(In)Equality in loot drops - Page 3 - Guild Wars Forums

Algren Cole · Jun 07, 2005, 06:05 PM // 18:05

Quote:

Originally Posted by Night

Actually you just answered your own question. See number 3 here and your above post describing how it changes when you switch to your other char.
I'll do my best to see if I can help you but in the end, sometimes perception is reality.

Anyway first things first. I highly doubt that the coders for ANET would hardcode something like "if Class=Ranger && SubClass = Mesmer then Rand_Number=(rand()%32) else Random_Number (rand()%6);" doesn't make sense although it's possible that through coding that an adverse affect on certain variable has thrown off the counts or drops rather. There is nothing that states that you are guaranteed 1/6th of all drops. For instance flip a coin. Let's say you get heads the first toss so 50/50 odds means that the next will be tails, right? Nope no guarantee as each toss is it's own individual event meaning that this toss has no corelation with the first toss. You could toss the same coin 10 times and get all heads or all tails. 50% or 1 in 2 simply means for that and only that toss, you have a 1 in 2 chance of getting heads or tails. What should see over a period of many tosses that the number fall in the 50% heads and 50% tails category but even that may be off until you hit large numbers like in the thousands.

It could be very well that between all of your chars you simply balancing out the approx. 1/6 (depending on where you fall as far party make up). It just appears that when you hit your Ranger, you get the short end of the stick. The only way you possibly test the true randomness would be to 1) run the same area with each character. 2) attempt solo then attempt with the same grouping of henchmem 3) Repeat it numerous times like several hundred per char.

Yes to truly be able to tell whether or not there is randomness, you have to do hundreds if not thousands of runs. Again keep in mind what random and what odds mean. Odds/Randomness doesn't mean that you are guaranteed anything, it simply means that on the drop that is occurring right then you have 1 in 6 shot (if in a party of that size) of getting loot. This could also be compounded if no drop is selected as that is an event as well. Run the above test and see.

You could be right that there is a bug that somehow keeps the particular class combo you have selected from getting the same share but I find it very unlikely. The reason why is that item generation and loot sharing is a function that has been around game programming since the inception of treasure/loot and drops, kind of like case statements or for loops and by now I am relatively certain that any issues that may have arised such as coding that causes adverse issues would have been discovered and corrected but who knows. Humans are the ones doing the coding so maybe someone did some sloppy coding.

I also doubt that Guild Wars was coded in Basis

lol

Dwiggit · Jun 07, 2005, 06:08 PM // 18:08

Night, Algren Cole, thanks for the helpful responses, and an open mind that something other than perception can be at play.

Sounds like both of you have some programming experience.

I was trying to avoid the scientific study of the situation, but I guess that is what it will take to convince people. I would rather just play the game and have fun, rather than logging all the loot drops for a month. I wonder if AN had some testing software that logs this stuff? It would make testing this a lot easier.

Thanks for helping!

Dwig

Night · Jun 07, 2005, 06:11 PM // 18:11

Quote:

Originally Posted by Algren Cole

I also doubt that Guild Wars was coded in Basis

lol

Basis? Not sure I follow that or did you mean Basic as in the language? Being that the rand() structure above is a C++ although the hastely thrown together If-Then statement was a lot easier to throw out then putting into a case statement as it would more than likely be coded in the game.

Dwig,

Unfortunately that would be the only known way to discover if what you see is happening is truly happening. As far as coding experiece, yep although mine is more on business aspects such as database mining/reporting and all that good stuff related to db programming.

Algren Cole · Jun 07, 2005, 06:12 PM // 18:12

Quote:

Originally Posted by Dwiggit

Night, Algren Cole, thanks for the helpful responses, and an open mind that something other than perception can be at play.

Sounds like both of you have some programming experience.

I was trying to avoid the scientific study of the situation, but I guess that is what it will take to convince people. I would rather just play the game and have fun, rather than logging all the loot drops for a month. I wonder if AN had some testing software that logs this stuff? It would make testing this a lot easier.

Thanks for helping!

Dwig

if I get some free time later this week I'll look into writing some rudimentary monitoring software to monitor all the drops that are assigned to each player. should make it easier to see what is actually going on....though I assure you it's a 1/# chance for every drop, and each drop has the same probability regardless of previous drops.(where #, again, is total party number)

Algren Cole · Jun 07, 2005, 06:14 PM // 18:14

Quote:

Originally Posted by Night

Basis? Not sure I follow that or did you mean Basic as in the language? Being that the rand() structure above is a C++ although the hastely thrown together If-Then statement was a lot easier to throw out then putting into a case statement as it would more than likely be coded in the game.

I did mean Basic. The If Then statement was Basic....and the rand() function is the same for .net and c++....and anyone that has coded in .net knows it's essentially a glorified basic language.

Santosh · Jun 07, 2005, 06:15 PM // 18:15

It is random. I've noticed whether I play with henchmen or with other players, the number of loot drops equally for each party member. Yes, randomly but equally. Not quality-wise, but number-wise. That's the difference. Some person in a group might get really nice drops during a mission, but it's all based on luck.

For example, if I was in a 6-man group, the drops would be in a rotation. Keep in mind that the mobs have already set loot drops in them for a particular mission when you zone in. Player 1 (ranger) gets the first 2 drops. Player 2 (war) gets the next 2 drops. Whatever each person gets depends on what that mob has on them. Coins or items. So on and so forth. It doesn't matter what class you play, it doesn't matter how much damage you do. So if random mob has a gold item, the lucky one will get that drop if it's their turn in the group rotation.

That's what I've noticed.

Arvydas · Jun 07, 2005, 06:16 PM // 18:16

Quote:

Originally Posted by Night

Actually you just answered your own question. See number 3 here and your above post describing how it changes when you switch to your other char.
I'll do my best to see if I can help you but in the end, sometimes perception is reality.

Anyway first things first. I highly doubt that the coders for ANET would hardcode something like "if Class=Ranger && SubClass = Mesmer then Rand_Number=(rand()%32) else Random_Number (rand()%6);" doesn't make sense although it's possible that through coding that an adverse affect on certain variable has thrown off the counts or drops rather. There is nothing that states that you are guaranteed 1/6th of all drops. For instance flip a coin. Let's say you get heads the first toss so 50/50 odds means that the next will be tails, right? Nope no guarantee as each toss is it's own individual event meaning that this toss has no corelation with the first toss. You could toss the same coin 10 times and get all heads or all tails. 50% or 1 in 2 simply means for that and only that toss, you have a 1 in 2 chance of getting heads or tails. What should see over a period of many tosses that the number fall in the 50% heads and 50% tails category but even that may be off until you hit large numbers like in the thousands.

It could be very well that between all of your chars you simply balancing out the approx. 1/6 (depending on where you fall as far party make up). It just appears that when you hit your Ranger, you get the short end of the stick. The only way you possibly test the true randomness would be to 1) run the same area with each character. 2) attempt solo then attempt with the same grouping of henchmem 3) Repeat it numerous times like several hundred per char.

Yes to truly be able to tell whether or not there is randomness, you have to do hundreds if not thousands of runs. Again keep in mind what random and what odds mean. Odds/Randomness doesn't mean that you are guaranteed anything, it simply means that on the drop that is occurring right then you have 1 in 6 shot (if in a party of that size) of getting loot. This could also be compounded if no drop is selected as that is an event as well. Run the above test and see.

You could be right that there is a bug that somehow keeps the particular class combo you have selected from getting the same share but I find it very unlikely. The reason why is that item generation and loot sharing is a function that has been around game programming since the inception of treasure/loot and drops, kind of like case statements or for loops and by now I am relatively certain that any issues that may have arised such as coding that causes adverse issues would have been discovered and corrected but who knows. Humans are the ones doing the coding so maybe someone did some sloppy coding.

The test runs you talked about also has to be ran with the same group and same alignment (meaning you're always the first/second/... person of the group). Like i said, if you do the die experiment, or the coin toss, you cannot get a correct result if you switch your choice around (joining another party). You have to always choose #1 or Head in order to test. So I guess it's very difficult to conduct a test in an online game that many odds (outside interference) could happen.

And I also don't think such a game that addresses balance would be coded with bias..

Algren Cole · Jun 07, 2005, 06:17 PM // 18:17

Quote:

Originally Posted by Arvydas

The test runs you talked about also has to be ran with the same group and same alignment (meaning you're always the first/second/... person of the group). Like i said, if you do the die experiment, or the coin toss, you cannot get a correct result if you switch your choice around (joining another party). You have to always choose #1 or Head in order to test. So I guess it's very difficult to conduct a test in an online game that many odds (outside interference) could happen.

And I also don't think such a game that addresses balance would be coded in bias..

not true. if you flip a coin and select heads as your choice you have a 50/50 chance of tossing heads....if you select tails on the next toss you still have a 50/50 chance of tossing tails.

Arvydas · Jun 07, 2005, 06:22 PM // 18:22

Quote:

Originally Posted by Algren Cole

not true. if you flip a coin and select heads as your choice you have a 50/50 chance of tossing heads....if you select tails on the next toss you still have a 50/50 chance of tossing tails.

Yup.. but if you are to count how many Heads you get, you can't switch to Tail all of a sudden.

Dwiggit · Jun 07, 2005, 06:26 PM // 18:26

Quote:

Originally Posted by Algren Cole

if I get some free time later this week I'll look into writing some rudimentary monitoring software to monitor all the drops that are assigned to each player. should make it easier to see what is actually going on....though I assure you it's a 1/# chance for every drop, and each drop has the same probability regardless of previous drops.(where #, again, is total party number)

Wow! If you could do that, it would be great!
Can you really do this without knowledge of GW code? (Or are you one of the game programmers?)

Thanks!
Dwig

stumpy · Jun 07, 2005, 06:32 PM // 18:32

The treasure drops are random ... but be patient your time will come ... for 1 week I tested the same issue with my R/Me ... all my friends we getting loads of purp and gold ... and me nothing ... damn what collector items ...

frusterating isnt it? ... well it paid off in the end because next week I started to get the gold drops ... and being the leader of a group I believe plays a plus in the equation ... after running the final mission with some noobs that invited me ... nothing ... not to mention the pug was terrible but others got alot spread fairly evenly. When I took the reigns ... and sorry to all others in my group at the time ... I got every rare drop. Every purple armor, every gold titans armor, etc. even a gold fellblade.

It is completely random ... going out on a map 2-3 times and getting frusterated and giving up wont help. Try 20 times then tell me you still got the shaft.

The algorithm will not be based on a 'group' ... think about it ... striking the server for a request every group with oh i'd say about 300,000 groups at any given time (*this would literally stop gw)

Instead I believe their algorithm is based on maps (so everybody in say in the glade ... *counts total people* ... *does calculation* and pre determines the treasure total treasure and holds it into a temp server for that map for a given time limit of say 2 minutes ... then recalculates the temp to reflect a more accurate number again) this lets them run there programs every few minutes and lightens the application pools.

Have you ever been able to run right outside of a town without a map change? ... i have ... and I am sure you have too ... guess what ...

Remember, number 1 in GW is security ...

FYI ... best time you will find a gold item is the busiest times of the day. (why ... because there are more people = more rare drops)

I can rake in 3-5 rare drops an hour during peak pst hours, but if Im on after 10 which i always am ... my drops tend to ... well be less than satisfying and
I tend to hit the arch to get rid of my loot.

Then again you could have pissed them off and your account is now cursed

Algren Cole · Jun 07, 2005, 06:35 PM // 18:35

Quote:

Originally Posted by Arvydas

Yup.. but if you are to count how many Heads you get, you can't switch to Tail all of a sudden.

but you're quantifying a probability...not the number of times that probability hits in your favor....either way it's 50/50 for every toss....there's no way to determine whether or not each individual toss is going to be in your favor based on prior history of tosses.

b1m2x3 · Jun 07, 2005, 06:36 PM // 18:36

im a monk, i've beaten the game, and i have NEVER gotten any gold or purple items while playing with others.

the only good stuff i've ever found was found alone.

Algren Cole · Jun 07, 2005, 06:36 PM // 18:36

Quote:

Originally Posted by Dwiggit

Wow! If you could do that, it would be great!
Can you really do this without knowledge of GW code? (Or are you one of the game programmers?)

Thanks!
Dwig

I'll just write an application that leaches onto the incomming connection and grabs the packets before they get to your client and analyzes who the drops are for. It'd really just be a basic packet sniffer...only it'd be customized specifically for Guild Wars.

Algren Cole · Jun 07, 2005, 06:38 PM // 18:38

Quote:

Originally Posted by b1m2x3

im a monk, i've beaten the game, and i have NEVER gotten any gold or purple items while playing with others.

the only good stuff i've ever found was found alone.

how much time did you spend playing alone in comparison to the amount of time you spent playing with other characters?

Arvydas · Jun 07, 2005, 06:45 PM // 18:45

Quote:

Originally Posted by Algren Cole

but you're quantifying a probability...not the number of times that probability hits in your favor....either way it's 50/50 for every toss....there's no way to determine whether or not each individual toss is going to be in your favor based on prior history of tosses.

Yup. There's no way to determine the individual toss. But what Night meant for the experiment to be conducted was to see is there a fair distribution for a particular character over time. So if you're counting head in one game and tail in another, you're not doing much.

But even so, I agree that history of toss cannot tell you what you'd get the next time.

Algren Cole · Jun 07, 2005, 06:52 PM // 18:52

Quote:

Originally Posted by Arvydas

Yup. There's no way to determine the individual toss. But what Night meant for the experiment to be conducted was to see is there a fair distribution for a particular character over time. So if you're counting head in one game and tail in another, you're not doing much.

But even so, I agree that history of toss cannot tell you what you'd get the next time.

ahh I gotcha now. I've completely dismissed the whole "fair vs not fair issue"....as I'm certain that it's random.

Arvydas · Jun 07, 2005, 07:02 PM // 19:02

Quote:

Originally Posted by Algren Cole

ahh I gotcha now. I've completely dismissed the whole "fair vs not fair issue"....as I'm certain that it's random.

I'm certain that it's random too.

C-Tzar · Jun 07, 2005, 07:06 PM // 19:06

I'm beginning to wish the drops were not completely random. If I had control over the loot drops, every account would have a "phat_lewt" field associated with it which would contain a percentage modifier that could be adjusted.

Spamming 1337 speak: -2% phat_lewt penalty.
Getting your whole party killed because you mistakenly think you're Superman: -3%
Propogating conspiracy theories about loot drop fairness: -2%
Throwing out vague, pretentious ideas about random number gerneration and/or chaos therory, etc.: -5%
Whining about phat lewt on a public forum: -4%
Being an arrogant jerk: -50% (wow, I'd be poor)
Creating a W/M: -3%

Wow, I could go on indefinately with this list.

Algren Cole · Jun 07, 2005, 07:12 PM // 19:12

Quote:

Originally Posted by C-Tzar

I'm beginning to wish the drops were not completely random. If I had control over the loot drops, every account would have a "phat_lewt" field associated with it which would contain a percentage modifier that could be adjusted.

Spamming 1337 speak: -2% phat_lewt penalty.
Getting your whole party killed because you mistakenly think you're Superman: -3%
Propogating conspiracy theories about loot drop fairness: -2%
Throwing out vague, pretentious ideas about random number gerneration and/or chaos therory, etc.: -5%
Whining about phat lewt on a public forum: -4%
Being an arrogant jerk: -50% (wow, I'd be poor)
Creating a W/M: -3%

Wow, I could go on indefinately with this list.

thanks alot! now everyone in my office is wondering exactly what it is about SMTP Protocol Patches I found so humerous.

Jun 07, 2005, 06:08 PM // 18:08	#42
Dwiggit Frost Gate Guardian Join Date: Apr 2005 Guild: Heroes Unlimited (HaCK) Profession: R/Mo	Some thought on the matter! Night, Algren Cole, thanks for the helpful responses, and an open mind that something other than perception can be at play. Sounds like both of you have some programming experience. I was trying to avoid the scientific study of the situation, but I guess that is what it will take to convince people. I would rather just play the game and have fun, rather than logging all the loot drops for a month. I wonder if AN had some testing software that logs this stuff? It would make testing this a lot easier. Thanks for helping! Dwig

Jun 07, 2005, 06:15 PM // 18:15	#46
Santosh Academy Page Join Date: May 2005 Guild: Cult Unseen Profession: Mo/E	It is random. I've noticed whether I play with henchmen or with other players, the number of loot drops equally for each party member. Yes, randomly but equally. Not quality-wise, but number-wise. That's the difference. Some person in a group might get really nice drops during a mission, but it's all based on luck. For example, if I was in a 6-man group, the drops would be in a rotation. Keep in mind that the mobs have already set loot drops in them for a particular mission when you zone in. Player 1 (ranger) gets the first 2 drops. Player 2 (war) gets the next 2 drops. Whatever each person gets depends on what that mob has on them. Coins or items. So on and so forth. It doesn't matter what class you play, it doesn't matter how much damage you do. So if random mob has a gold item, the lucky one will get that drop if it's their turn in the group rotation. That's what I've noticed.

Jun 07, 2005, 06:32 PM // 18:32	#51
stumpy Jungle Guide Join Date: Feb 2005 Location: Canucklehead BC, Canada Guild: Advanced Necro Undead Society	The treasure drops are random ... but be patient your time will come ... for 1 week I tested the same issue with my R/Me ... all my friends we getting loads of purp and gold ... and me nothing ... damn what collector items ... frusterating isnt it? ... well it paid off in the end because next week I started to get the gold drops ... and being the leader of a group I believe plays a plus in the equation ... after running the final mission with some noobs that invited me ... nothing ... not to mention the pug was terrible but others got alot spread fairly evenly. When I took the reigns ... and sorry to all others in my group at the time ... I got every rare drop. Every purple armor, every gold titans armor, etc. even a gold fellblade. It is completely random ... going out on a map 2-3 times and getting frusterated and giving up wont help. Try 20 times then tell me you still got the shaft. The algorithm will not be based on a 'group' ... think about it ... striking the server for a request every group with oh i'd say about 300,000 groups at any given time (this would literally stop gw) Instead I believe their algorithm is based on maps (so everybody in say in the glade ... counts total people* ... does calculation and pre determines the treasure total treasure and holds it into a temp server for that map for a given time limit of say 2 minutes ... then recalculates the temp to reflect a more accurate number again) this lets them run there programs every few minutes and lightens the application pools. Have you ever been able to run right outside of a town without a map change? ... i have ... and I am sure you have too ... guess what ... Remember, number 1 in GW is security ... FYI ... best time you will find a gold item is the busiest times of the day. (why ... because there are more people = more rare drops) I can rake in 3-5 rare drops an hour during peak pst hours, but if Im on after 10 which i always am ... my drops tend to ... well be less than satisfying and I tend to hit the arch to get rid of my loot. Then again you could have pissed them off and your account is now cursed

Jun 07, 2005, 06:36 PM // 18:36	#53
b1m2x3 Ascalonian Squire Join Date: May 2005	im a monk, i've beaten the game, and i have NEVER gotten any gold or purple items while playing with others. the only good stuff i've ever found was found alone.

Jun 07, 2005, 07:06 PM // 19:06	#59
C-Tzar Krytan Explorer Join Date: May 2005 Location: North AL, USA Profession: E/Mo	I'm beginning to wish the drops were not completely random. If I had control over the loot drops, every account would have a "phat_lewt" field associated with it which would contain a percentage modifier that could be adjusted. Spamming 1337 speak: -2% phat_lewt penalty. Getting your whole party killed because you mistakenly think you're Superman: -3% Propogating conspiracy theories about loot drop fairness: -2% Throwing out vague, pretentious ideas about random number gerneration and/or chaos therory, etc.: -5% Whining about phat lewt on a public forum: -4% Being an arrogant jerk: -50% (wow, I'd be poor) Creating a W/M: -3% Wow, I could go on indefinately with this list.