Forum:Rare,Very Rare, Ultra Rare..Etc

From the RuneScape Wiki, the wiki for all things RuneScape
Jump to: navigation, search
Forums: Yew Grove > Rare,Very Rare, Ultra Rare..Etc
Archive
This page or section is an archive.
Please do not edit the contents of this page.
This thread was archived on 8 April 2010 by Calebchiam.

In many articles where the monster has quite a few drops, what determines in the article what is rare,semi-rare, ultra rare? I think that there should be some sort of a standard. If I am to understand right, all monsters have common drops and a group of rare drops. Take the Ankou for example: Deaths and bloods are common drops, where as Rune javs and uncut gems may be less. I think that all of these prefixes (semi, very, fairly, ultra, extremely, uncommon) should be regulated to make it easier on the reader ‎Easter egg.pngAtlandy 14:26, 17 May 2009 (UTC)

Discussion

Support - I support the principle, but I can see this being difficult to implement. Concensus would need to agree with what Rare, common, uncommon etc. mean and the requirements for them e.g

Label Drop Rate
Common 1 in 25 kills
Very Rare 1 in 1000 kills


Karis Talk to me 18:11, 17 May 2009 (UTC)


Support - I would like to suggest this:

Label Drop Rate
Very common 1 in 5 kills
Common 1 in 10 kills
Semi-common 1 in 20 kills
Uncommon 1 in 50 kills (not sure about this one)
Semi-rare 1 in 75 kills
Rare 1 in 100 kills
Very rare 1 in 500 kills
Extremely rare 1 in 1000 kills

Or something like that. Ancient talisman.png Oil4 Talk 18:28, 17 May 2009 (UTC)


Comment Just to add, there is already a topic on the YG about this, however no discussion has started -Forum:Metrics Karis Talk to me 18:44, 17 May 2009 (UTC)

This has long been something that has bothered me. If we could agree (rofl) on what numbers get what prefix, then we would have arguments over what numbers are actually accurate. I think we need to totally drop all the prefixes in order to maintain a more uniform standard. If people wish to collaborate and gather some data (such as we have on Charms) then the raw numbers themselves could be included in a small subsection on the bottom of pages, as long as we aren't making complete drop logs or anything like that. --Degenret01 04:21, 18 May 2009 (UTC)

I Agree that "less is more". The table above has 8 different classes. I think 1 would suffice (rare)‎Easter egg.pngAtlandy 11:02, 18 May 2009 (UTC)

Support & Comment 10 kills isn't enough to judge if an item is very common, semi-rare etc.

I think we should create a 'drop log' containing 500+ kills, as you in theory could get 5 Draconic visages in a row from a Iron dragon ... FredeTalk 11:12, 18 May 2009 (UTC)

Comment - But how can we do that? We don't want to take up extra space on a page with a drop log. Perhaps assign a few users to get these drop statistics? ~MuzTalk 11:15, 18 May 2009 (UTC)
Reply We could use some of the existing drop logs [1] This one will be very helpful.... But yes, it seems we have to 'assign' a few volunteers to make a drop log. FredeTalk 11:28, 18 May 2009 (UTC)
Reply Drop logs are not the answer. ‎Easter egg.pngAtlandy 13:03, 18 May 2009 (UTC)

Support - When I see rare or common in the drop lists of a monster, I get confused to what it really means. If we have a solid use of these, we can make for sure that the number is what it is. ~MuzTalk 11:15, 18 May 2009 (UTC)

Support & Comment - I support having a metric system like this. However, I think that we should do away with very rare/extremely rare (or at least get rid of one of them) as it is rather difficult to draw the line between the two. For items that are rare to such an extent, a player might be able to get the drop after 400-500 kills while another may get the item after 1000+ kills. I'm fine with the very common/common as those are fairly obvious. C.ChiamTalk 11:17, 18 May 2009 (UTC)

Comment I am against drop logs on the pages, but I think if we looked at two words only, Common and Rare, it would make it a whole lot easier. ‎Easter egg.pngAtlandy 11:24, 18 May 2009 (UTC)

I only believe that drop logs are fun uses and not really accurate research. It kinda contradicts on what I put on my logs however.=S Santa hat.png Powers38 おはようヾ(´・ω・`) 14:03, 18 May 2009 (UTC)

Comment - Would we do this for every single drop on a particular creature's drop list? I have noticed that some drops don't have "common" or "rare" next to them (any reason why?), so we would have to implement this system for those drops as well.  Tien  15:42, 18 May 2009 (UTC)

Comment I would think the idea is that to have common only next to common items such as herbs from chaos druids, and rare on only rare items (Unholy symbols from chaos druids). Other could be left blank assuming the all drop at the same rate‎Easter egg.pngAtlandy 15:54, 18 May 2009 (UTC)

Support - We need a standard for drops. On all of the articles with drop lists, we'd probably have to put a link at the top of the subsection to whatever policy this turns out to be. I propose something like this:

Label Drop rate Explanation
Always 1:1 If the item is received every one kill, it is thus always a drop. You'll probably stop picking them up after a while
Very common 1:2 The way I interperate "very common" is something that you get almost all the time, but not quite every time.
Common 1:4 Something that is common is between being very common and semi-common. 25% is large for a repeatedly-used probability. You're not quite getting it every kill, nor even almost every kill, but you'll get quite a lot of them if you kill the monster for a while.
Semi-common 1:10 This is something that you'll get decent ammounts of when training on a single monster for a while, but only half as often as a common item at best.
Uncommon 1:20 5% is a reasonable probability if you're going to metaphorically roll the non-existance dice of monster drops when training. Training on a monster for long periods of time will get you many of these.
Semi-rare 1:100 If you train on a monster for a period of time, you might get this a few tmes.
Rare 1:500 Something that is rare is something that not quite everyone will get, and is a reasonable feat to get it. If you train on a monster for at least a few hours, you might get something like this once or twice. Thus, it is a rare drop.

Tell someone that you got one as a drop, and they might compliment you.

Very rare 1:2,000 A very rare drop would be something that would take probably multiple sessions of training on the monster, each being a few hours. It would be an accomplishment to get this.

Tell someone that you got one as a drop, and they'll congratulate you.

Extremely rare 1:10,000 Something that is extremely rare is something that will take either large ammounts of luck, or lots of dedication to get. Getting a Dragon chainbody from Dust devils is a 1:16,000 chance, as I heard. It will take you a long time to get one as a drop, and most people won't ever get it altogether.

Tell someone that you got one as a drop, and they'll be in awe.

My spin on things. I really don't think that we should label drops as "extremely rare" if you get one every thousand kills. If that's so, you can get one a day. If you can get them that often, they're not all that rare. Dragon medium helm! Whaddaya know?Chiafriend12Better than rune!Loon is best buttlord 21:33, 18 May 2009 (UTC)

Common While the chart is nice, I really do not think we need 9 levels of how common something is. Can anyone give me a monster where all of these would be used? If I think of Ankous for example Deaths would be common where as Laws would be rare. Do we really need any more than that? ‎Easter egg.pngAtlandy 22:55, 18 May 2009 (UTC)
Support - Chiafriend's idea is great. Statistics.png Lvl 3 skils3 Choice! Talk~ Holiday Signup ~Hiscores 00:23, 19 May 2009 (UTC)
Comment - I think the "very common", "common", and "semi-common" (and possibly "uncommon") could go together.  Tien  00:55, 19 May 2009 (UTC)

Comment - I was wondering if we do it like this


Label Drop Rate
Always Always
Common Somewhat 1 in 3-5 kills
Uncommon Somewhat 1 in 10-20 kills
Rare 1 in 100 kills
Very rare 1 in 500+ kills

I decided to remove extremely rare because...it is extremely rare and would most likely be difficult to obtain one. I also shortened it to 5 so maybe we might not have much difficulty choosing.=S Santa hat.png Powers38 おはようヾ(´・ω・`) 00:39, 19 May 2009 (UTC)

It'd still be nice to know whether an item is "extremely rare" or not. If we leave it out, then people might be eagerly expecting one while they kill the monster, but get frustrated after, say, 5000 kills without that particular item as a drop.  Tien  00:55, 19 May 2009 (UTC)
Which is why I added 500 + kills. Although I see what you mean, it is difficult to draw the line between very rare and extremely rare.Santa hat.png Powers38 おはようヾ(´・ω・`) 01:07, 19 May 2009 (UTC)


Could we have the actually kills:drop ratio as well as the rating? This was a final solution for the money making metric, where the levels of effectiveness didn't quite give precise enough information. As for figuring out what the ratio acutally is, is anyone on the wiki a good enough programmer to be able to lift the numbers from the game engine, and if not, couldn't we ask jagex? --Serenity1137 22:24, 22 May 2009 (UTC)


The idea of common and rare should really just give someone an idea of how hard/easy it will be for someone to abtain that particular item. It will take several tests and probability simulations before you can accuratly determine the rarity of an item. For example I know people who recieved visages in only a few attempts and I recieved a few rare items myself on only a few attemmpts. Therefore I agree with the above table because it gives a general idea.kudos to who created this...Rune crossbow.png Hess36talk Ancient staff.png 20:32, 26 May 2009 (UTC)

Support Chia's I find that one to be the best by far. This needs more regulation, and Chia's looks perfect --Aburnett(Talk) 20:36, 26 May 2009 (UTC)

Comment Look at Iron dragon before I fixed it. There were at least 10 different levels of rarity. If we made it that only 2 modifiers, it would be so much easier (rare and common). —The preceding unsigned comment was added by Atlandy (talk).

But it doesn't give as much detail as some might want, draconic visage is one hell of alot rarer than a rune square shield  —The preceding unsigned comment was added by Serenity1137 (talk).

SupportChia has a grat idea! The best for me so far.-- Defence Waterkunai1 Talk # fluffy bunniesArmadyl symbol.png 20:03, 29 May 2009 (UTC) Comment While Chia's idea on paper looks great please consider the following: There are how many? (400+ monsters) in RS? with such a large breakdown of drops, it would take countless hours to determine the difference between 1 in 10 and 1 in 20. The reasoning behind my original post was to make the list smaller and easier on the average wikian. To split drops into 9 different parts makes it even harder. Every page will have to come with an explanation on the differences between semi-common, common, uncommon and semi-rare. If we kept it to 3 Common, Rare, and very rare it would make it easier on the entire wiki. ‎Easter egg.pngAtlandy 20:43, 29 May 2009 (UTC)

Support Chia's We really should have a standardization. And that is a pretty good one.Joe Click Here for Awesomeness18:10, 1 June 2009 (UTC)

Idea Why don't we put the ratio instead of rare etc.?Joe Click Here for Awesomeness18:11, 1 June 2009 (UTC)

Well ratios aren't always consistent... And on a separate note, we do have to remember kill rate is not the same as drop rate. If a Goblin drops 2 nats 1:100, that would be labeled the same as the Armadyl Boss dropping a hilt 1:100, but I do not know the actual ratios. If this comes through don't let the item influence people. Prayer.png Jedi Talk HS Log Tracker Summoning.png 18:31, 1 June 2009 (UTC)

Comment Look at Lesser demon. There are no less than 7 different modifiers with its drops (common, most common, very uncommon, semi uncommon, semi-rare, rare, and extremely rare) Some drops have nothing at all. ‎Easter egg.pngAtlandy 22:23, 2 June 2009 (UTC)

Support Chia's, as long as it's accurate. White partyhat old.png C Teng talk 11:53, 9 June 2009 (UTC)


Support if applicable Really what would be nice (and I know I say this alot) would be to just lift the information out of the game engine. If you can't do that then it is going to be very hard to get these numbers right. p.s. I think items/per kill is preferable to drops/kill.--Serenity1137 14:26, 10 June 2009 (UTC)

Support if willing to test Sounds like a great idea but... who's gonna be willing to test the drops for every past monster? Like the Ogre, who wants to test that?! --Quest point cape.pngLil Diriz 77 Talk Summoning-icon.png 12:11, 21 June 2009 (UTC)

Support Chia - and testing it would not be a bad idea as well. Chia just came up with a great sysytem. Magic-icon.pngStelercusIlluminated Book of Balance.png 09:37, 23 June 2009 (UTC)

Comment I do not think that more prefixes makes it more clear cut. Also, drops as we know, are different for different players. What I suggest is only 3 prefixes (common, Rare, and very rare) That should cover all the bases ‎Keg of beer.pngAtlandyBeer.png 13:19, 23 June 2009 (UTC)

Comment - Err...It's not possible to get the droprates. And if you wanted to statistically estimate them, no-one here has a clue how to. So I don't see how this would work? JalYt-Xil-Vimescarrot 13:59, 23 June 2009 (UTC)

Reply That was my point initially. Why don't we just call things common, rare and the occasional very rare ‎Keg of beer.pngAtlandyBeer.png 14:45, 23 June 2009 (UTC)
Drops are different for different players?!? --Serenity1137 15:37, 23 June 2009 (UTC), p.s. I would say there are more than three levels that we can tell apart, p.p.s. It is good to have numbers even if we can't do it accurately, otherewise we get caught up in me thinking rare is 1 in 10, very rare is 1 in 100 and you thinking rare is 1 in 100, very rare is 1 in 1000.
Make a drop log that anyone can add to. After every 100 - 500 (depends on how fast one can kill the monster) drops, we calculate the drop rates so far and put them in the articles. Then, after another 100 - 500 kills, we calculate everything again and update the article, making it more precise. Seems possible to me (although a bit annoying for things like the Queen or the Beast). Ancient talisman.png Oil4 Talk 18:36, 23 June 2009 (UTC)
Reply Drop logs are not the answer. Drops are random. You could receive a visage on your first iron dragon kill and I could kill 5000 and not receive one. We have already determined that drop logs for each monster would not work ‎Keg of beer.pngAtlandyBeer.png 19:13, 23 June 2009 (UTC)
I like the idea of cumulative drop logs, though they are random, the odds of getting an unlikely thing (i.e. draconic visage)

on the first drop are very low. I thought drop logs were rejected on the idea that one person couldn't do enough kills to make it a well sized sample. Somethings would end up with 1000s of things on the log. I would support this if drop logs were semi protected (the thing were only oldish accounts can edit) --Serenity1137 22:10, 23 June 2009 (UTC)

Even well-meaning additions to a group drop log would be biased. "Wow! I just got a visage! Wait...maybe I should add these 100 steel dragons' drops to the drop log!" Conveniently forgetting the six thousand steel dragons killed a month earlier. There is no practical way to reliably find out drop rates, so stop trying to find one. JalYt-Xil-Vimescarrot 10:15, 24 June 2009 (UTC)

Of course there is a practical way, simply ask people not to start drop logs because of a particular drop... In other words say that you must decide whether or not your going to add a drop to the log before you finish killing the monster. --Serenity1137 15:33, 24 June 2009 (UTC)

Support - 3rd OptionShort, Sweet and simple. --Ringmaster hat.png Celtic-Creations Talk to me Dragon hatchet.png 09:30, 23 August 2009 (UTC)

System Trail

Essentially if there is enough support for this, I will make a template that turns a drop log table into stats, and start a drop log for one monster , if after a few weeks this has been shown to work, we integrate it with the page relevant page and start drop logs for all monsters. The table would like something like

Total Kills Bones Diamond- uncut Steel bars ...
18 18 1 2

The table would have one column or row for each item that is known to be dropped by the monster, and the log would be counted as accurate for a particular item when more than 5 had been dropped (less for very rare items, more for items dropped in large quantities).

Either it would work or it wouldn't, if it doesn't seem to be an accurate representation, no harm no foul, we just delete it and move on.

--Serenity1137 15:46, 24 June 2009 (UTC) p.s. this would allow us to calculate average profit from killing a certain npc


Support as nominator --Serenity1137 15:46, 24 June 2009 (UTC)

Neutral, leaning towards oppose - Vimescarrot convinced me that logs are really not the answer. By the way, how can we make these for hundreds, even thousands of different enemies? Ancient talisman.png Oil4 Talk 18:58, 24 June 2009 (UTC)

Easy, simply create I'll simply create a template that means all we need is a sub page on each monsters page. Contributing to logs would just be something to do if you felt bored and wanted to be vaguely productive. Also, I've said this before but I'll say it again: Vimes' argument is that unlikely things happen, they do - but they are unlikely --Serenity1137 00:07, 25 June 2009 (UTC)

Slight Oppose - Per Vimes, drop logs are going to be statistically biased towards rarer items unless very carefully done, which is going to be next to impossible. --Quarenon  Talk 00:49, 25 June 2009 (UTC)

This won't be a problem if we make very clear rules for when you can add to the drop log, I think these rules would be followed were it semi-protected. It would be made very clear that you must decide whether a drop was going to be entered before getting it, and that any contributions that didn't fit in with that would be not only not useful but actively destructive. If a good drop alerts you to the idea of drop logs, then you should start doing the drops logs from the next kill... --Serenity1137 06:36, 25 June 2009 (UTC)
Eh... No. It just seems to me like people would only add to certain logs. For example, how many people would go and kill some high-level, high-hitting, low-xp monsters just for the sake of a log? Ancient talisman.png Oil4 Talk 18:48, 25 June 2009 (UTC)
Drop logs can create statistically valuable data if done in a large enough scale. At a minimum just thinking off the top of my head as an engineer of 16 years. You would need at a minimum of 100 kills per monster with ideally 10 people providing the data for 100 kills each. This would provide percentages based on 1,000 kills. At that amount of data the probabilities would be in general fairly precise. This would not cover all monsters because some monster have such a large amount of potential drops. Anyway If people would be interested in undertaking that amount of data I would certainly be willing to find a few various calculations that would determine within what percent error drops would be consider relevant. Will this cover the percentage of very rare drops such as dragon chain bodies or draconic visage's? No but it should cover most all other levels of drops. Again this is just a off the top of my head estimate. Strength-icon.png Hammer2092 Time |KOTA Bandos godsword.png 20:04, 25 June 2009 (UTC)
I do think that there are quite some people who are bored in RS and who might want to do this. For example, me. Ancient talisman.png Oil4 Talk 21:17, 25 June 2009 (UTC)
If we can get 10 people to say they would have an interest in doing this I would think starting a project page specific to logging monster drops and the appropriate charts would be easy enough to do. Ultimately this project page would take months if not years to complete but would make for a really impressive bit of info after all is said and done. Strength-icon.png Hammer2092 Time |KOTA Bandos godsword.png 21:26, 25 June 2009 (UTC)

Changed to neutral, leaning towards support. I just hope it's possible. Ancient talisman.png Oil4 Talk 06:58, 26 June 2009 (UTC)


I don't think we need anywhere near as much as a thousand before it starts to be useful, statistically speaking I would say 100 kills is more than enough for the common drops, and gives a rough idea of semi-common drops (up to one in 20, you only really need an expected value of 5) as such I disagree that it will be years before we have useful data in some areas (like flesh crawlers will could esily end up with a 100 kills inputted every few days). Obviously the very rare drops wouldn't be accurate for a long time, but without using this or a similar system we would be stuck with peoples random guesses for ever. As for the maths I can handle that, and the moment I've got the go ahead I'll start work on making a template to convert easily input raw data into stats in a convienient form. And finally, I suspect monsters with high hit points/defence/attack and bad drops won't be done much, but they are less important as less people will want to kill them. --Serenity1137 14:26, 26 June 2009 (UTC) (p.s. sorry if any of this post makes no sense, I kept adding bits in)

Well, count me in as a volunteer if you need one. I think that we can get enough volunteers to update the articles often. And because everyone can edit the logs, we don't need to rely on volunteers only. Ancient talisman.png Oil4 Talk 19:04, 26 June 2009 (UTC)

System Trial

I know there wasn't consensus to do this, but I don't see how trying can do any harm. I have produced a template to turn drop logs into a nice drop table, here is an example of how it works

This input:

{{Log_Table
|Bones|1||||||||
|50
|Gold|150
|Steel bar|25
|Diamond|2
|Clue Scroll|1
}}


Produces:

{{Log_Table
|Bones|1|||||||| 
|50
|Gold|150
|Steel bar|25
|Diamond|2
|Clue Scroll|1
}}


I've started a drop log [[Goblin/F2P_Drop_Log|here]] for goblins (it is currently deleterated, I'm trying to get it undeleterated), starting with something with a small range of drops thats easy to kill. Feel free to start adding information.. (This is a log for f2p worlds.)


--Serenity1137 17:46, 29 June 2009 (UTC)

Comment / Question - I like it. If it is decided that this will be used, I'll be happy to contribute. Just a little question: the input is 'drop|amount', right? How does this work for different amounts? Like a monster that drops 5gp at once but sometimes drops 20000gp at once. How will you add that? Ancient talisman.png Oil4 Talk 18:50, 29 June 2009 (UTC)

You add 20,000 to gold when it drops 20,000 and you add 5 when it drops 5. --Serenity1137 20:49, 29 June 2009 (UTC)
But doesn't the template count it as 20,000 single coin drops then? Ancient talisman.png Oil4 Talk 21:14, 29 June 2009 (UTC)
Well you've added 20,000 coins to the total drop, and one to the total kills. You aren't recording separate drops, you're recording total drops and total kills --Serenity1137 22:51, 29 June 2009 (UTC) p.s. check out the one I've started, im fairly sure stuff like coins per kill is accurate to atleast 1 sf.

Would it help if I wrote a decent guide on how to use the drop logger? --Serenity1137 07:53, 30 June 2009 (UTC)

It would help a lot, I think. The part with the coins was a bit confusing to me. Ancient talisman.png Oil4 Talk 19:04, 30 June 2009 (UTC)
Its [[Template:Log Table/Guidelines|here]], note that the actual logs are currently deleted, and until the go ahead is given we can't recreate them.--Serenity1137 19:17, 30 June 2009 (UTC)
Your link was red, so I fixed it. Anyways, I hope that we get a green light. Would be a waste of the template if not. Ancient talisman.png Oil4 Talk 20:54, 30 June 2009 (UTC)

We totally should, consensus isnt permanent, and there is no harm in a trial of the system --Serenity1137 21:29, 30 June 2009 (UTC)

righty, we are back in business. Is the guidelines page clear enough? --Serenity1137 07:35, 1 July 2009 (UTC)


Ok, I think that this drop log system will take several thousand drops before we know the drop percentages of the rarer items. However, the average drop worth has been static since around 50 kills. I propose we keep adding data till all the things have settled down --Serenity1137 08:44, 1 July 2009 (UTC)
Okay. Why doesn't anyone post here? Ancient talisman.png Oil4 Talk 20:16, 1 July 2009 (UTC)

Looks alright; any way you could only show rows that are needed? I would contribute to a log, but I'm concentrating on non-combat skills at the moment. Next time I go slaying or something I'll log my drops. Quest.png Gaz Lloyd 7:^]Events!99s 21:36, 1 July 2009 (UTC)

Not that I can tell gaz, unless parser functions have some subtlety that is unknown to me --Serenity1137 00:07, 2 July 2009 (UTC) p.s. the current log is a test run, and until there is consensus to change policy on drop logs, it is probably good form to just have the one (which is currently level two goblins in a f2p world because I could kill those quickly and they are quite densely packed in lumby)
Short of hard-coding the table each time (we could use #vardefine: to set the kill count, but it would still be a nasty job with the large amounts of drops and GEPrices), you could use a lot of #if:'s (but the code would be rather large), or maybe some way of specifying how many rows you need (I'm not sure how to do what, maybe #switch:?).
If and when I make a drop log I'll probably write it down by hand (or maybe save in Excel or something), then probably test out the log table with it in my userspace until consensus is reached. Quest.png Gaz Lloyd 7:^]Events!99s 11:26, 2 July 2009 (UTC)
The problem lies with putting | inside a parser function without it interfering, you can't use <nowiki> because then it won't be any good for use in the table. As for the code being large, you might be able to get away with using subtemplates like {{Log Table Row}} --Serenity1137 11:45, 2 July 2009 (UTC)
Template:! is used to put pipes in parsers (for use in constructing tables etc). Quest.png Gaz Lloyd 7:^]Events!99s 12:37, 2 July 2009 (UTC)
Fixed it, Smile--Serenity1137 14:21, 2 July 2009 (UTC)
And by fixed it, I mean broke it Grrrr --Serenity1137 14:44, 2 July 2009 (UTC)
I got it, looks great now. Quest.png Gaz Lloyd 7:^]Events!99s 14:47, 2 July 2009 (UTC)
How very perculiar, I'm not entirely sure what I was doing wrong, maybe I was doing it for the wrong ones. Nicely done though, and indeed it be awesome now. Thanks --Serenity1137 15:06, 2 July 2009 (UTC)

Should I knock the max items up to 200 (from 50) now that it won't take up a page? --Serenity1137 15:09, 2 July 2009 (UTC)

Charm droprate

I think it would be valuable to add some extra data on droprates of charms. Something else than just the word very common. Im thinking of a template that looks something like this.


It would have to be added to all monsters and it should be used as an indication not exact numbers.--Ruderion 14:17, 11 July 2009 (UTC)

Done. Template:Charm table is available to do this now. Full documentation on the page. The simplest way to generate the above table is as follows:
Input   Result
{{Charm table
|kills=20
|gold=6
|green=2
|crimson=3
|blue=1
}}
 
Charm drop percentages
No Charm Gold charms Green charms Crimson charms Blue charms
40% 30% 10% 15% 5%

Based on a sample of 20 kills. 1 charm is dropped at once.
Note that due to the small number of kills, the percentages may be slightly inaccurate.
Please add to the log!

Vyrewatch
{{Charm table
|kills=397
|gold=31
|green=17
|crimson=17
|blue=19
}}
 
Charm drop percentages
No Charm Gold charms Green charms Crimson charms Blue charms
79% 8% 4% 4% 5%

Based on a sample of 397 kills. 1 charm is dropped at once.
Please add to the log!

Waterfiends
{{Charm table
|kills=154
|gold=22
|green=6
|crimson=118
|blue=5
}}
 
Charm drop percentages
No Charm Gold charms Green charms Crimson charms Blue charms
2% 14% 4% 77% 3%

Based on a sample of 154 kills. 1 charm is dropped at once.
Note that due to the small number of kills, the percentages may be slightly inaccurate.
Please add to the log!


Any further improvements needed? Quest.png Gaz Lloyd 7:^]Events!99s 18:16, 28 July 2009 (UTC)
Looks great, the only I see adding is center back. Should we add a Category:Missing_Charm_drop_rate for incomplete charm drop data? - TehKittyCatTalk Wikian-Book 14:42, 29 July 2009 (UTC)

Numbers have been centered - I added a few more examples (from my investigations) above. A category could be useful, but we would need to add it to every monster's page, kind of an arduous task.

I'd like to draw attention to here and especially here, where I have been testing out the Log Table template - so far logged 47 goraks, 154 waterfiends, 397 vyrewatch cremations and an slightly more unorthodox use on 150 magpie pouches (I have a dust devil task now so shall be logging them).

As Serenity said on my talk page, these tables (expecially the vyrewatch one due to the higher number of cremations) give a good idea of the general tables in the early stages. It can still be innaccurate (e.g. the dragon spear from vyrewatch came in the most recent lot of 50, so if I started then it would seem a lot more common, etc), but once the killcount reaches very high numbers (in excess of 10,000) then the small anomalies will be insignificant, and general rate will be pretty much correct. After about 100 or so kills the average value of a kill and the more common drops (like coins to 1 sig. fig.) should stablise.

If we overturn the no drop logs point, I think (or rather agree with Serenity) that subpages should be created for each monster with a log table on (one for F2P, one for P2P, one for P2P with a ring of wealth, maybe). These pages should be kept an eye on, one unnoticed vandal and BAM, there goes your accuracy. Semiprotection could be used, but I'm not 100% sure on that yet (popular monsters (e.g. things like the KBD) will probably need it, though). Quest.png Gaz Lloyd 7:^]Events!99s 17:34, 29 July 2009 (UTC)

There is no way to verify any claim anyone makes. Everyone who contributed to this would have to be on some kind of trust list. (Fully protecting the pages would be enough - the "trust list" would be all admins.) Plus, systematic biases would make it inaccurate. ("Woot! I got a D spear! I'll upload this log to the Wiki!" Conveniently forgetting the six thousand previous logs with which nothing has been done.) For common drops maybe - for rare drops, unless only statisticians are allowed to contribute - no. JalYt-Xil-Vimescarrot 18:27, 29 July 2009 (UTC)
Mm, that is a point. The benefit of having a decent drop chance listed (derived from a log) outweighs the possible inaccuracies of the log, in my opinion. We could work around it - full protection would work, but I feel we have plenty of reliable non-admin users who would want to contribute; we could also ask users to state that they are doing a log and of how many monsters (give or take a few), before doing it, on the talk page (or for already-completed logs users should post a link to them on the relevant talk page before adding them).
Might be a good idea to open a project for interested users to create logs for mainspace use. Quest.png Gaz Lloyd 7:^]Events!99s 19:39, 29 July 2009 (UTC)
Guaranteed inaccuracies, for rare items. JalYt-Xil-Vimescarrot 18:27, 30 July 2009 (UTC)

Here's my two cents worth:

Display Pros Cons
What we have now (users entering "rare", "common", etc.) Easy to edit No consistency as to what those words mean, absolutely no idea if "common" means 40% or just "more common than some other drops".
Not displaying any drop information until we have logs of 100,000 kills per monster Accurate Won't have data on most monsters for years
Display exact percentages, using Gaz' template Easy to read Not really accurate, if you kill 100 monsters and get 3 gold charms, you really can't say the drop rate is exactly 3%.
Display range of percentages (see following sections) Statistically valid A little more complicated to read

Cscooper 21:27, 23 August 2009 (UTC)

Displaying confidence range

I've expanded Gaz's original template so that it displays a statistically valid range of values (see Template:Charm_table_stats), based on the number of kills reported:

Charm drop percentages
No charms Gold charm.png Green charm.png Crimson charm.png Blue charm.png

75–82%

6–10%

3–6%

3–6%

3–7%

Represents 90% confidence range, based on a sample of 397 kills.
1 charm is dropped at a time.

This kind of implements the idea that Vimescarrot proposed that "only statisticians be allowed to contribute". The statistician information has been congealed into a macro, and the information presented in a relatively simple way.

We could even expand this idea to display "Rare", "Common", etc. (assuming we ever come to agreement on what those terms should mean) in addition to the percentage ranges.

Cscooper 21:27, 23 August 2009 (UTC)

Multiple drop logs

I then created an idea that lets multiple users upload drop logs, and it combines those drop logs into a single output, it looks like this: <pre>{{charm_drop_start}} {{charm_drop_log |user = Username of the player reporting the drops |date = Date on which the drop log was finished and/or reported |kills = Number of kills contained in this drop log |gold = Number of gold charms acquired |green = Number of green charms acquired |blue = Number of blue charms acquired |crimson = Number of crimson charms acquired }} (repeat charm_drop_log as many times as there are different drop logs) {{charm_drop_display}} </pre> So you could have Wiki code that looked like this: <pre>{{charm_drop_start}} {{charm_drop_log|user=Killer|date=03Aug2009|kills=100|gold=10}} {{charm_drop_log|user=Gaz|date=05Aug2009|kills=150|gold=20}} {{charm_drop_log|user=Csc|date=07Aug2009|kills=500|gold=47}} {{charm_drop_log|user=Noob|date=03Aug2009|kills=5|gold=5}} {{charm_drop_display}} </pre> and gave this output: {{charm_drop_start}}{{charm_drop_log|user=Killer|date=03Aug2009|kills=100|gold=10}}{{charm_drop_log|user=Gaz|date=05Aug2009|kills=150|gold=20}}{{charm_drop_log|user=Csc|date=07Aug2009|kills=500|gold=47}}{{charm_drop_log|user=Noob|date=03Aug2009|kills=5|gold=5}}{{charm_drop_display}}

Notice that the noob who was all excited, he killed 5 monsters and got 5 gold charms, is allowed to contribute, but his information is included with the other drop logs and thus is evened out.

What I'm thinking, is that this way we have visibility into each user who uploads drop log information - and if we see one that looks way out of line, we can delete just that one without losing all of the log information.

Cscooper 21:27, 23 August 2009 (UTC)

Summary

This page is getting wicked long, what do people think about moving the discussions into these categories?

Cscooper 19:50, 10 August 2009 (UTC)


Terms

Option 1

Label Drop Rate Percent
Very common 1 in 5 kills 20% to 100%
Common 1 in 10 kills 10% to 19.9%
Semi-common 1 in 20 kills 5% to 9.9%
Uncommon 1 in 50 kills (not sure about this one) 2% to 4.9%
Semi-rare 1 in 75 kills 1.3% to 1.9%
Rare 1 in 100 kills 1.0% to 1.29%
Very rare 1 in 500 kills 0.2% to 0.9%
Extremely rare 1 in 1000 kills 0.1% to 0.19%

Option 2

Label Drop rate Explanation
Always 1:1 If the item is received every one kill, it is thus always a drop. You'll probably stop picking them up after a while
Very common 1:2 The way I interpret "very common" is something that you get almost all the time, but not quite every time.
Common 1:4 Something that is common is between being very common and semi-common. 25% is large for a repeatedly-used probability. You're not quite getting it every kill, nor even almost every kill, but you'll get quite a lot of them if you kill the monster for a while.
Semi-common 1:10 This is something that you'll get decent ammounts of when training on a single monster for a while, but only half as often as a common item at best.
Uncommon 1:20 5% is a reasonable probability if you're going to metaphorically roll the nonexistence dice of monster drops when training. Training on a monster for long periods of time will get you many of these.
Semi-rare 1:100 If you train on a monster for a period of time, you might get this a few times.
Rare 1:500 Something that is rare is something that not quite everyone will get, and is a reasonable feat to get it. If you train on a monster for at least a few hours, you might get something like this once or twice. Thus, it is a rare drop.

Tell someone that you got one as a drop, and they might compliment you.

Very rare 1:2,000 A very rare drop would be something that would take probably multiple sessions of training on the monster, each being a few hours. It would be an accomplishment to get this.

Tell someone that you got one as a drop, and they'll congratulate you.

Extremely rare 1:10,000 Something that is extremely rare is something that will take either large ammounts of luck, or lots of dedication to get. Getting a Dragon chainbody from Dust devils is a 1:16,000 chance, as I heard. It will take you a long time to get one as a drop, and most people won't ever get it altogether.

Tell someone that you got one as a drop, and they'll be in awe.

Option 3

Label Drop Rate
Always Always
Common Somewhat 1 in 3-5 kills
Uncommon Somewhat 1 in 10-20 kills
Rare 1 in 100 kills
Very rare 1 in 500+ kills

Presentation

(What do we display on the final Wiki page - "1 in 50" or "extremely rare"?)

Support Option 2' - I would say "extremely rare" with each of the terms linking to a page explaining the Wiki's drop nomenclature system. - Pith helmetDuckmartins 19:39, 27 August 2009 (UTC)

Which alas requires a decision on what that drop nomenclature system is :-)

Maybe it's time to pick one and say "this is the standard" and see if anybody complains.

Cscooper 12:42, 28 August 2009 (UTC)

Support "Option 2" - It seems more realistic. Zaros symbol.pngChaos Monk Talk SignCoins 250.png 23:41, September 25, 2009 (UTC)

Support Option 2 - As long as it's written in third person. — Enigma 06:22, October 18, 2009 (UTC)

Comment - Is there really any drops which is 1:10,000? --Nup(T) 14:29, October 27, 2009 (UTC)

Dragon chainbody from dust devils is supposedly 1/16,000. Things like dragon spears and left half of the dragon square shield, from low level monsters (hill giants, skeletons, etc) I would guess also fall into this region, too. Quest.png Gaz Lloyd 7:^]Events!99s 17:21, October 27, 2009 (UTC)

Template

Good Idea about the template, I could see this implemented on Wikia, although, would there be a way that in the template, it put everything in categories, if we could categorize every dropable items, or atleast the most common ones, such as charms?

--Quest point cape.pngTalk Newbie856 edit count Nomad guideMusic icon.png 18:36, September 25, 2009 (UTC)

Support - Per above. I think categorizing them would be very beneficial to the wiki. Zaros symbol.pngChaos Monk Talk SignCoins 250.png 23:41, September 25, 2009 (UTC)

Support - Categorizing would be great for the wiki, and would organize things very well, which would generally make it easier to redo the rarity on the articles once the time comes.--Cheers, Off-hand ascension crossbow.pngYodaAscension crossbow.png 01:29, September 26, 2009 (UTC)

Option 2 seems to me like a better option. I think it's easier to understand. I also think that a charm drop chart would be great for people looking for charm drops. Just my 2 cents. Ghrim's book.png WildWings Dragon plateskirt.png}} 04:20, October 25, 2009 (UTC)

Support - Yea great idea, Option 2 Is (In my opinion) The best it would just take time, adding this in a page, and research. Supawilko 11:12, October 27, 2009 (UTC)

Support 2 - Option 2 seems best. Ancient talisman.png Oil4 Talk 21:02, October 27, 2009 (UTC)

Option 2, but minor change - I like option 2 except that it is rare to see something get dropped 1:2 or even 1:4 times. I think this number should be bumped up a bit. I think very common could even be more like 1:5 with things bumped up appropriately. What do people think? Air rune.png Tollerach hates SoF Fire rune.png 04:30, October 28, 2009 (UTC)

I think cockroach soldiers drop mithril ores at about 1:4. Ancient talisman.png Oil4 Talk 16:04, October 28, 2009 (UTC)
The point of me opening this discussion was to streamline and simplify the drop system. Option 2 has way too many options and will do nothing but cause confusion to the new wikian. Keeping it simple is the way to go ‎20px‎AtlandyBeer.png 18:30, November 1, 2009 (UTC)

Support modified version of 3 - I think 3, if modified a little (making it clearer when something is rare or not etc), would be better, because it is simple. Ancient talisman.png Oil4 Talk 20:14, November 1, 2009 (UTC)

Support modified version 2 - I think 3 would need to be expanded upon greatly. For 2, I think just merging some of the categories works well (common, semi-common, and uncommon, maybe?). Too many is confusing and a big hassle, but if there's two few, why bother with it anyway? 20px‎ Kudos 2 U Talk! Edit count! Contribs! 22:37, November 3, 2009 (UTC)

Support modified version 2 - I think that 1:5 is an appropriate measurement for very common, and maybe 1:10 for common. all the others are good, except that very rare might be bumped up slightly (such as instead of 1:500 it is 1:750 or possibly even higher)--Cheers, Off-hand ascension crossbow.pngYodaAscension crossbow.png 01:02, November 23, 2009 (UTC)

Closed - Further discussion being continued here. C.ChiamTalk 09:28, April 8, 2010 (UTC)