug.
Thanks for your contribution Nechtan.
I ran two trials and found essentially the same thing in both cases. Discounting coin drops, slightly less than half of the total drops were shared. This is what you would expect to happen if two identical sets of drops were each randomly split two ways. Therefore, I'm going to go out on a limb and say that if you could look in Ogden's inventory, you'd find the majority of the missing drops. Needless to say, loot distribution is not fixed in the same way that drops are. The governing algorithm re-seeds the random number generator or otherwise injects another source of "randomness" into the process.
Also, I wanted to mention that on the second run the chest spawns did not correspond, although the creature spawns were identical. That's never happened to me before on a successful sync, but I remember someone asked about the possibility of this happening. That's interesting. Based on just that one test I think it is quite safe to assume that -what- is going to be dropped is decided at instancing but -who- is going to get the drop is decided on the fly (thus the random distribution between your runs) or at least it doesn't depend on the instancing seed/time.
I can't read the whole thread again, but has this been used to test the time of death bug? This could throw very interesting results and maybe even confirm the existence of the so called 'Time of Death Bug'.
I know the title of the thread is SoV vs SV but I have no idea if those trigger the ToD bug or if you would need something like a nuker or SS to test this bug.
Thanks for your contribution Nechtan.
psykick5
thats amazing... my cousin and i should try the same thing and see if that happens to us too. would be interesting if that is true...
Truegen
Great research...this issue is more or less like a philosophical question in the sense that there may never be one distinct answer accepted by all, but now there's at least a framework here that a majority can agree on
Bob Grunter
lol this is cool but i believe it to be fake
Skinny Corpse
Still some unbelievers, huh?
Ah well, I've done what I can to show it's not fake.. and as
I'm not playing GW anymore I won't update this thread with
further results.
Ah well, I've done what I can to show it's not fake.. and as
I'm not playing GW anymore I won't update this thread with
further results.
T1Cybernetic
Quote:
Originally Posted by Bob Grunter
lol this is cool but i believe it to be fake
What leads people to believe this stuff? all the evidence is there for everyone to see, just go and test it yourself before crying fake!
There is no substitute for experience!
There is no substitute for experience!
byakko
how did your brother get a necromancer tome to drop in normal mode? (first pics)
Skinny Corpse
Quote:
Originally Posted by byakko
how did your brother get a necromancer tome to drop in normal mode? (first pics)
Cuz it ain't Normal mode.
He's just customized the UI a bit, moving the HM icon to the bottom of the screen.
He's just customized the UI a bit, moving the HM icon to the bottom of the screen.
FrAnt1c??
Now we only have to find out at what time an everlasting tonic drops from the zaishen chest 
jetlaw
Yes, I know I am late to the party, but having just spotted this thread for the first time, I gotta say that the results are rather amazing. Very nicely done! 
I can only imagine a massive exploit of this by having large groups sync and then have one person tell the rest it it was worth their time/lockpick. However, that would only occur IF syncing were easy to accomplish. But, as it stands, since syncing is such a pain, it seems that the possibility of doing runs with more and more people would get exponentially more difficult (since all would have to sync perfectly, and the likelihood of each additional person syncing would make the odds of a successful sync of a large group VERY unlikely).
I also appreciate knowing that builds using one at a time kill skills (such as SV/IP) really don't impact the drops versus a AoE insta-kill build (such as the pre-SF nerf Raptor Farming AoE Ele/A builds). Since the real difference is speed, it makes me wish that you could use AoE without scatter for faster kill rates.
As for the question posted on the Z key sync, I think the main problem would be the inability to know if you were synced. Unlike other explorable areas, when doing Z key runs there are no variable enemy spawns to give you the ability to tell if you are precisely synced. As a result, it would seem to me that people who tried to sync for Z key runs really would be rolling the dice every time they open a chest.
I can only imagine a massive exploit of this by having large groups sync and then have one person tell the rest it it was worth their time/lockpick. However, that would only occur IF syncing were easy to accomplish. But, as it stands, since syncing is such a pain, it seems that the possibility of doing runs with more and more people would get exponentially more difficult (since all would have to sync perfectly, and the likelihood of each additional person syncing would make the odds of a successful sync of a large group VERY unlikely).
I also appreciate knowing that builds using one at a time kill skills (such as SV/IP) really don't impact the drops versus a AoE insta-kill build (such as the pre-SF nerf Raptor Farming AoE Ele/A builds). Since the real difference is speed, it makes me wish that you could use AoE without scatter for faster kill rates.
As for the question posted on the Z key sync, I think the main problem would be the inability to know if you were synced. Unlike other explorable areas, when doing Z key runs there are no variable enemy spawns to give you the ability to tell if you are precisely synced. As a result, it would seem to me that people who tried to sync for Z key runs really would be rolling the dice every time they open a chest.
squiros
this actually stands up to real programming. for those of you who are good at computers:
the default measurement of time, for most API's is seconds. the current time is actually the total amount of seconds that have passed since midnight of january 1st, 1970. the is pre-made, and included in time.h standard library. to get the time, you use a predefined class. a class is is a collection of data. when your character is created is based on that as well.
int player_loot_seed = srand( time() ) ;
the resolution is SECONDS. time returns a class time_t. srand takes a time_t class. so if you zone in at the same second - which is well within most server lags, it's synched. it has to be the same second, obviously, not just within 1 second.
srand is also a pre-made function. Knuth was the one who essentially designed random number theory. knuth designed a system that willl appear random to all known statistical analysis. the algorithm IS known, the values created are not. for example, take the current time as the number of seconds passed, and simply take that digit of pi. if 1 million seconds have passed, take that digit of pi. if you can predict that, your name and idea will be added to every book about computers for the rest of time =). but anyway, linear congruency applied by knuth's modulus method is more than sufficient for a random number sequence of 2^31. that means there will be more random numbers than everyone's farming runs combined. if anet didn't want people predicting it, a cipher could be applied. of course, there are hash collision attacks, etc.
http://en.wikipedia.org/wiki/Linear_congruence_theorem
http://members.cox.net/srice1/random/knuth_pg24-25.gif
http://members.cox.net/srice1/random/knuth_pg26-27.gif
http://members.cox.net/srice1/random/knuth_pg28-29.gif
http://members.cox.net/srice1/random/knuth_pg30-31.gif
http://members.cox.net/srice1/random/knuth_pg32-33.gif
the default measurement of time, for most API's is seconds. the current time is actually the total amount of seconds that have passed since midnight of january 1st, 1970. the is pre-made, and included in time.h standard library. to get the time, you use a predefined class. a class is is a collection of data. when your character is created is based on that as well.
int player_loot_seed = srand( time() ) ;
the resolution is SECONDS. time returns a class time_t. srand takes a time_t class. so if you zone in at the same second - which is well within most server lags, it's synched. it has to be the same second, obviously, not just within 1 second.
srand is also a pre-made function. Knuth was the one who essentially designed random number theory. knuth designed a system that willl appear random to all known statistical analysis. the algorithm IS known, the values created are not. for example, take the current time as the number of seconds passed, and simply take that digit of pi. if 1 million seconds have passed, take that digit of pi. if you can predict that, your name and idea will be added to every book about computers for the rest of time =). but anyway, linear congruency applied by knuth's modulus method is more than sufficient for a random number sequence of 2^31. that means there will be more random numbers than everyone's farming runs combined. if anet didn't want people predicting it, a cipher could be applied. of course, there are hash collision attacks, etc.
http://en.wikipedia.org/wiki/Linear_congruence_theorem
http://members.cox.net/srice1/random/knuth_pg24-25.gif
http://members.cox.net/srice1/random/knuth_pg26-27.gif
http://members.cox.net/srice1/random/knuth_pg28-29.gif
http://members.cox.net/srice1/random/knuth_pg30-31.gif
http://members.cox.net/srice1/random/knuth_pg32-33.gif
Daft Shifty
^^ what he said but less programmer speak... knew i should have kept learning perl...
Some_guy
I wasn't really paying attention, but.. Where was that farmed at?
Dreamwar
So, we went to Fahranur to do some tests in the name of science.. me playing my monk with Shield of Judgement, and my brother
It's in the first sentence....
It's in the first sentence....
Marty Silverblade
I'm still wondering why you chose SoJ. It's not exactly what you want if you want to kill everything at once. Wouldn't it be possible if you did it in NM, just because eles can wipe those massive groups with big fire nukes? I know loot scaling works a bit differently in HM but would it matter?
Nechtan Thaumaturge
It's a bit tricky to get simultaneous kills in Fahranur for several reasons. The groups aren't homogeneous, which means some monsters have fewer hit points than others and are likely to die quicker. Also, some will have less armor and therefore take more damage. Some of the monsters have healing/life-stealing skills which allows them outlive the ones that don't have them (or don't use them).
The best I could do was to use the Visage spells to drain away energy and therefore healing ability, then whittle the monsters down with armor-ignoring damage, and finally spike them to death with Crystal Wave. I think I was using Ether Prism for energy and Kinetic Armor + Stoneflesh Aura to stay alive. That seemed to get the simultaneous kills, but it takes awhile to work. Also, I didn't try NM, because I wanted to match up as closely as I could against the original results.
The best I could do was to use the Visage spells to drain away energy and therefore healing ability, then whittle the monsters down with armor-ignoring damage, and finally spike them to death with Crystal Wave. I think I was using Ether Prism for energy and Kinetic Armor + Stoneflesh Aura to stay alive. That seemed to get the simultaneous kills, but it takes awhile to work. Also, I didn't try NM, because I wanted to match up as closely as I could against the original results.
Fay Vert
Quote:
Originally Posted by Marty Silverblade
I'm still wondering why you chose SoJ. It's not exactly what you want if you want to kill everything at once. Wouldn't it be possible if you did it in NM, just because eles can wipe those massive groups with big fire nukes? I know loot scaling works a bit differently in HM but would it matter?
Indeed, that is the way it always used to be proved. I used to farm Zelnehlun with this build:
[Zelnehlun Farm;OgQVUsDbz7gPJPE6zQN8NxA6OOA]
Basically disease them all to drop health to 90% and SS Zelnehlun so when SS takes one out, they all drop to the EoE bomb. Prior to the loot nerf you got 11-13 drops per run. When the loot nerf hit you get 3-4.
Okay, thats just he loot nerf doing what its supposed to do you would think.
But no, witht he same build, but this time you just stand on the bridge and PoF +RH them down one by one (about 1 kill per 30 seconds) and bingo, they are now dropping 11-13 again.
Loot scaling? No
Rate of kill? Yes
Here is a couple of screenies I just made
Of course, you could do the simultanious entry trick if you wanted further proof, but anyone can do this if they don't believe it.
Fast kills
Slow Kills
[Zelnehlun Farm;OgQVUsDbz7gPJPE6zQN8NxA6OOA]
Basically disease them all to drop health to 90% and SS Zelnehlun so when SS takes one out, they all drop to the EoE bomb. Prior to the loot nerf you got 11-13 drops per run. When the loot nerf hit you get 3-4.
Okay, thats just he loot nerf doing what its supposed to do you would think.
But no, witht he same build, but this time you just stand on the bridge and PoF +RH them down one by one (about 1 kill per 30 seconds) and bingo, they are now dropping 11-13 again.
Loot scaling? No
Rate of kill? Yes
Here is a couple of screenies I just made
Of course, you could do the simultanious entry trick if you wanted further proof, but anyone can do this if they don't believe it.
Fast kills
Slow Kills
Fay Vert
Okay, I tried the above but this time syncing zone entry out of Rilohn Refuge with a guild mate, who was running the one kill per 30 second, while I EoE bombed the lot within 1 second. Syncing was easy as when you zone there are variable spawns to your left and right to check a match with.
Drop numbers were as before, I got 2-4 each time, he got 10-13 each time. But the intersting thing is that not once were my drops a subset of his, not a single matching drop, except once we both had a spider web (they are common drops). No item or cash drops were the same, and no loot scaling excempt items were the same either.
We were not able to reproduce the OP findings, and we were at it for ages.
Drop numbers were as before, I got 2-4 each time, he got 10-13 each time. But the intersting thing is that not once were my drops a subset of his, not a single matching drop, except once we both had a spider web (they are common drops). No item or cash drops were the same, and no loot scaling excempt items were the same either.
We were not able to reproduce the OP findings, and we were at it for ages.
Marty Silverblade
If you're only farming Zehnehlun, you won't have enough drops to prove much. If you were to clear the entire zone, maybe you'd see more similarities.
Fay Vert
Your missing the point. This is about sync entry drops, not about that old rubbish Gaile used to spout about the whole zones drops.
Every one of my 2-4 drops each run should have been present in my friends 10-13 to be consistant with the OP. We both did the same run at the same time synced up. But not one was the same over a number of sync runs. I could not reproduce the OP results.
Every one of my 2-4 drops each run should have been present in my friends 10-13 to be consistant with the OP. We both did the same run at the same time synced up. But not one was the same over a number of sync runs. I could not reproduce the OP results.
Age
I don't know how you keep your inventories so clean as I have stuff in mine and few slot to put drops in.It looks all even to me I will have to give that place a try.
go cubs
well...just saw this now...thats pretty incredible....ive never seen this farm sadly...wanna try it out looks good....just wondering how long the run is....sounds sweet i wanna try this out with my bro see if it works
looks good....just wondering how long the run is....sounds sweet i wanna try this out with my bro see if it works
killress
This is very interesting. Good research.