Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 50

Thread: Randomness of loot tables.

  1. #21
    Quote Originally Posted by Sromp View Post
    "Being factual" and "asking for a factual debate" seem very different, then.
    I'm offering suggestions.
    You're mocking the guy.

    Long live the difference!
    I said give reasons and proof, he retorts with namecalling. If that's the way you start a proper debate, then I guess there's no harm done with a little mocking of the superstitious. And as with so many other snake oil salesmen, he hasn't delivered yet. Maybe he will. Had he had a huge spreadsheet detailing the last 3 months of him killing a round of bosses with his friends on, showing a deterioration in drop% chances that fall outside the relatively random spectrum that should be expected, why I would've stood right there beside him saying "seems something's not right here". And I would in fact be very interested in seeing such a thing!
    But he didn't, and the very notion of being asked for it was seen by him as some great attempt at trolling and being an ass or whatever. Meh.
    Last edited by Mastablasta; Apr 24th, 2012 at 18:13:23.
    Thor Mastablasta Hammersmith - Level 220, AI 30, LE 70 Clan Atrox Nano Technician - Setup
    The Red Brotherhood

    I'm a Nano-Technician, don't ever expect me to fight unbuffed, alone or fair.

    Means: about f'ing time :P
    Satenia: heresy <3
    Znore: Mastablasta <3
    Kinkstaah: I have agro from many mobs ;(
    Madarab: we are aoe class, we are supose to use pistols
    Marxgorm: the NT toolset does not fit into my raiding tactics

  2. #22
    Tone down the personal stuff and talk about the actual statistics, if you please.
    Coordinator Anarrina
    Public Forums Moderation Team Lead
    Community Relations Department

    Advisors of Rubi-Ka

    Forum Social Guidelines // Social Events with ARK // Rubi-Ka Marriage Registry // ARK Tours // Join ARK!

    A reader lives a thousand lives before he dies. The man who never reads lives only one. -- GRRM

  3. #23
    Quote Originally Posted by Mastablasta View Post
    I said give reasons and proof, he retorts with namecalling. If that's the way you start a proper debate, then I guess there's no harm done with a little mocking of the superstitious. And as with so many other snake oil salesmen, he hasn't delivered yet. Maybe he will. Had he had a huge spreadsheet detailing the last 3 months of him killing a round of bosses with his friends on, showing a deterioration in drop% chances that fall outside the relatively random spectrum that should be expected, why I would've stood right there beside him saying "seems something's not right here". And I would in fact be very interested in seeing such a thing!
    But he didn't, and the very notion of being asked for it was seen by him as some great attempt at trolling and being an ass or whatever. Meh.
    To try to please you I will go into that factual thing you suddenly seem to like so much. You started by ignoring what I said in my first post and went straight to the amusing part of superstitions and earsay. You started by moking the guy and are now trying to sound reasonable and rigorous with your demand of facts.

    But since there is nothing more important in this world then to satisfy your need for factual accuracy I will try my best to please you just this once. In at least the last 3 years I have noticed the occurence of an increased drop rate of some rare dyna loot after a server restart in 3 SL playfields, on at least 10 different locations. I am sure of at least 13 but more likely more then 15 events in that an item (or a set of items of the same type) dropped in significantly higher quantities then before. The items range from rare (something that I never saw drop except after a server reset) to common loot (stuff that drops at least 3 or 5 times in a week. In one location it happened 3 times and in another couple of others it happened twice with most of the locations with one event. I can not unfortunatly quantify the exact amounts but I think that saying that in the first day or 2 after a server reset the amount of items dropped was equal or higher then the amount dropped during the following month or even longer. I can also clarify that this did not happen to all the bosses I farmed and that I did not farm all bosses in SL.

    You will ofc with rigorous mathematical precision and irrefutable logic prove that this is not only completly within the limits of complete randomness (I suggest quoting Mandelbrot, fractals are very cute and impress the ladies), but also that I am a total and complete liar because I didnt give the name, age, adress, social security and ID card numbers of the deceased mobs as well as the exact time of their deaths and the name of at least 3 witnesses and didnt present documental evidence of their demise.

    Well I have to be frank with you, as much as it pains me. I DONT CARE FOR WHAT YOU THINK OR WHAT YOUR OPINION IS ON THE SUBJECT.

    And now for the really painful part. YOUR OPINION IS OF NO USE TO THE DEVELOPERS. Contrary to what you might think the developers will be able to evaluate the situation much better if you (or any other poster for that matter) are able to present your OWN experience on the subject.

    So, how many boss mobs did you kill lately on SL? Any phats?

    PS: As a consolation prize I present you with the "Grand Honor" of being the only person on my ignore list. I know you are good old chap and will do me the exact same favour. Have fun.



    Or not. I dont really care.

  4. #24
    Quote Originally Posted by Gimpynoob View Post
    But since there is nothing more important in this world then to satisfy your need for factual accuracy I will try my best to please you just this once. In at least the last 3 years I have noticed the occurence of an increased drop rate of some rare dyna loot after a server restart in 3 SL playfields, on at least 10 different locations. I am sure of at least 13 but more likely more then 15 events in that an item (or a set of items of the same type) dropped in significantly higher quantities then before. The items range from rare (something that I never saw drop except after a server reset) to common loot (stuff that drops at least 3 or 5 times in a week. In one location it happened 3 times and in another couple of others it happened twice with most of the locations with one event. I can not unfortunatly quantify the exact amounts but I think that saying that in the first day or 2 after a server reset the amount of items dropped was equal or higher then the amount dropped during the following month or even longer. I can also clarify that this did not happen to all the bosses I farmed and that I did not farm all bosses in SL.
    Information this vague is essentially worthless.
    You will ofc with rigorous mathematical precision and irrefutable logic prove that this is not only completly within the limits of complete randomness (I suggest quoting Mandelbrot, fractals are very cute and impress the ladies), but also that I am a total and complete liar because I didnt give the name, age, adress, social security and ID card numbers of the deceased mobs as well as the exact time of their deaths and the name of at least 3 witnesses and didnt present documental evidence of their demise.
    How about you just provide some relevant data? You are making some rather big claims (loot tables are universally flawed or broken) with nothing to back them up and then getting hostile when people question the validity of your claims.

    Even something simple like a list of the drops from 100 kills of the same creature (with time and date stamps) would be a start.
    Well I have to be frank with you, as much as it pains me. I DONT CARE FOR WHAT YOU THINK OR WHAT YOUR OPINION IS ON THE SUBJECT.

    And now for the really painful part. YOUR OPINION IS OF NO USE TO THE DEVELOPERS. Contrary to what you might think the developers will be able to evaluate the situation much better if you (or any other poster for that matter) are able to present your OWN experience on the subject.
    Well, actually, seeing as how he's a professional, his opinion is apparently of some use to the devs. And your attitude is a bit unwarranted considering that you have yet to provide anything that would help the devs "evaluate the situation" (or that would even indicate that there is a "situation" to begin with).
    I am wiser than any god or scientist, for I have squared the circle and cubed Earth's sphere, thus I have created 4 simultaneous separate 24 hour days within a 4-corner (as in a 4-corner classroom) rotation of Earth. See for yourself the absolute proof.

  5. #25
    Quote Originally Posted by drainbamage View Post
    Information this vague is essentially worthless.

    How about you just provide some relevant data? You are making some rather big claims (loot tables are universally flawed or broken) with nothing to back them up and then getting hostile when people question the validity of your claims.

    Even something simple like a list of the drops from 100 kills of the same creature (with time and date stamps) would be a start.

    Well, actually, seeing as how he's a professional, his opinion is apparently of some use to the devs. And your attitude is a bit unwarranted considering that you have yet to provide anything that would help the devs "evaluate the situation" (or that would even indicate that there is a "situation" to begin with).
    Yes the information is vague. I dont have to give you, him or any other poster on these forums any more information. You are not going to check the code therefore you dont need any more information. However you can contribute with some information yourself. Did you notice the same pattern? Did you notice the same pattern at a different time? Did you notice any other repeatable patterns? Did you even kill any mobs?

    And considering that he is a professional I am sure that the developers will take his opinion in consideration... in matters that concern the rebalance of the profession.

    If my attitude is unwarranted then what would you call his attitude. Or yours for that matter? What have you contributed to the subject? He makes fun of my post, you demand more information. Why? Will you solve the problem if I tell you exactly which mobs I killed and when? Your opinion is of no use, your information, even vague could be. If enough people posted claiming to have detected the same problem, or a different problem or no problem at all then that could be enough to get the attention of the devs and maybe get them to investigate the matter some time in the future if necessary. If I post complete information on one mob or 2 then all that will bring is a few more campers to the mobs that drop valuable stuff.

  6. #26
    Quote Originally Posted by Gimpynoob View Post
    Yes the information is vague. I dont have to give you, him or any other poster on these forums any more information. You are not going to check the code therefore you dont need any more information. However you can contribute with some information yourself. Did you notice the same pattern? Did you notice the same pattern at a different time? Did you notice any other repeatable patterns? Did you even kill any mobs?
    What same pattern? You haven't provided anything other than a vague "hey I think loot is messed up sometimes things drop more often than they do other times (I think)!"

    But you're right, you don't have to give us information, and in turn we don't have to take you seriously.
    And considering that he is a professional I am sure that the developers will take his opinion in consideration... in matters that concern the rebalance of the profession.

    If my attitude is unwarranted then what would you call his attitude. Or yours for that matter? What have you contributed to the subject? He makes fun of my post, you demand more information. Why? Will you solve the problem if I tell you exactly which mobs I killed and when? Your opinion is of no use, your information, even vague could be. If enough people posted claiming to have detected the same problem, or a different problem or no problem at all then that could be enough to get the attention of the devs and maybe get them to investigate the matter some time in the future if necessary. If I post complete information on one mob or 2 then all that will bring is a few more campers to the mobs that drop valuable stuff.
    My attitude is entirely warranted. You are making a claim, and you are providing nothing to back it up with. You are doing this on a forum where anyone can reply. And when people (quite understandably) express doubt and ask for some real evidence, you get defensive (or outright juvenile, in the case of your last reply to Mastablasta).

    Frankly, I doubt you have anything at all other than anecdotes, which are likely tainted with confirmation bias anyways. If you did you have addressed the devs directly instead of looking for support from players.

    So tell us why the devs should care at all about one player claiming the entire loot system is broken when he can't even provide a solid example?
    I am wiser than any god or scientist, for I have squared the circle and cubed Earth's sphere, thus I have created 4 simultaneous separate 24 hour days within a 4-corner (as in a 4-corner classroom) rotation of Earth. See for yourself the absolute proof.

  7. #27
    The next person in this thread to make a comment about someone else's relevancy, attitude, personality, or whether or not you like his aunt's dog gets a 24 hour vacation to cool off. As does everyone who follows up on it.

    I told you guys to cool it, now cool it. Discuss the problem, not each others attitude problems. At least, don't discuss each others attitude problems in this thread.
    Coordinator Anarrina
    Public Forums Moderation Team Lead
    Community Relations Department

    Advisors of Rubi-Ka

    Forum Social Guidelines // Social Events with ARK // Rubi-Ka Marriage Registry // ARK Tours // Join ARK!

    A reader lives a thousand lives before he dies. The man who never reads lives only one. -- GRRM

  8. #28
    Lets do it your way.
    3 hours after the server went online I killed the Nippy slither for 5 hours in a row and it dropped 3 Nippy John Stilletos. I did the same on the next 5 days and it didnt drop anymore. Therefore I conclude that the whole looting system is broken.

    Do you believe me now? Do you think there is a problem with the loot system? Do you think the devs will take it into consideration?

    If your answer is no to any or all of the questions or if your answer is yes to any or all of the questions or if your answer is maybe to any or all of the questions or if your answer is "I dont know" to any or all of the questions then the result is exactly the same. The devs will most likely say it was a freak accident and ignore it and maybe you and certainly a bunch of other people will park their highest dd toon near the mob and check every 30 seconds if the servers are up again the next time a server goes down.

    I posted on my first post, that you apparently did not read before demanding details, information that is much more relevant albeit less precise to determine if there is a problem with the loot generation mechanism. I said that I noticed a repeated pattern of increased drops on some bosses after a server reset. A repeting pattern is very unlikely to be caused by a random system, it becomes even more unlikely when it is tied to another specific event. A much more reasonable explanation is that the system is not as random as it should be. This is the crucial part that neither you nor Masta seem able to grasp. Camping a mob for some time and having a lucky day is unlikely but possible and bound to happen sooner or later. That it happens regularly after some external event however will be MANY orders of magnitude less likely.

    I posted my first post because I am aware of the limitations of my observations. I didnt check with rigorous scientific rigour which mobs drop what and when. I didnt go check all the bosses for an extended period of time after each server reset. That is where you and others where supposed to come in. Posting your own observations to prove or disprove me. I didnt claim that the loot generation mechanism is broken, I posted some observations that may be explained by it.

    Similar observations by other people would give more weight to my own observation and maybe make the devs take some action. Observations by other people that contradicted my own would disprove my explanation. And the devs would not need to do anything. Maybe I made a mistake or maybe I lied or maybe it was a freak accident. Wathever.
    But you in your post start from the opposite direction, if I dont give you all the details then you dont believe me. Which means nothing. Just because you dont believe doesnt make me a liar, or wrong or mistaken or anything. The only thing that you can do to prove me right or wrong is to post your own findings. And if you never farmed the mobs or never noticed anything then there is nothing to it.

    Im done with this topic and with posting any kind of feedback on these forums. From now on I will do what almost everybody else does. If I still can.

  9. #29
    Has somebody ever forced a moderator's hand to umm...

    oh I'm not cool im HAWT

    Edit: I agree with Masta
    * Cryborg Nano-Technician - Have a shoulder to cry on!
    Aramlash Fixer - Can't catch me!
    Aramsunat Engineer - 4 Blockers of the Apocalypse


    Devil Inside

  10. #30
    Quote Originally Posted by Gimpynoob View Post
    To try to please you I will go into that factual thing you suddenly seem to like so much. You started by ignoring what I said in my first post and went straight to the amusing part of superstitions and earsay. You started by moking the guy and are now trying to sound reasonable and rigorous with your demand of facts.

    But since there is nothing more important in this world then to satisfy your need for factual accuracy I will try my best to please you just this once. In at least the last 3 years I have noticed the occurence of an increased drop rate of some rare dyna loot after a server restart in 3 SL playfields, on at least 10 different locations. I am sure of at least 13 but more likely more then 15 events in that an item (or a set of items of the same type) dropped in significantly higher quantities then before. The items range from rare (something that I never saw drop except after a server reset) to common loot (stuff that drops at least 3 or 5 times in a week. In one location it happened 3 times and in another couple of others it happened twice with most of the locations with one event. I can not unfortunatly quantify the exact amounts but I think that saying that in the first day or 2 after a server reset the amount of items dropped was equal or higher then the amount dropped during the following month or even longer. I can also clarify that this did not happen to all the bosses I farmed and that I did not farm all bosses in SL.

    You will ofc with rigorous mathematical precision and irrefutable logic prove that this is not only completly within the limits of complete randomness (I suggest quoting Mandelbrot, fractals are very cute and impress the ladies), but also that I am a total and complete liar because I didnt give the name, age, adress, social security and ID card numbers of the deceased mobs as well as the exact time of their deaths and the name of at least 3 witnesses and didnt present documental evidence of their demise.

    Well I have to be frank with you, as much as it pains me. I DONT CARE FOR WHAT YOU THINK OR WHAT YOUR OPINION IS ON THE SUBJECT.

    And now for the really painful part. YOUR OPINION IS OF NO USE TO THE DEVELOPERS. Contrary to what you might think the developers will be able to evaluate the situation much better if you (or any other poster for that matter) are able to present your OWN experience on the subject.

    So, how many boss mobs did you kill lately on SL? Any phats?

    PS: As a consolation prize I present you with the "Grand Honor" of being the only person on my ignore list. I know you are good old chap and will do me the exact same favour. Have fun.



    Or not. I dont really care.
    I should comment that you shouldn't post anything just to please me, but to make your post work as a point of feedback for the developers.

    And if you want to make fun of the notion of giving statistics as a basis for argumentation, then that's all fine and well, but then how can you expect anyone to take you seriously on a real topic? I was interested in factuality all along, but you didn't provide any from the start and your statement is so preposterous that without any fact to back it up, how can I be anything but amused?

    And don't you think the devs require some kind of proof to warrant spending time looking at code? How would you suggest that you find reduction of loot chance based on server up time in code? If there was indeed such a thing and it was easy to find, don't you think it would've been found by now, what with AO having been mostly rewritten several times? And since they haven't, it must be a larger undertaking, and as such nothing to go around doing without some sort of statistical backing to determine what is going on. Saying things and expecting developers to spend their time looking at that subject just because one player ever says it... that's pretty outlandish. But then, perhaps they DID find that this is true and that's why we have the server "maintenance" downtimes all the time...

    Oh, and the devs probably care as much about what I say as they do about what you say. I don't think that any special prevalence is given to anyone, based on just one post alone. Had you been able to back up what you're saying it would've made more sense for them to listen to what you had to say.


    PS. you can't ignore me - moderator status <3 I've also never ignored anyone on the forum. And over 11 years ingame I've maybe 3 ignores in total. I wouldn't want to miss an opportunity for good drama
    Last edited by Mastablasta; Apr 25th, 2012 at 18:55:38.
    Thor Mastablasta Hammersmith - Level 220, AI 30, LE 70 Clan Atrox Nano Technician - Setup
    The Red Brotherhood

    I'm a Nano-Technician, don't ever expect me to fight unbuffed, alone or fair.

    Means: about f'ing time :P
    Satenia: heresy <3
    Znore: Mastablasta <3
    Kinkstaah: I have agro from many mobs ;(
    Madarab: we are aoe class, we are supose to use pistols
    Marxgorm: the NT toolset does not fit into my raiding tactics

  11. #31
    So it was asked about supporting or not supporting this idea. I will try to be detailed enough to explain my observations and the methods I used with out crushing anyone with huge amounts of raw data. Frankly I don't think I have all the data anymore since I have migrated computers since then.

    My data is older from before the bounty nerf in S10. It is data about bounty drops and when they were better or worse. The data set consist of over 2000 points of data taken in over 120 hours of game play spanning 35 days. So here goes.

    During the period of the data tracking my normal life was out of whack because of a schedule change that had me playing AO at very strange hours for my timezone. It was because of this that I found something that made me very rich because it was repeatable and I made use of it. But it made me want to understand it as well, after all I am a software engineer and these oddities perk my interest.

    I was farming S10 with a 150 enforcer and would log in and start farming. The first hour or so would see drop rate of 5 to 7 bounties an hour. The second hour would see drop rates of 20-24 bounties an hour. Huge difference. Something is up and I tracked it for several days. I expanding my testing to include starting an hour earlier when I could and playing an extra hour longer. In these 4 hour periods I would see two hours of 5 to 7 bounties per hour and 2 hours of 20-24 bounties per hour with the second hour of high bounty drops sometimes dropping to 18-21 bounties per hours.

    As such I adjusted my farming to two solid hours that corresponded with the high drop rate hours and was amazed that the high drops continued. So it was not time spent in zone related that effected the drops. Must be something else and that is when I started looking at what it could be. And was it ever obvious once I saw it.

    The time period I was playing and seeing the change in drop rate occur corresponded directly with the 00:00 time of the timezone the servers are located in. It took a little digging to locate the timezone of the servers but back during the last migration of AO servers to open AOC servers the host company name was divulged and after that it took just a little more digging to verify the timezone.

    So why would the local time of 00:00 matter when we constantly hear about GMT time from FC and such.
    1. Generally computers clocks are set to local time by hosting companies.
    2. New day scripts that rotate logs for the servers and the database are probably scheduled by the local clock time.
    3. Other house keeping scripts probably run at the same time as well.
    4. Back a few years ago a script that runs to up the SL XP award on mobs stopped working and Means related the funniest of tale to the players. There is a thread about it. There is a rough time of when it started to occur. Guess when it corresponds too? Yeah server 00:00 time.

    So I dug deeper to get a better understanding of what I was seeing and try to get more data to support or disprove my idea. First I came across an old GD post in which there was discussion of the rarity of an item drop and the GD posted that at the time 3 of said item had dropped and that was consistent with the rarity FC was looking for the item. So based on that I assume the database for AO tracks how many times items have been spawned (dropped) as loot.

    Having had a hacked account I know FC can back track all your character transactions and recover lost items. So I assuming there are extensive logs being kept is a safe bet.

    Means had said that the mail system created database transactions when you sent something. And that was why there was time delay on deliveries, so the database server did not get crushed. So I assume that the database is at the core of all the server subsystems. If it is not I am really surprised.

    OK a few assumptions yes but reasonably based they are.

    So why was I seeing higher drop rates right after this local 00:00 time. I mean loot tables have a percentage chance for an item to drop right? That is what the whole argument about loot table always comes down too.

    Well it is here I have insight into loot generation. I am a software engineer and have 10 released game titles under my belt and loot generation is a part of many of them. And just trusting a random number generator to give you reasonable loot generation is amateur at best and poor design at worst. So how do you it and keep actual rare drops rare and protected from streaking of the RNG? I mean a 10% drop rate could see rolls of 1 1 2 9 10 3 5 6 10 1 3 4 5 6 3 4 10 10 1 1 1 1 1 3 8 9 4 in a row and the RNG be working properly. So in 30 kills that list of 28 percentages would be a stupid high drop rate for the day and appear that something is really really wrong. Other methods of control have to be included.

    From here on is assumption of how I think AO does loot rolling and why external events such as local 00:00 time and server restarts impact loot drops. Hate on it if you want it is my view and based on actual implementations I have used to create smooth loot drops in release games I have worked on.

    Example:
    Item A on Mob X has 10% chance to drop each time mob is killed.
    Item A has a max per day value of 5. This is a control field to help make items rare.
    Item A has a daily counter that is reset in the 00:00 new day house keeping scripts.

    So the actual drop % chance is 10 - ( Daily Count / Max Per Day Value ).
    The result is that the 5 items that do drop had a scaled drop chance of 10, 8, 6, 4 and 2% over the course of the day. This keeps a rare item rare. In this case these numbers are more like a limited or semi-rare item. But the concept is solid.

    We can see that having Daily, Weekly, Monthly and even life time drop limits can be used to scale the drops accordingly. Does this mean that every items in every drop table is scaled this way? No it does not. In fact it would make sense that these special controls would be used on loot that FC wants to control. Think of the code like this

    dropRateModifier = 0 ;

    if ( itemA.dailyLimit > 0 )
    {
    dropRateModifier += ( dailyDropCount / itemA.dailyLimit ) ;
    }

    if ( itemA.weeklyLimit > 0 )
    {
    dropRateModifier += ( weeklyDropCount / itemA.weeklyLimit ) ;
    }

    if ( itemA.lifeTimeLimit > 0 )
    {
    dropRateModifier += ( lifeTimeDropCount / itemA.lifeTimeLimit ) ;
    }

    actualDropRate = itemA.baseDropRate - dropRateModifier ;

    percentageRoll = CallFuncomRNG( PERCENTAGE_VALUE ) ;

    if ( percentageRoll < actualDropRate )
    {
    // Drop the item the for the poor chap.
    }

    So by setting a limit to 0 they are simply turning off that possible modifier. This method actually works very well. But it does have problems and server restarts and 00:00 new day operations are just two of them.

    So take my bounty data and you can figure that there was a daily limit on bounty drops. And because I just happened to be playing when it reset I get the benefit of the best case drop rate. Since my data shows that the drop rate decreased as time since 00:00 increased I made sure I farmed then and no other time. Until RL decided to change my schedule again. At which time late nights on the weekend became the norm.

    So the OP suggested that server restarts cause this abnormally high drop rate. A server restart would almost certainly involve a restart of the database server. At this time it is safe to say all the new day scripts and other normalizing scripts would run and that all counters than are not forever lasting would be cleared.

    Why do I think this? Consider that Tara and all other mobs are up at server restart. Clearly their re-spawn timers are being reset. Taking that idea a step further and assuming that a server restart is based off the same scripts that would run the first time a new server is setup is reasonable and again something I have done in other distributed programming projects I have worked on so I consider a viable possibility.

    Frankly, I was not going to post any of my observation as they could be considered as proof that I was gaming the system, but the lack of any true prove or disprove efforts left me feeling the need too. Also since bounties are no longer in S10 this data does not help anyone get a leg up on S10 farming. I will refrain from identifying the other items I have found this observation to be true as those items are still in game. The OP has seen a trend that I have been aware for over 3 years now. I used it to farm billions in S10 bounties easily and in the fastest possible manner.

    So have at my post. Tear it apart as you see fit.
    Lheann
    President of When I Grow Up

    Lhisa - MA - RK1
    MaxKillz - Enf - RK1
    Namaru - Enf - RK1

    "If you find yourself loosing a fight, your tatics suck."

  12. #32
    I'd like to see the raw data, nonetheless
    I am wiser than any god or scientist, for I have squared the circle and cubed Earth's sphere, thus I have created 4 simultaneous separate 24 hour days within a 4-corner (as in a 4-corner classroom) rotation of Earth. See for yourself the absolute proof.

  13. #33
    Excellent post.
    Thor Mastablasta Hammersmith - Level 220, AI 30, LE 70 Clan Atrox Nano Technician - Setup
    The Red Brotherhood

    I'm a Nano-Technician, don't ever expect me to fight unbuffed, alone or fair.

    Means: about f'ing time :P
    Satenia: heresy <3
    Znore: Mastablasta <3
    Kinkstaah: I have agro from many mobs ;(
    Madarab: we are aoe class, we are supose to use pistols
    Marxgorm: the NT toolset does not fit into my raiding tactics

  14. #34
    Nice post, Lheann.

    Thread.
    Contra
    Urynt
    Malraux
    Fontane
    Critbull
    Cleanex
    Fontane2

  15. #35
    Quote Originally Posted by Lheann View Post
    stuff.
    have you also tried checking if it's snowing in oslo when the drop rates increase? or lunar cycles? i found the highest droprates to occur on thursdays when it's lightly drizzling in hawaii myself so before hitting up s10 i'd always kiss my lucky rabit's foot and check the honolulu weather channel

  16. #36
    Quote Originally Posted by Lheann View Post
    Lots of interesting stuff
    I have to say I'm impressed at the amount of effort you've put into this post. Thank you.

    I do however see you're "stacking assumptions"; that is, assuming something to be true, then using that assumption to assume something else to be true, and so forth.
    This is fine as long as there's data to back it up; And I'm sure there are.

    It would however be nice if you provided access to said data, so that others can go over it and see if there may not be other explanations for what you're observing.
    ::: My Tools & Stuff :::
    ::: Cratine Savagedheals Enfine Zoewrangle Demoder :: Solitron Demotionform :: IRC Demoder Savagedlight :::
    ::: AOItems :: Blog :: CIDB :: HelpBot :: ItemsBot :: PlanetMap Viewer :: Tower Wars :: Twitter :::

  17. #37
    Quote Originally Posted by Demoder View Post
    I have to say I'm impressed at the amount of effort you've put into this post. Thank you.

    I do however see you're "stacking assumptions"; that is, assuming something to be true, then using that assumption to assume something else to be true, and so forth.
    This is fine as long as there's data to back it up; And I'm sure there are.

    It would however be nice if you provided access to said data, so that others can go over it and see if there may not be other explanations for what you're observing.
    Thank you Demodder. This thread had gotten out of control and I wanted to come back to the OP and address the question asked. So to do that and avoid more of the same this thread offered I had to take the time to write up a very articulated post. Minus the data everyone wants.

    You are right about stacking assumptions and that does even bother me about the entire summation of my efforts. However I can say that using the data to determine this ideal drop rate time I was able to take billion upon billions out of S10 in less time than people I know that farmed it for periods of 5 to 8 hours daily for weeks at a time. The only real difference was they farmed at other times than these ideal hours. So either I was stupid lucky day after day for week after week or there is some truth to this. Now does that mean it is all correct? Nope as I said it is all assumptions past a certain point.

    As I said the data is old and the easy to present excel sheets are lost as they went with the hard drive failure that led me to upgrade to a new computer. And really recovering family photo's, bank data, professional items and email ranked much higher in importance than any AO related data. Though I do make regular backups of my pref folder now as you know that is a PITA to redo on many toons. I know, backup backup backup.

    Anyway, when I get done with RL work travel I will spend some time putting my scribbles back into at least a limited form and presenting a subset of the whole for people to review. It really was simple note taking.
    Hour 1 (22:00-23:00) X number of drops
    Hour 2 (23:00-00:00) X number of drops
    Hour 3 (00:00-01:00) X number of drops
    Hour 4 (01:00-02:00) X number of drops

    Did not always have data for hour 1 and hour 4 as schedule did not allow. At least not full hours worth. After about a week of Hour 1 and 2 being low and Hour 3 and 4 being high I pretty much only played in hours 3 and 4.

    So here is an example of one of the days data. It really is simple note taking. Excuse the crudity of the following but trust me it is better looking than my notes. My scribbles make a doctor's hand writing look excellent.
    Code:
    Hour 1           Hour 2                 Hour 3            Hour 4
    xxxxx            xxxxx                  xxxxx             xxxxx
    xx                                      xxxxx             xxxxx
                                            xxxxx             xxxxx
                                            xxxxx             xxxx
                                            xxxx
    Where each x was a tick mark when a bounty dropped. It took all of 3 days to say it was a pattern. The rest of the data collection was support. I mean in terms of scientific method data collection the hand written grid and tick mark is time tested and proven. If the difference had been statistically insignificant I would have never noticed. A range difference of 5-7 and 4 to 8 is really inside the fluctuation of any RNG system. But the range difference I noticed was like WTF and deserving of more research.

    At the end of the data I simply averaged each hour column and I think I had color coded the high and low numbers in each column to define the extremes of the range.

    The hardest part of this was forum post digging to find the post that gave up clues like the hosting company and the timezone the servers are in. The old GD post about them knowing how many times a super rare item had dropped. Suffering a hacked account to realize that wait a minute they track a huge amount of data. Watching for 11 years after a patch or server restart that all the mobs are up and all the timers are reset. Remembering that Means had posted details about the SL XP boost script and finding the post and dissecting it and then reading the complaint thread and mapping the first reported issues back to the servers 00:00 time or close to it.

    So I am going to ask everyone as well, have you seen anything like this? Have you taken the time to verify it or collect more data? There is clearly more to loot generation than a simple RNG roll but what that is will only be revealed to those that work at FC. The rest of us have to be build theroy based on data.
    Lheann
    President of When I Grow Up

    Lhisa - MA - RK1
    MaxKillz - Enf - RK1
    Namaru - Enf - RK1

    "If you find yourself loosing a fight, your tatics suck."

  18. #38
    If I read you C/C++ code correctly, you are trying to break randomness...in itself it's not bad.

    In statistics when you enter a new roll, the results of previous rolls don't matter.. seems to me you are trying to address that..


    Why not!

  19. #39
    Quote Originally Posted by Ocene View Post
    If I read you C/C++ code correctly, you are trying to break randomness...in itself it's not bad.

    In statistics when you enter a new roll, the results of previous rolls don't matter.. seems to me you are trying to address that..


    Why not!

    In statistics you are correct. In a MMORPG where the idea of rare and semi rare or controlled drop rates do matter then the previous results have to impact the future results. In fact the MMORPG model is almost the opposite of what statistics would suggest.

    Example: If the JEPP and JSPP have for agrument a 3% drop rate. Then you could estimate that there should be X number of them in game due to the Y number of kills over the years (Y Kills * 3% = X number). Of course statistics tell us that no matter how big Y is it will not effect X as each event is isolated. In fact Y = infinity and X = 0 is a statistical possibility. Likewise statistics says that X = Y is a valid outcome. I am sure X = Y is not an event FC wants to see actually happen anymore than Y = infinity and X = 0. So we have to think of drop rate as drop rate over time with time actually being kills. Because we would all love Beast to see a streak and every beast kill for a month or more drop a BOC does not mean FC would. Statistics says X = Y is valid but I am pretty sure FC says X = Y is broke in this example. That is why there has to be other controls.

    So my pseudo code is more like trying to reduce the apparent streaking that all RNG's have. Now I don't know how AO does it but other larger systems generally have an RNG running all the time. And it generates numbers between say for example 0 and 1000000. That RNG runs at a pace of 1000's of numbers per second. When a subsystem needs a random number it calls the RNG subsystem and the current number in the stream is bounded into the requested range and returned.

    What this does is prevent streaking within a single subsystem because no single subsystem can dominate the RNG number stream. And anyone that has played with C based simple RNG's will know that if you use every number generated in place the streaks and patterns will be obvious. But this shared RNG approach introduces a new type of streaking effect. It is possible for a subsystem to request numbers and those numbers be all the same or very close to one another even though the RNG itself has produced 1000's or 10,000s of number between each of them.

    This effect is best described as noise. And the laws of noise within any system states it will exist. Further that any attempt to filter the noise still results in noise as the output of the filtering.

    So how do you deal with this noise. You deal with it by using limiters in your code. Now the rules of noise in a system still apply. Filtered noise is still noise and limited noise is still noise. So in the end all these controls do is just reduce the noise (streaking in an RNG) to statistically irrelevant levels.

    But the pseudo code tries to go further and remove the X = Y and Y = infinity and X = 0 statistical possibilities and create an over time (over n kills if you will) probability of something dropping. If the drop rate starts higher than the desired average value and then declines it is more likely to avoid Y = inf and X = 0 because of the high starting value and likewise the lower and lower following values will ensure that X = Y does not occur as well.

    There is something else that I missed expressing in my first post. We talk about percentage chance of dropping and people assume that is 0-99 or 1 to 100. But what if 100% actually equaled a value of 100000? Does this smooth the RNG effects some (ie a type of filter)? The answer is yes. When A RNG runs the wider the range of validity for the return value the more likely that streaking will not be as apparent (ok this is more masking than it is filtering). Think of it like this. Take 50 pennies and put them on a sheet of paper. The chance of penny being where you touch the paper is VERY high. Now spread those same 50 pennies across a football field. The likely hood of touching one is now VERY low.

    I guess all these long post come to one point, FC used experienced game designers. To think that the loot generation code uses a single value and a call to the C function rand() and that the code could look like:

    int dropRate = lookUpDropRateForItem( itemID ) ;
    int percentage = ( rand() % 100 ) + 1 ; // 1-100 range percentage
    if ( dropRate >= percentage )
    dropTheFats() ;

    Would suggest that a game with such other complex systems went cheap on one of the most critical subsystems is laughable. I think like AAD and AAO and AR and DR that the overall loot drop system and math is very complex. My own experience says you have to have all these types of controls if you want the game to maintain a steady consistent feel.

    And that brings me back to my long post. No matter how complex you make it, external events are very likely to play a significant part in upsetting the smoothness of the loot system operation. After all, these filter values have to be stored and reset at some point. A daily limit would need daily resets. Another type of limit might only get reset at server restarts. So if one is observant, collects some data, uses information that is available they could unlock at least a couple of small pieces of the loot system operation. And that is what my data collection resulted in and is what I think the OP was trying to drive the discussion towards.

    Sorry for geeking out on you all. I design large scale distributed systems for the military as my day job and these type of discussions get my blood flowing and I can go on for some time about them. As always, feel free to rip, flame, compliment or disregard my post as you see fit.
    Lheann
    President of When I Grow Up

    Lhisa - MA - RK1
    MaxKillz - Enf - RK1
    Namaru - Enf - RK1

    "If you find yourself loosing a fight, your tatics suck."

  20. #40
    This guy above me knows his stuff.
    Contra
    Urynt
    Malraux
    Fontane
    Critbull
    Cleanex
    Fontane2

Page 2 of 3 FirstFirst 123 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •