Curse

Luminarus · Mar 26, 2008, 09:36 AM // 09:36

HAHA Suck Din :P

JKS

Dunno why.

tmakinen · Mar 26, 2008, 09:40 AM // 09:40

Hey strcpy, I remember discussing a related issue with you before, namely whether drops are determined at the creation of an instance or at the time of the kill. The evidence by that time was somewhat inconclusive, but just recently I saw this thread which demonstrates pretty well that drops are indeed determined when the instance is created (the only other option being that there would be separate seeds for drops and other random events which doesn't sound likely, and even in that case the drops would be determined if not actually calculated at the creation).

How this relates to the OP, if you suspect that the game is withholding your well deserved loot because ANet has some personal grudge against you, there is an experiment that you can do to (dis)prove your case. Read the linked article about sync farming and do the same with some lucky friend of yours. Record everything that drops and compare results. If the data show that you're in sync but still missing something on the exempt list that your friend got then you have a case. Happy experimenting

Joe L. · Mar 26, 2008, 03:09 PM // 15:09

Quote:

Originally Posted by tmakinen

Hey strcpy, I remember discussing a related issue with you before, namely whether drops are determined at the creation of an instance or at the time of the kill. The evidence by that time was somewhat inconclusive, but just recently I saw this thread which demonstrates pretty well that drops are indeed determined when the instance is created (the only other option being that there would be separate seeds for drops and other random events which doesn't sound likely, and even in that case the drops would be determined if not actually calculated at the creation).

How this relates to the OP, if you suspect that the game is withholding your well deserved loot because ANet has some personal grudge against you, there is an experiment that you can do to (dis)prove your case. Read the linked article about sync farming and do the same with some lucky friend of yours. Record everything that drops and compare results. If the data show that you're in sync but still missing something on the exempt list that your friend got then you have a case. Happy experimenting

Actually, I read this same thread, and it piqued my curiosity, so I gave it a go. I was running SoJ smiter, and my partner was running SS. It only took one run to make me a believer. All drops, and I mean ALL were identical, and to be honest it was one of the better runs I've had in a long time. We did Fahranur, and I remember getting 2 dead swords,and at the very least 4 other gold drops. Alas I do not have a screenshot, since I've wiped my HD since then, and figured I would not need GW screens.

On another note, I did not create this thread to cause any kind of argument. I know there is very complex math involved to actually create what drops and when, that is not my issue. My issue is if this actually changes based on how often you happen to farm. The whole "anti-farm" thing is allegedly based on punishing solo-farming, yet I get horrid drops when I go with a full party;At this rate, I may as well solo-grind forever, at least I'd get 1 stupid gold eventually. I believe my impatience is justified, since: 2 Slaver's full runs = 5-6 hrs, and yields nothing but onyx and diamond, with purples from chests. Thanks A-net for making team playing rewarding.

strcpy · Mar 27, 2008, 05:31 AM // 05:31

Quote:

Originally Posted by tmakinen

Hey strcpy, I remember discussing a related issue with you before, namely whether drops are determined at the creation of an instance or at the time of the kill. The evidence by that time was somewhat inconclusive, but just recently I saw this thread which demonstrates pretty well that drops are indeed determined when the instance is created (the only other option being that there would be separate seeds for drops and other random events which doesn't sound likely, and even in that case the drops would be determined if not actually calculated at the creation).

Yes, read the thread a long time ago. Drops are interesting.

I don't really know what to think of that at the moment. There is a great deal of evidence that drops are affected by how fast you kill things and it works nearly 100% of the time. However we also have that thread and I don't really see them as lying (though I do not think that they could actually sync their entering a mission to milisecond - some other trigger would be required - I was rather surprised that they had a coarse enough seed that this was possible as that is generally a big no-no).

As such if I happen to jump into that argument now I tend to tell both sides as I see it needs more data/time. Right now my opinion is that there is something to both sides and a third option is the correct one - as to what that is dunno. While possible (just as it is possible I may get hit by a meteor tomorrow) I have a hard time seeing that kill speed has *nothing* to do with it due to the amount of data that shows otherwise. However, it is obvious that a great deal of drops are determined at the creation of an instance.

You will also note that I never said this person drops were *not* degrading, just that his evidence shows no such thing and is perfectly within chance (in fact it is even expected to occur). Anet - the people who actually know all the answers to these questions - says random and they do not degrade drops for any reason. The whole thing about loot being equal between the two characters VERY much shows that what Anet says is correct.

Bluefeather · Mar 27, 2008, 07:08 AM // 07:08

Quote:

Originally Posted by strcpy

Considering that I had quite a number of graduate courses in mathematics (I'm a computer scientist), yes that is pretty simple stuff. You know, someone somewhere had to understand that to write it and it isn't even written to a standard a higher level undergrad college course would accept, let alone someone who actually works with that type of math at a professional level.

Just because you can't follow some math doesn't mean everyone else on the planet also can't do it. The point being that until you even have that basic a level of understanding of what is going on you do not have the background to make an educated assertion about the issue and are simply guessing. Same is true of I look at an airframe and tell a aeronautical engineer his plane will not fly - what the heck do I know about designing and airframe?

No, but then since you obviously don't understand what is going on I can see how you wold think so. Better if you point out that PRNG's are not truly random but are deterministic so that variance doesn't count. While you still obviously don't understand the math behind them and are nearly 100% wrong you at least have a very basic understanding of some of the principles involved and you aren't fully 100% incorrect (to note decent PRNG's pass all of these statistical tests and even poor ones do not fail in the way you, and others, are saying they do).

Of course, I don't particularly expect lack of knowledge to stop people from telling mathematicians how mathematics work - it never has in the past. I'm sure your lack of understanding gives special insight on the matter at hand and you could easily tell that 3 dimensional RANDU sequences fall into 15 2 dimensional planes by noting that your gold drops aren't as good as you expect.

Someone told me that if a person uses so many jargons, do not listen because his real intention is not to be understood. And I ask why, he said, because he doesnt understand what he is saying.

The first time I worked on Hard Mode I got a gold drop from the first gargoyle in Ascalon. In my excitement, I went back in to tell people that I got a gold drop from my first kill in HM. Now, i frequent Sardelac Sanitarium to kill that stone-something-boss with "coobble.." or something in his name. It drops gold weapons, gold and sometimes nothing. And I understand why, because it's random.

Off-topic, u said ur a computer scientist. You must have a lot of drops in game. Hahaha i'm still laughing. I think I'm going to visit this thread more often to read funny jokes hahahaha.

Bluefeather · Mar 27, 2008, 07:13 AM // 07:13

Quote:

Originally Posted by MirkoTeran

Actually, I do. I'm a CS graduate so we did have a bit of math.

No, but if you were better at it you would know what you are getting bad drops.

I could write a bit more but strcpy explained it quite well.

And this one, too.

CS Graduate also. And thinks that Math knowledge has something to do with good drops.

I'm going home. Too many nerds today.

Inger · Mar 27, 2008, 07:28 AM // 07:28

Quote:

Originally Posted by Bluefeather

And this one, too.

CS Graduate also. And thinks that Math knowledge has something to do with good drops.

I'm going home. Too many nerds today.

They're not saying you have to even remotely understand that page in order to get good drops. Quite clearly its not your lack of understanding of the mathematics that is affecting you but your lack of understanding of the english language.

Basically the underlying message behind posting that page was just to indicate that the drop rate is RANDOM variance is just the mathematical representation of randomness. If you understand the mathematic so much the better, you understand things will be random. If you don't then it doesn't affect your luck at all. The math page was just to indicate things will be random, you don't have to understand math in order to get good or bad drops, its simply an "explanation" or reasoning behind why you can get a string of bad drops or on the other hand a string of good drops.

Davros Uitar · Mar 27, 2008, 08:19 AM // 08:19

Funny though - you never see the thread that says wow I just got 8 gold drops that run, or the one that says wow my last 12 chests were gold and the picks never broke.

Believe me though - it does happen in the game.

tmakinen · Mar 27, 2008, 08:26 AM // 08:26

Got five rares out of HM chests with the same pick yesterday, I consider that pretty good (about 0.3% chance). On the NM side my luckiest pick has so far been 14 chests opened (0.25% chance), not all rares, though.

Also, I'd be interested to find out exactly how coarse the instance seed is. 10 ms, 100 ms, 1 s? Note to self, should do some experiments

Perkunas · Mar 27, 2008, 02:36 PM // 14:36

Everyone can argue back and forth about lootscaling, about the mathmatics behind it, I don't care. If lootscaling did not change the drops rates of gold items, then how come "pre-lootscaling" - 60 chests opened, 43 items ID'ed; "post lootscaling" - 128 chests opened, 57 items Id'ed? Please explain that with your math.

This is on the same character.

Esan · Mar 27, 2008, 02:40 PM // 14:40

Quote:

Originally Posted by Perkunas

Please explain that with your math.

Random numbers are random?

Joe L. · Mar 27, 2008, 02:50 PM // 14:50

Quote:

Originally Posted by Davros Uitar

Funny though - you never see the thread that says wow I just got 8 gold drops that run, or the one that says wow my last 12 chests were gold and the picks never broke.

Believe me though - it does happen in the game.

Not for my monk it doesn't. And now how are the rare drops going for a different character, from the same Slaver's boss chests? Also forgot to take a screen, but had 2 gold staves drop from Duncans on HM. So out of the last 15 slavers chest drops with my monk: 0 golds, last 15 across tank, and ranger/ursan 13 gold drops. Monk us BUSTED! lol

I had always figured it was just a bit of bad juju, seeing how my monk used to be a farmathon, but as months go by, and the horrid luck continues in dual-farm, and fulll party scenarios, you can see how the scepticism would sink in. I guess the biggest problem is, that since the drops are so bad, I'm not motivated at all to keep farming other spots with my monk to try out different theories, when it would just waste time that I could spend using any of my other characters and get many more rares.

P.S. on the one screen, my first elemental sword ever.

kranius · Mar 27, 2008, 03:08 PM // 15:08

What i think is funny, it's all the posts talking about the randomness of drops. Because GW is a program, thus coded by some people. But, randomness can't be coded, otherwise it wouldn't be randomness... If it was coded, with some algorithm, then the coder should get Fiels medal, because none of the greatest mathematicians could never "mathematicize" it.

So i think most of the people who talked about drops randomness with some mathematical backup are just wrong.

Sorry if i am not clear or made mistakes, English isn't my first language.

ischuros · Mar 27, 2008, 05:18 PM // 17:18

Ok, to settle this question once and for all, an experiment. First, we run a farm, a one man, raptors for instance, with four or so different people using the same build, same place time etc., trying to take out all variables. Then, two of them continue farming the area for an hour, or so, after that all for do another run, comparing their drops to the first run.
We do this over a couple of days, and see how it turns out. Since it dosn't take into account previous farming time, only relative to the first 'instance' it should be free of any bias due to the 'anti-farming code' if it exisits.

Ekelon · Mar 27, 2008, 05:22 PM // 17:22

To be honest, I wouldn't be surprised if they still have a little bit of the farm code in there... It is funny how easily convinced we are that they took farm code off by merely announcing it!

creelie · Mar 27, 2008, 07:36 PM // 19:36

You've just had a good run of bad luck buddy. Don't catch any superstitions while you're swimming in the ocean of tears here.

HawkofStorms · Mar 27, 2008, 07:44 PM // 19:44

Quote:

Originally Posted by Joe L.

To have true variance, then the pattern I am seeing should be ever-present, not progressively worse. To be progressively worse, means some of the base criteria for what does or not drop must be changing, therefore, not random. I guess that's the beauty of "worsening drops" (if) Anet actually did administer some code, is it's impossible to prove.

To address an issue I overlooked and has been brought up, Lately, in the past 5-7 months, I have actually not done much solo-farming at all. Not even an hour a week, so if anything, I would assume my drops would have gotten, if nothing else, a bit better, not the opposite.

Gambler's Fallacy.
You think you are having progressively worse bad luck, but have no actual data to back up your claims.

Dallcingi · Mar 27, 2008, 10:03 PM // 22:03

Bad luck.. i got all purple drops in gwen during the preview =p

manitoba1073 · Mar 28, 2008, 12:16 AM // 00:16

Yeap there definitely is still AFC in place not.

look in the chat log for how long I was playing. those drops were right afterwards. And yes it was straight farming with very few breaks.

strcpy · Mar 28, 2008, 03:35 AM // 03:35

Quote:

Originally Posted by Bluefeather

Someone told me that if a person uses so many jargons, do not listen because his real intention is not to be understood. And I ask why, he said, because he doesnt understand what he is saying.

Uh huh - doctors are quacks too since they use even more jargon than pretty much any profession on the planet. Given that there is no common slang for those ideas expressed the *only* way to express it is through the appropriate jargon.

I'm not the one who couldn't understand the link so this somehow makes me stupid?

Quote:

Originally Posted by kranius

What i think is funny, it's all the posts talking about the randomness of drops. Because GW is a program, thus coded by some people. But, randomness can't be coded, otherwise it wouldn't be randomness... If it was coded, with some algorithm, then the coder should get Fiels medal, because none of the greatest mathematicians could never "mathematicize" it.

Once more this has to be explained - a little knowledge is a dangerous thing. Yes, PRNG's are deterministic. However what generally makes them still a *random number* is that they express random when any and all statistical analysis is applied to them. That is, (assuming they didn't choose and algorithm close to RANDU - a very famous one that fails miserably), you can not tell the difference between them and a true random sequence. For all intents are purposes they *are* random and there are quite a few algorithm out there that do well in this regards. If a large super computer given months to calculate on it can not find a pattern in the sequence chances are you aren't going to come up with one - computers are quite good at finding those types of patterns and we are not.

Further even the bad ones do not exhibit the type of behavior that is complained about. Even Randu would just have drops that were never chosen and some that were chosen consistently more often than they should. Even then it took many thousands of samples viewed at a single time before the pattern was found.

There are fairly fast cryptographically secure pseudo random number generators out there, even some that are commonly used in games are fairly close (and if those guys aren't finding patterns then, again, for all intents and purposes they *are* random). In fact, there are VERY VERY few natural occurrences that are as statistically random as the decent PRNG's and they are all very expensive (flipping a coin is horrid, MUCH worse than even the majority of poor PRNG's out there).