Talk:Drop rate

Initial Conversation
Good idea, Karlos! Two notes from my side: -- 19:31, 22 February 2006 (CST)
 * Dye: You make it sound like the drop rate depends on the price. Don't you think it's the other way round: The less often a dye type drops, the lower the supply (in relation to demand), and hence the higher the price?! So off course you can make the reverse conclusion and use the price as an indication of drop rate, but I find the current wording a bit confusing.
 * Should we start collecting data about what type of "Equippable Items" are dropped? There seem to be specific relations. Just one example: From my experience, Skeletons in the Catacombs and in Kryta drop Sephis Axes more often than any other creature.


 * Thank you. I fixed the dyes wording. I agree, it soundedlike price was the determining factor.
 * As for collecting more specific drops info, as I suggested in the project points. Create a second table under the main one with the breakdown of the equippable items, if you want that. You can go as specific as you want (e.g. how ofen they drop blue or purple or even what are the stats of the weapons they drop). This will be useful for those looking for Fellblades and Chaos Axes and so forth. I never found it interesting to I never collected it, but the more info we have, the better players we will be. --Karlos 19:41, 22 February 2006 (CST)


 * I can see this kind of data being useful to some people, but I'm afraight most types of creatures aren't farmed enough to collect the necessary amount of data to come up with a statistic that has any significance. Especially for "rare" items like Fellblades, Sickles and such. You'd have to kill hundreds of foes of each type, and write down every single drop. It'd be hard work. It would be nice to have, but probably it won't happen. -- 19:47, 22 February 2006 (CST)


 * It just so happened that I have been 2 manning the UW extensively when I was saving for my eles FoW armor. So, I started to jot down this stuff. I am now trying to focus on the FoW critters. Will start noting their drops as well. I have this unconfirmed theory that Shadow Army drops more than Skeletal Army which drops more than the plants and the spiders. Will see what the numbers say. --Karlos 23:38, 22 February 2006 (CST)


 * Good idea! I'll smush some trolls after school :D 21:14, 22 February 2006 (CST) EDIT: possibly this should go to Drop rate


 * The article drop rate should definitely exist. If we wish to move the actual data collection to the Guildwiki name space, that's fine with me. --Karlos 23:38, 22 February 2006 (CST)


 * Nice idea! A little tedious for me, but nice otherwise. I wish we could get a "thottbot" like program that just watches the drops. But no UI, mods, etc. so it will have to be done this hard way. Good luck with the project :) --Ravious 21:46, 22 February 2006 (CST)


 * I agree the data should go in its article, re: Drop rate. Kinda like a notepad till the numbers are finalized for each monster's page. Can't wait to sic my 55 monk on some areas, maybe the desert, south Kryta, and Snake Dance. :D Good idea!!
 * I do have one little problem. Areas where full or almost full parties are needed, say Summit groups or Murasaat, you would have to have a full "human" party, no hench, to get the correct data. If you had hench and only could record the drops you recieved(8/8 party with all hench), wouldn't that askew your findings since you wouldn't be able to see all the drops? --Gares Redstorm 02:59, 23 February 2006 (CST)


 * Having NPCs in the party would not be a problem (if they simply take one "drop spot") as long as you report the number of NPCs with you. Basically we would adjust the drop rate by scaling it up with the number of non-reporters in the group. Of course, if you go with 7 NCPs, you would have to kill 8 times as many monsters to get the same amount of data comparable to going alone. --Xeeron 03:07, 23 February 2006 (CST)


 * For high-difficulty areas like these, then small tightly built groups are the best.. For example, Oro farming groups for SF or 5 man FoW groups if you can't do 2 man. The problem with going with other players is that some data might not be easy to capture unless all party members are collecting data too. For example, if the item is white, blue or purple, you will not know from the message on the chat log.
 * Personally, what I do in such bigger groups is that I record the drops that dropped for me, that's it. As Xeeron says, it will take 8 times the number of kills. But it's better to have little data than no data. Avoid big groups and hectic situations, like thunderhead Keep's siege finale since errors will likely occur. In time, errors will be insignificant as the averages will begin to shape themselves. --Karlos 08:56, 23 February 2006 (CST)

Take no disencouragement from this please, but note that the amount of work (ie the number of monsters to be killed) that needs to be put into this to yield half reliable statistics is HUGE. So dont expect any quick results from this. It is a very worthwile task, but it is going to be some beast of a task as well. --Xeeron 03:07, 23 February 2006 (CST)


 * True, I want to know if my theory on Aatxe is true or not. It is amazing to me that they have been consistently dropping more Ecto than Smites even though they are encountered earlier on in the adventure (hence they have NEVER dropped any gold items for me and easily drop nothing at times). There is no question that there are variations in drop rates though, as evident by farming Gypsie Ettins in Kryta vs farming Maguuma Centaurs. The Ettins drop a LOT of gold and purple armor, the centaurs rarely ever drop any such colors. --Karlos 08:56, 23 February 2006 (CST)


 * If you want a LOT of easy gold armors, kill the beastriders in the Borlis Pass, solo'd with a lvl 20 and I've never gotten less than 50% of their drops be gold armors. --Rainith 12:35, 23 February 2006 (CST)

Question: Does the level of a monster affect the drop rates? Example: Me and a couple of others in my guild notice that lvl 8 Charr give more Charr Hides than lvl 5/6. Should creatures of the same species but different level be tested as well? From my experience, I would have to say yes. --Gares Redstorm 04:06, 23 February 2006 (CST)


 * I believe it does, though not in terms of hides. For example, I once got a Lump of Charcoal from killing a Hulking Stone Elemental, ONCE in over a few hundred kills over 5 characters. Yet, killing Imps in Kryta (who drop granite or scales) gives Charcoal fairly easily. On the flip side, killing the Mergoyles (who are of the same level) gives no rare materials. As fr hides, I think it's pretty consistent even across species. The Ettins, Charr and Centaurs all drop them at roughly the same rate for me. --Karlos 08:56, 23 February 2006 (CST)


 * I believe you may have misunderstood my question/statement. What I was referring to is if there is a difference in drop rates between species of one level and the same species of another level. Do you think level 8 Charr have the same drop rates as level 5 Charr? Or for a more widespaced example, level 10 Stone Summit and level 24 Stone Summit? Obviously, the level 24 Stone Summits will give better/higher priced items than their level 10 counterparts, but would they have different drop rates as well. I, myself think that different leveled, but same species monsters have different drop rates. Just wanted to get a concensus. --Gares Redstorm 09:36, 23 February 2006 (CST)


 * I understood the question. Allow me to answer a shorter, simpler answer: No, I do not think that is the case. :) I think that Charr will drop carvings and hides and armor at the same percentage. I, however, do not have any data to support this, just my feeling based on fighting them. --Karlos 12:25, 23 February 2006 (CST)


 * I think it would be helpful if, on the statistics gathering page, you listed what level the creatures were. Then we could (hopefully) confirm your feeling and stop noting it or discount it and increase the burden of this huge undertaking even further. So, data being logged? Drop, creature, location, and level? ;) Shandy 19:14, 23 February 2006 (CST)

Nothing drops reduce in probability with the more time the user spends in an area? This is very interesting. Is it dependent purely on time, or does it decrease with each monster kill (or some other way)? Because if time is the only factor, it seems to be a good idea to enter the Underworld then go make a coffee and read the paper then come back and farm Aatxes. --Qian Khan 07:05, 25 February 2006 (CST)


 * I am under the impression that the number of mobs killed is the decisive factor. It definitely isn't the time alone. --Eightyfour-onesevenfive 07:48, 25 February 2006 (CST)


 * That's a shame. If anyone can confirm this, it ought to be changed from "and drops quickly as players spend more time in the area" to something like "and drops for each foe killed in the area" --202.164.194.254 07:59, 25 February 2006 (CST)


 * That is what I meant by spend more time. I never meant time standing around doing nothing. :) I changed it. --Karlos 19:38, 25 February 2006 (CST)

Maybe a category for drop rate data gathering articles? It would seem sensible to have an index of the research progression. Shandy 21:49, 27 February 2006 (CST)


 * Someone could test if characters at different levels get different drops (Better when lower level). I have noticed that my level ~10 characters might get gold and purple items from Mergoyles outside the Gates of Kryta, but my level 20 characters don't. Gem 01:48, 3 March 2006 (CST)


 * I think that's the "new comers get a freebie" theory that I have. See my charr doodoo theory in th section immediately below. --Karlos 05:42, 3 March 2006 (CST)


 * I'd like to discuss on what theories you all may have considering this: When you zone into a map, the possibility of (some or all) equippable items are already fixed (as well as enemy mob placement).  Of course there are exceptions and exclusions, but generally what I'm speaking about is that certain enemies are holding equippable items that can be seen, and killing them has a chance of dropping that item.  For instance, when I was fighting a smite crawler in the UW the other night, it had a 'Protective Icon' one of it's legs (I took a screenshot since it was so unusual to see).  I killed it, and it dropped it.  I assume there is the chance it could of dropped something else, e.g. Phantom Residue.  I assume that this theory is true, taking from past enemy encounters, and also from the bosses in SF, where they have items in which they drop, that relates specifically to them.
 * Of course, enemies may drop equippable items that they are not showing/holding. When farming the FoW, I killed Doubter Dryders and they have dropped gold Chaos Axes and Shadowblades, when usually they are holding some sort of Cane, and regularly drop 'Jeweled' items.  This is the exception that I mentioned, since most, if not all enemies in FoW can drop FoW-related items which are the Chaos Axes and Shadowblades.  My conclusion is that in addition to the area-specific equippable item, enemies can also drop items in which they are wielding, and this is important to know because it is determined at the loading of the map, and also determines any possible mods.
 * In the past, I had made runs in Riverside Providence to the Hill Giants. I would have to run through some White Mantle, and some of the Rangers would have Vampiric Bows.  I knew this because I saw the 'Life Stealing' effect on myself, and could determine the ratio.  Unfortunately, Vampiric mods are one of the few in which you can determine so easily from enemy groups.  Other weapon mods such as HP+30 are impossible to figure without killing the enemy.  This also relates to the 'anti-farming' pop-up message that you get from rezoning and killing a certain enemy mob (although I think rezoning has more of an effect).  Every time you zone in a map, you are given a certain "percentage" number of recieving drops.  You can say this determines how 'lucky' you are on this particular run.  Every time you rezone, your percentage drops, and you are more likely of getting 'nothing' drops, and a lot of whites.
 * Another note, I am unsure if rezoning after dying will lower your drops. Specifically I mean when dying in areas where you are forced back into town, like the FoW.  I had completed over 10 solo runs in FoW, dying purposely each time that I was finished with the run.  I was still able to receive Obsidian Shards after the 10+ runs, although I did not keep track of my gold or purple drops.  I plan to do more study on this, and if I am able to receive gold or purple drops after so many runs, I may come up with a definitive conclusion. --XGZ 10:28, 26 March 2006 (CST)
 * I have also noticed that monsters tend to be using the items they dropped. When farming trolls outside droks certain trolls die a bit slower and these trolls always drop a shield. I haven't noticed vamp however I've never found vamp weapons either. Also, with henchies I noticed a shield on a monster, took a screenshot and thought it would drop but obviously the henchmen got it instead of me. On the Shing Jea islands fighting a boss with some henchmen I noticed he was doing a lot of damage and he dropped his green bow. This might hold true for any item and be useful for casters by checking for that 20% of whatever skills they want.--Apocrypha 20:29, 20 June 2006 (CDT)


 * These are interesting points. Here are my thoughts:
 * Yes, monster will only drop what they are alrwady equipped. i.e. their drops are determined at area rendition. Once the area is rendered, any monster in it will only drop the equipable items that have been generated for him by the system. You can notice this very distinctly with Smite Crawlers. Sometimes you'll see your SS doing 41 across the board, and then there is this one loaner who is getting only 39 damage. He dies, and low and behold, he drops a -2 damage shield. However, they do not ALWAYS drop that visible item. Sometimes the crawler with the shield would drop gold or Phantom Residue.
 * On point number two, I think that each monster is given some items. i.e. 2,3 or maybe even 4 equippables. These equippables may even be sorted in value. i.e. each creature has a gold item, a purple item, a blue item and a white item. However, the chances of dropping the gold item are, say, 5%, the purple item 10% and the blue item 25% (within the overall chance of dropping an equippable item to begin with). This leads to the next point.
 * I don't think "percentage number" is the right term, because I think the game tries to alternate drops between party members randomly. However, it maybe that the above mentioned theory I mentioned (5% gold, 10% purple) gets affected by how much you come to an area and how long you stay there.
 * I have done countless FoW 2-man runs and have found the shard drop rate to be consistent. You can get a run where only one shard will drop after clearing 3 areas, then get a run where you get 3-4 shards just clearing the tower of strength. I believe shards and ectos are misleading in calculating gold and purple drops. I think these two things are apples and oranges. I DO believe that the rate of gold drops is affected by frequently farming an area, I do not believe the rate of crafting materials dropped is.
 * --Karlos 01:16, 21 June 2006 (CDT)


 * In my eoe-barrage solo nolany runs, (in which I open the nolani gate, cast eoe, shadowstep down, and barrage around 15-20 charr), I have notices a major decline in the amount of drops per monsters, after 20 or 30 runs. I could go on a run where only 4 our of 15 charrs droped something. this could be a good place to expiriment the effects of death vs maping. (can use vamp weapon, or better yet, wait for a smaller second wave of charr). Foo 03:43, 21 June 2006 (CDT)


 * I tried Nolani with my E/Mo and Phoenix a while ago. I killed everything that came through the gate (usually about 10-15 Charr, including 1 boss) plus 2 or 3 more Charrs outside. I sold all the stuff I got (except for 1 colour). First run yielded a few hundred gold, but it quickly declined to about 50-80g, and after about 10-15 runs I came to the point where I killed those charrs and got NOTHING, no gold, no material, just nothing. I "checked" the other groups of Charr farther away and almost all of them dropped loot. By the way: A different character of mine experienced normal drop rates after this in Nolani, but they also declined just as quick. What I wonder: If somebody clears the entire Charr army, will all of them stop dropping loot? Because then it is not about placement in the map but just how often a certain group is killed. --Ts 12:14, 11 July 2006 (CDT)

Dynamic drop rates
I don't know the details, because I've never been much of a farmer, but from what I've read on forums ANet have implemented a dynamic drop rate system to counter farming. The more people farm a certain mob type in an certain area, the lower the drop rates. If that is true, then the whole drop rate project is somewhat pointless, because not only will the drop rates depend on what you farm, but also where you farm and when you farm. Does anybody know the details of the dynamic drop rate system? -- 22:50, 27 February 2006 (CST)


 * I've read that, too. In my opinion those are rumors, started/spread by paranoid people ("Oh noes! I didn't find three superior runes this ettin run! Anet must've nerfed the drops!"). Maybe there is some truth to it, but if so, this project may be the one which confirms it and makes the rumor become fact. ;) There seems, however, to be some kind of anti-bot-farming code in effect. It never happened to me, so this, again, I know only by hearsay. When you farm one spot very excessively (say, buy repeatedly killing only the very first mob in an area), you will appearantly get a hint window which tells you that your drops will degrade. I've seen a screenshot of that somewhere in the guildwarsguru forums. --Eightyfour-onesevenfive 23:05, 27 February 2006 (CST)


 * Yes, that pop-up window is definetly confirmed. I've seen screenshots of it, too. But the not-so-obvious dynamic drop system running hidden in the background server-side may be true too. It is my experience too, that you get the best drops if you explore remote areas where very few people farm regularly. And I seem to recall somebody from ANet talking about a dynamic anti-farming system in an interview or so. -- 23:19, 27 February 2006 (CST)


 * As someone who farms ANet's two most precious farming locations, FoW and UW, I can attest that this theory of nerfed drops is, as you might expect me to say, Charr doodoo. I 2-man farm both places every evening for 2-4 hours. Drops have been fairly consistent. There will be the occassional Ectoless run and the Ectofull run as me and my buddy like to call them, but it evens itself out. Shards drop more than ectos because everything in FoW drop them, while roughly half the creatures in UW 's farm areas do not drop them.
 * If you do kill the first few monsters in an area you'll get this warning: (hangs head in shame)
 * [[image:farming_warning.jpg]]
 * The text is very clear on what gets nerfed. Farming runs usually entail more than killing a few creatures then mapping out.
 * Now, I do have a Charr doodoo theory of my own, which is that new comers into an area usually get good loot early on. It has not always held true for all my characters, but 3 of the 4 got early shards and ectos in the Realms the first time they were there. Also, whenever one of my characters has been away from either area for sometime, when they go back, they often (not always) get a few good drops early on. Increasing the feeling by the others in the party that their drops have been nerfed. Overall, however, drops are very consistent. When I post the data for FoW (over 1500 kills recorded) you'll see that (with the exception of the Dryders and the Skales) everything follows an obvious pattern. --Karlos 08:50, 28 February 2006 (CST)


 * After farming the first mob on a map for a while, I got that first farming warning. I continued to farm them for a few more times to make sure it wasn't a scare tactic, then decided to solo Fort Ranik (somewhere I hadn't been since I was level 4), in an attempt to increase the drop rates by killing different monsters and completing a map. It was pure chance that, while killing the first mob of Fort Ranik, Guild Wars gave me this warning:
 * [[image:Farming_Warning_2.jpg]]
 * This suggests that the dynamic drop system is more affected by the relative location in the map of the monsters you're killing than by what/where the monsters actually are. --Hammer Bro. 17:18, 27 June 2006 (CDT)

Hides
One small thing that just came to my attention: Hides (as in the type of drop as specified in the article) are not always called "hides". Some kinds of spiders drop spider webs (i.e. Maguuma Spider Web), which bear the same characteristics as hides. Nothing serious, just wanted to mention. :) --Eightyfour-onesevenfive 23:14, 27 February 2006 (CST)
 * Yes, and there is the half-digested food too. See: Salvage Item. -- 23:19, 27 February 2006 (CST)
 * Technically, you are correct. These "Salvageable Collectable" drops are the same. I was obssessed with hides cause they drop WITH armor, the webs and the hal-eaten blobs drop as opposed to armor. What should we call that group? "Stackable Salvage Items"? --Karlos 08:52, 28 February 2006 (CST)
 * That's what I called them on the excel sheet I started last night, "Stackable Salvage." I dropped items, felt redundant at that point.  --Rainith 09:38, 28 February 2006 (CST)
 * I have Rainith Attunement on me!! :) Whatever name we decide on here just stick it in the article. I put "salvageable remains" as a temporary name better than hides. --Karlos 09:50, 28 February 2006 (CST)
 * Wait until you check out my respose on Talk:Heads Up Display. I'll start have people asking me in-game if you and I are the same person again.... ;)  --Rainith 10:07, 28 February 2006 (CST)

Droprate that change over time
It is theoretically possible that drop rates can be altered in teh background by Anet w/o us knowing it.

I advocate that the date of drop rate data be added, and we make drop rate tables analyzing for each of the following (if available): over the past week, over the past month, over the past 3 months, over the past year, overall. -PanSola 01:39, 2 March 2006 (CST)


 * I was thinking about that. Not because of drop rates being altered (if my experience in the fissure over the past 7+ months is any indication, they don't change much), but it would be better to have each data entry batch in a separat row. This way we can also see if a batch had strange data, or if a new tendency in drop rates evolves (or account for things such as Candy Cane Shards dropping for a short while). So, I will alter the steps of the process in the article now. --Karlos 13:33, 2 March 2006 (CST)

Drop data in articles..
I am thinking we have to find a better way to keep the monster article updated with the drop data. I suggest that we have the data page be all except the part that has the summary of the drops and percentages, that part would be includable. Then the monster article would include the drop rates article of that monster, and thus only the percentages would show up. How is that? --Karlos 05:27, 3 March 2006 (CST)


 * I think the percentages are a bit too specific for this. Maby the note above the drop rates portion of the monster page could be more specific and state that these percentages are calculated from drops received by some wiki members. Dyes and keys propably shouldn't be shown on the monster page as every monster has a small possibility to drop them. The 'no drop' -perccentage should also be shown, even if it varies acccording to the amount of monsters killed earlier in the instance (Or am I mistaken?) because the percentage of 'no drops' varies hugely on some monsters. --Gem 05:54, 3 March 2006 (CST)


 * The "no drops" definitely needs more research first. When I look at the data collected so far, I see one thing: Mobs that you can't find at the very beginning of an explorable area (i.e. Smites) have a very low percentage of "no drops", while others that you tend to find at the beginning (i.e. Minotaurs) have a very high percentage. I would pet my IDS, that would you collect data on Minotaurs, after first killing a dozen or so other mobs, you would find the "no drops" rate dramatically decreasing. --Eightyfour-onesevenfive 06:02, 3 March 2006 (CST)


 * A good example of this might be the Smite drops of me and Karlos. Karlos has never seen any zero drops from Smites, but I see them frequently. I tend to kill the minimum amount of enemies before progressing to the Smites, which might be the explanation for this. --Gem 06:17, 3 March 2006 (CST)


 * Well, this is the heart of this project, to find out what data is interesting. For example, if dyes and keys have exactly the same rate in any area with any mob, then you are correct, there is no point in listing them under every mob. Just mention their rate in the key and dye articles. But.. If a creatures or an area drops more dyes than usual, then I think that should be note don their drop section.
 * The includable section will need to be customized per monster. i.e. For an Aatxe, it should say Piles of Glittering Dust and Ecto, not Common and Rare materials. This way, when it is included, it makes sense in the article. --Karlos 07:13, 3 March 2006 (CST)


 * So we should edit the tables for all of the monsters so that they do not include item types that can't be dropped and so that the general names are replaced with the specific ones. Ie remove Salvagable armor from Aatxes and change rare materials to Ectoplasmas. --Gem 09:18, 3 March 2006 (CST)

Not sure what section to post this under so I stuck it down here. Will all or most monsters be covered in the drop rate reseach or just those that have rare drops. Or is it basically whatever creatures are entered into the research article Wortan 07:19, 21 March 2006 (CST)


 * Any creature you feel like tracking. It's driven by interest. I play the Realms almost exclusively, so that's what I monitor, you can monitor any critters you like. As long as they are fuzzy and cuddley. :) --Karlos 07:52, 21 March 2006 (CST)

Chest data?
Could we start recording chest drops? Professions, rarity and so on, I just saw a GWG thread about that =) 02:31, 22 March 2006 (CST)


 * I actually suggested this a long time ago somewhere in the wiki. I really think we should do this atleast for the most precious chests (Phantom, Miners, ...) 02:34, 22 March 2006 (CST)


 * I say for any that drop gold items as they might not be max damage, but can have max mods. I have salvaged more than a couple of perfect mods from gold items found from Kryta chests. Thats as far down the "chest ladder" I have found gold items in. --Gares Redstorm 02:47, 22 March 2006 (CST)


 * Please go ahead. I'll start a template for Obisidan Chests and you guys can modify that and then use it for other chests as you like. --Karlos 04:37, 22 March 2006 (CST)


 * Hey if you are talking about the chest run for the Darkstone chests on GWG, yes that was me that started that topic. Although it was a short sample, someone else was able to do a seperate sample and provide more data.  The data is still in that thread, so if someone could do it for me and input the statistics, that would be great.  I am new to wiki and don't feel like learning it. :) --XGZ 10:30, 26 March 2006 (CST)

Your project is flawed. There is a correlation that alters the drop rate for chests and that will yield nearly 100% gold drops at least for the Southshiver Peak and Fire Island areas. I have been testing it by farming chests in that area for several days and it has steadily held true.


 * Do you mean that those chests now only drop gold items? --Gem [[Image:Gem-icon-sm.png|Gem]] 05:47, 8 May 2006 (CDT)


 * A theory unstated is by-design, irrefutable. I cannot refute, prove or disprove your correlation if you do not state it. To say that there exists "a correlation" that "nearly" makes all those chests drop gold and not state it means I have to do one of two things: a) Guess, or b) Beg you for your correlation. I will do neither. If you wish to state it for the benefit of others, please do. I am sure everyone would like to have their 600 gold keys give back gold items. thanks. --Karlos 06:33, 8 May 2006 (CDT)


 * This is where my almost-native englich skills shows up. Everyone uses fancy words and I'm not sure if I understand it all correctly, although I try my best. --Gem [[Image:Gem-icon-sm.png|Gem]] 06:44, 8 May 2006 (CDT)

My comment stands. There is a variable that alters the rate that gold items drop at in the Fire Islands and Southshiver Peak areas. We have limited experience outside these areas but are collecting data when we play there. Currently I can not garrantee a gold item drop every time as my data are still not sufficient. Hence, we are not ready yet to publish it. It is sufficient to say that we now beleive we know when to use a key with a very high probability of a gold drop and when not to. I have to admit that greed is at war with the desire to publish on this issue as I have contributed in the past. At present greed is winning. We have demonstrated this to a small group of players without telling them how we know and they have experienced the phenomenon for themselves, multiple gold drops from a single chest and a sequence of chests. Needless to say we now have a small loyal group of players that are anxious to add us to their gaming groups. The technique is not hard to discover if you understand the philosophy of the game. I can not take credit for it as my spouse made the original observations. If you require a demonstration, one could be arranged. Find us and we will show you. We are online almost everyday for at least 4 hours, usually 8 or more, but only two of us will acknowlede the contact and conduct the demonstration. Oh, there also seems to be a similar but less obvious correlation for creature drop as well but our data is less convencing at present due to a smaller data set. In the spirit of the game we remain: D.F., D.M., V.B.A., K.H.L., S.D.


 * Why do you come to a wiki bragging that you have found a secret to know if a chest drops gold items or not, but don't share the secret or your names so that we could contact you? A wiki is not a place for this kind of behavior. If you want to help, please share your secret or atleast your in game names so we could see you in action, but don't just come here to say "We know something you don't. Screw that! We aint gonna tell you!" --Gem [[Image:Gem-icon-sm.png|Gem]] 08:47, 8 May 2006 (CDT)


 * Again, more boasting, less facts. Nothing I can do really. I don't even have contact info to watch you "demonstrate" this theory. You have my contact info in-game on my user page. If you believe you have something to show me, feel free to contact me. On the flip side, even if you do show me that you opened X chests and they were all gold, I am not really sure what to put in the wiki to describe that... "Joe Schmoe has demonstrated to the wiki that they can hack Shiverpeak chests?"
 * Like I said, I will neither waste my time guessing nor my energy begging. You need to discover the worth of sharing this information with others on your own. --Karlos 08:59, 8 May 2006 (CDT)

Hey, I know it! there IS a correlation with the value of 1 between the chance that I'll get a gold drop from a chest, to the chance the one who opened it before me, got a gold drop from it! :p Foo 10:05, 8 May 2006 (CDT)

A good start Foo but not nearly the whole story and we are not sure that number is one but it is close to it. We will publish our work when our data set is more complete. Hate making errors. Kind of compulsive in that way. Typical of scientists. By the way, no hacking involved. Just a particular style of game play readily adapted to by the average player and not not likely to be changed by the game designers without major disruptions to their game design. Not a farming technique per say. D.F., D.M., V.B.A., K.H.L., S.D.


 * Well, if you need a mathematician to figure it out, you know where to find me. M.F 14:33, 8 May 2006 (CDT)

Questions for M.F.: What is the required size of a data set to prove with 95% and 99% confidence the validity of a hypothesis with the only possible results from the data points being true or false? All the data points to date have been true. Want to know how much more data we need. I should know where to get this data but it has been a number of years since I had statistics and my profession does not require any regular use of it. By the way, nice linked page. Hope to see you in the game. We will say hi. D.F., D.M., V.B.A., K.H.L., S.D.
 * A fast answer would be ~15 for 95% and ~95 for 99%. a more sure answer would take me another day or two. Foo 16:23, 8 May 2006 (CDT)

Thanks. Should have enough data for single player and henchies to be 95%confident in about a week. Current data set is 9 of 9 when conditions are right to get gold drop using multiple character mixes. Type of drop data less revealing. Majority of items are not useful to the character running the henchies but with 8 character types that is to be expected. Still a bias for drops less useful to the active character seems to be emerging. We have purposely not generated one character type to facilitate the test. Less data on negative condition set as it is more expensive to test that condition. Worried about the economic effects of revealing our results too soon as the knowledge is funding the research. Imagine the effect of every key producing a gold item on item prices. Guildwars is sensitive to impacts on their economic system and it would also have impact on the players, both positive and negative. Still debating ethics vs enevitability of the spread of the discovery and impact on the larger game environment. Harder to get data for full parties as some character mixes are not readily accepted into groups. Data set smaller there but growing. Very furstrating watching the parties create the wrong conditions early in the run. Don't want to be labeled as leaver but wastes a lot of time. Makes testing a very time consuming process.

Something changed with chest drops in the recent update. Prior to update had a record of 13 of 13 gold drops when there had been no player or henchie deaths in the party prior to opening a chest in the Fire Island and Southshiver Peaks areas. This observation no longer holds since the May 10 update. Now get mix of gold and purple in roughly a 1 to 2 ratio. Limited data is not sufficient for statistical analysis at this time. D.F., D.M., V.B.A., K.H.L., S.D.


 * I could have proven you wrong a long time ago. I have wondered in Perdition Rock and Southern Shiverpeaks with my running build and opened some chests, not all golden. No party, no deaths. Sorry, but your 13 chests were a nice coincidence only. --[[Image:Gem-icon-sm.png|User:Gem]] 11:12, 15 May 2006 (CDT)


 * Wow, wish you would have said that long ago, I would have told you it makes absolutely no sense to tie btween deaths and chest drops as well as the fact that ususally the first chest a party opens (usually with no deaths) is often purple. --Karlos 18:35, 27 June 2006 (CDT)

Some Chest Drop Theories
Ok, I have been hanging out with some of the Treasure Hunting crowd in those Faction Farming runs, here are some theories/myths that they have about chest drops: Anyone has any experience regarding either and/or more theories? --Karlos 18:02, 27 June 2006 (CDT)
 * 1) Related to the apparent wealth of the player opening the chest: Some believe that if you have Gold items in your inventory this negatively affects the chances of the chest being gold. It has been demonstrated to me a few times, but it is not like automatic. i.e. I have opened chests first with no gold items in my inventory and got purple, and I have opened chests first with gold in my inventory and got gold.
 * 2) Related to Party size: i.e. to counter chest runners, the game rewards a full party with more gold chests than a party of one. This seems more likely to be true. You can FF solo or with a party of 8, and I can attest that at least when I am running solo, I seem to get a lot less gold Kurzick chests, than when I am in a party.


 * The first is not true according to my experiences. I carry all my gold bows with me all the time and I still get about the same amount of gold items as before. Party size however might affect the gold/purple status of chests. --[[Image:Gem-icon-sm.png]] 01:36, 28 June 2006 (CDT)
 * Lets start recording this in our data. But keep in mind that point 2 is really self fulfilling, every chest will drop purple or gold items, so if you go in large parties only one member has to take a risk on a chest, thus you will get a higher rate of gold items per key. --Draygo Korvan (Yap) 11:12, 28 June 2006 (CDT)

Drops of shards and ectos
I have heard a rumer that ectos and shards have stoped droping. that was while me and another guildie was seperatly soloing the fow cave. we did not get any shards, (The chance that a total of 102 Armored Cave Spider and 24 Snarling Driftwood will not drop any shards, is 1.9%). in the same time, traders' prices are going up, and it seems that in the big cities there are many buyers and very little to non sellers. does anyone know anything about it? I'd like to know if other people have expirienced that, or if they got drops as usual. Foo 15:58, 26 May 2006 (CDT)
 * Never mind, it was a fixed bug. May the prices go up still. Foo 05:09, 27 May 2006 (CDT)

Confidence level?
Would it be a good idea to add statistical error in the estimated rate? For example, if killing 1000 certain mob yields 10 gold drops, the rate is estimated to be 10/1000 = 1%, but the statistical error of 10 is the square root of it, i.e. about 3, or 0.3%, so a more accurate measure of the rate is 1&plusmn;0.3%, i.e. 0.7 ~ 1.3%. This wouldn't matter for high statistics, but for something like 1 observed drop out of 100 kills, it becomes 0 ~ 2% instead of just stating "1%", so as to clearly indicate the unreliability of available data. Or is this too nerdy? -- Ledrug 02:58, 28 June 2006 (CDT)


 * Don't understand where the math is coming from. Where do you get the formula that the statistical error of 10 is the square root? --Karlos 03:34, 28 June 2006 (CDT)
 * http://en.wikipedia.org/wiki/Poisson_distribution Square root is actually pretty common in statistics, but I don't think we want to get into too much math&mdash;btw, it's an estimate of the fluctuation, not exact, particularly for small numbers. -- Ledrug 04:18, 28 June 2006 (CDT)
 * Unnecessary imho. --[[Image:Gem-icon-sm.png]] 11:07, 28 June 2006 (CDT)