Reply to Thread
Page 2 of 5 FirstFirst 1 2 3 4 ... LastLast
Results 11 to 20 of 46

Thread: VL Adv Search – Alarming Change In Drop Rates?

  1. #11
    Retired Community Leader
    Join Date
    Feb 2013
    Posts
    292
    World
    Ares
    Quote Originally Posted by Perwyn View Post
    One of my guildmates has been tracking his results since after Easter and shared them with me. I've taken screen shots. He's sending out 18 explorers at a time and he's getting the kind of results you would expect. The percentages are in balance and he's not getting 13% secluded experiments.

    Is it possible, (I know it is, and even highly likely), that the search algorithm has been modified and as long as you're sending out a mass of explorers at a time it's working perfectly, BUT, if you're sending out ones and twos at a time, the results are sporradic. (sic!)

    I'm sorry all my friends, but the results I'm getting can't be explained away by bad RNG. Bad RNG is exactly what the label implies, Random. It's not constant and doesn't last for 5 weeks at a time. The results I was getting with just 3 tavern explorers prior to Intrepid were entirely satisfactory. Results have totally hit the fan since.
    Unless there was some truly horrific coding, it would not be possible for there to be two different sets of tables based on the number of other explorers sent out. It would take having the game store the number of explorers for each person and their recent activity. There would have to be addition variables to control for the time gap between them, as well as player-specific adjustment factors. A normal algorithm for using a lookup table for a random result would essentially one line, with the only variables being the type of search and the random number generated. An algorithm to provide for the tables to adjusted for specific players and for their sending patterns would be at least an order of magnitude more complex.

    The answer to the observed data still is most consistent inadequate sample size. This is reinforced by your contrast example, as the other player is obtaining a sampling with at least five times the amount of data and getting percentages that are more consistent with what is expected. This is normal behavior with dealing with a discrete set of returns, and you are still at about a third of size of the sample size needed to look for the change of a single element of the table.

  2. #12
    Mayor Perwyn's Avatar
    Join Date
    Feb 2012
    Location
    The Shire
    Posts
    1,559
    World
    Zeus
    First: 206 (total of comparison data) divided by 89 (total of my data) = 2.31 not 5 so how is that 5 times the amount of data? If you're going to spout numbers back at me in order to prove me wrong, at least be accurate.

    Second: how is it that the addition of one explorer skewed results so suddenly and badly?

    Third: What does data sample size have to do with one explorer skewing results so suddenly and badly?

    Fourth: Given what I assume is your knowledge of statistics, how many more searches will I have to do before the rate on Secluded Experiments drops down to the expected 5% range and YWC moves up to the expected 9% range?

    Fifth: If this was happening to you, would you just sit back and say nothing?
    "Very little is needed to make a happy life; it is all within yourself, in your way of thinking." ~ Marcus Aurelius

  3. #13
    Retired Community Leader
    Join Date
    Feb 2013
    Posts
    292
    World
    Ares
    1) The reason I used 5 times the amount of data was based on the difference in what you were reporting as far as generals being sent out. You have been reporting 3 explorers, tavern type, for the search. These have a cycle rate of 72 hours at base, or 3 days. The Intrepid one is twice as fast, so 36 hours base. Assuming no books, this gives 1.6 cycles per day for the very long search, or counting the intrepid as 2 explorers (reasonable if the 2 returned are independent), 2.3 cycles/day. The other group is using 18 explorers. Using the total number of explorers available, this would mean 3 lucky explorers (x3), 3 tavern explorers (x1), 1 Intrepid (x2), and 11 savage/experienced (x2). Setting the intrepid as a double, this should give 12.7 cycles/day. I was using a rough estimation in comparing 4 explorers to 18 explorers, but this gives a potential sampling rate difference of 5.5. Both sets assume no books were used, which is reasonable for the tavern explorers, but less so for the premium one. Adding books to increase the speed would increase the potential sampling rate difference even more.

    2 & 3) These are essentially the same question. Using a standard random function and lookup table, each result is independent of the previous. With independent returns, the odds of an event happening on the next occurrence does not change based on what has happened before. For example, say a coin was flipped 3 times, and heads was obtained each time. The odds of the obtaining heads again on the next coin toss is still 50%. The law of averages does not really exist, but it is shown for groups of sufficient size; with the size depending on how unlikely the specific events actually are. While another explorer may have been added to increase the sampling rate, it is still a relatively low rate, and a few outcomes cannot be said to skew such a low powered group.

    4) The sample size I posted before to start looking at Ali Baba returns with a power of 80% was 239 searches. That was based on a return rate of 10%. However, this was based on a binomial distribution, comparing just 2 groups - getting Ali Baba vs. not getting Ali Baba. Without going through a multivariable analysis, there are some easy ways to look at it. Taking your "normal" group, the returns before the easter event, you were beginning to approximate the numbers from the other fan sites. The number is fairly close to the number of returns reported from the other same set from the 18 explorers. This would be a good starting number before you can start claiming significance, but should likely be larger as there is still significant deviation between your numbers and the baseline.

    5) The underlying assumption is that something is happening specifically to you. From the other data set with the larger number of returns, it is not happening to others. If I really felt strongly that there was a change in the numbers, the best method I can think of to prove it in a reasonable amount of time is to go to the test server, buy as many explorers as possible, in as many test accounts as I could reasonably do, and generate numbers. As one can get a minimum of the 10 premium explorers and the 3 taverns (not including the ones in the event being tested), with a few accounts running you could generate a fairly reasonable sampling. Just the premiums would give 10 returns each every 1.5 days, or more than 40 in a week (per account). What I would not do is start making accusations without a meaningful sample. As there are no reported tables, it cannot even be said that BB is providing inaccurate information. What I would be doing is enjoy playing the game.

  4. #14
    Mayor Perwyn's Avatar
    Join Date
    Feb 2012
    Location
    The Shire
    Posts
    1,559
    World
    Zeus
    Quote Originally Posted by Aurboth View Post
    4) ... there are some easy ways to look at it. Taking your "normal" group, the returns before the easter event, you were beginning to approximate the numbers from the other fan sites. The number is fairly close to the number of returns reported from the other same set from the 18 explorers. This would be a good starting number……....
    Exactly my point. I had decent results from a small set of explorers before the introduction of the Intrepid.


    Quote Originally Posted by Aurboth View Post
    1) The reason I used 5 times the amount of data was based on the difference in what you were reporting as far as generals being sent out…...
    It shouldn't make any difference how many explorers I'm sending out. And that was the point of my 2nd to last post. You've already admitted that my results prior to Easter were fairly close both to the fan sites and the sampling of someone using 18 explorers. All of this was done with 3 tavern explorers, so from my point of view, the total number of results, not the number of explorers used, is the important factor.


    Quote Originally Posted by Aurboth View Post
    2 & 3) …... While another explorer may have been added to increase the sampling rate, it is still a relatively low rate, and a few outcomes cannot be said to skew such a low powered group.
    Well, that's been your objection all along. That I don't have enough data. I think most people would look at the results and say “Wow! You're getting a whole lot of Secluded Experiments all of a sudden”. For some reason you simply can't, and have to prove to me and the rest of the world that I know nothing about statistics and shouldn't bother posting here. I don't believe that 89 results can be dismissed as too little data if 127 results are coming out close to what's “normal”.

    As far as playing and enjoying the game, I advise you to do the same. I'm done with you. I will continue to post results as I accrue them. If you don't like it, I really don't care.
    "Very little is needed to make a happy life; it is all within yourself, in your way of thinking." ~ Marcus Aurelius

  5. #15
    Mayor Raubhautz's Avatar
    Join Date
    Aug 2013
    Location
    Comporellon
    Posts
    2,203
    World
    Zeus
    All this aside; two things I might throw out there that has been brought up in other threads:


    1. Sounds like your data indicates getting results not in line (distribution) of other players; one issue could be what TSO is using for its RND seed. I say this because it seems that some players who get crummy (not in normal distribution) results get them consistently for certain actions, on a frequent basis. Is it the bad luck of the seed being off your account #?! idk, I have noticed that in most things TSO, I get 'normal' results over a period of time; however, for some functions, e.g. 'Lucky Explorers', mine have a horrible return rate of anything valuable (they run as Artifact Finders only) and further they return a 'Lucky' slot at the rate of 33% - WAY lower than the expected 50-55%.
    2. The only responses on these sticky issues are from other players. Perhaps they mean well; perhaps they even are providing accurate information, which just results in confusion and frustration as nobody responding is an authority on the behind the scenes mechanics of the game; what would be REALLY nice is some feedback from the source - Blue Byte employees, even if not be the designers or developers, but someone who would research the issue and discuss it with them.
    “Many that live deserve death. And some die that deserve life. Can you give it to them? Then be not too eager to deal out death in the name of justice, fearing for your own safety. Even the wise cannot see all ends.”
    J. R. R. Tolkien

  6. #16
    Mayor Perwyn's Avatar
    Join Date
    Feb 2012
    Location
    The Shire
    Posts
    1,559
    World
    Zeus
    In the “You've got to be kidding me department”…… So my Intrepid comes home last night and she has………

    Secluded Experiments & her bonus adventure is YWC. Well at least I got a Woodcutter out of the deal right?

    Then 2 of my 3 tavs get back overnight and this morning there are map frags and…. Wait for it….. Secluded Experiments.





    Let me be absolutely clear for anybody out there who might be thinking “why doesn't he just send out a ton of explorers and be done with it?” . I realize I could start sending out 18 explorers at a time and probably begin to get results more in line with what's “normal”. I mean I've even provided examples of that working for someone else. But that is completely beside the point.

    The point being that I was sending 1 tavern explorer per day, 3 in total, so that once per day one would come back with something. Eventually they would overlap because they came back in the middle of the night, so I was sending 2 at a time sometimes, but I eventually would separate them out again so it was for the most part 1 per day. I did that for close to 3 months and was getting “normal” results.

    The only thing that's different now is that the game was changed with the addition of a new explorer. In theory, nothing in my results should be different. I should be able to carry on in more or less the same manner I was prior and get similar results altho a little bit faster. But obviously the similar results aren't happening, and the overabundance of Secluded Experiments is coming from all explorers.

    So sending out 18 explorers might appear to “fix” the problem. But there was no problem with my method prior to the new release so why is there one now? What's happened? Changing everything I was doing to try and get the results back to where they were is not a solution. That's not how you test something. So I'm choosing to stay with what was working before and hope that either <1> eventually things even out and my proportion of found adventures falls into line, or <2> we hear from BB that Oops, something's not working quite as we intended and it will be fixed in an upcoming game update.

    There's no amount of stretching the imagination that can get me to even begin to consider that the results I gathered over the course of almost 3 months using the method I was using are somehow incorrect and that what I'm getting now is the way it should be working based on my method. I don't believe the game is intended to work that way. The number of explorers sent out should ultimately have no effect on the results over the course of time. If the only way to get “normal” results is to send out armies of explorers at a time, why was it not the case for me prior to the Intrepids release?

    Something I realize is that I've gotten completely obsessed and compulsive over this matter and it's ruining my sense of fun in the game. I want my game play to function consistently, even when a new feature is introduced. But I can't control what BB, or anybody else does or doesn't do. So if BB continues to not offer any indication that these posts are even being read, or if any of my friends or other players can't seem to grasp the essence of what I'm saying is the problem, when really, it's quite simple and quite visible, well…. I just gotta laugh and go run some adventures.
    "Very little is needed to make a happy life; it is all within yourself, in your way of thinking." ~ Marcus Aurelius

  7. #17
    Mayor Perwyn's Avatar
    Join Date
    Feb 2012
    Location
    The Shire
    Posts
    1,559
    World
    Zeus
    Aurboth, I need to apologize for losing my temper the other day and firing off hasty responses to your comments. I should have logged out and cooled off rather than letting my frustration with a game mechanic spill over like that.

    From your comments about statistics and your link to a site on them, it's quite obvious you're knowledgeable and probably used to dealing with this kind of data. So the first contention between us was my first set of results. Maybe you saw 127 results and instinctively thought “phhttt, that's not enough data”. I don't know that for sure but it's a pretty good guess. Had I sent out a gang of 18 explorers 7 times to get those results, I'd be in complete agreement with you. Gathered in that manner, it's not enough data.

    However, It was gathered over the course of ¼ of a year. I found a screen shot I'd taken of my level up screen when I hit 56, the level that opens up the very long search. The time stamp was late in the evening on 12/5. So I started doing searches the morning of 12/6. The manner of search was one tavern explorer per day. You may not agree, but to me, logging one result a day over the course of 3 months 9 days is as accurate as sending out gangs of explorers for a lesser amount of time. I think either method will capture the essence of the game mechanic. How else can it be explained that my percentages are in the ballpark with the various wikis and the example of gang searches I provided. We may never agree on this, but either one of us acting like jackasses over the point doesn't further the investigation of the problem if there indeed is one.

    The other thing we may have to agree to disagree on is the number of results since the Intrepid. To me, there should have been no real change in the ratios when another explorer was added. I find the very large increase in the amount of Secluded Experiments quite troubling, and in all honesty don't see how it's going to fall back in line with what it should be any time soon. That was the point of me starting this thread to begin with, the increase in the occurrence of SE. I mention YWC, yes, but my main concern is SE. What happened? Why the sudden spike? I haven't changed anything I'm doing other than adding 1 explorer, so it's way out of proportion for some reason. I guess only time will tell since I'm sticking to a small number of explorers.

    At any rate, I don't expect to be best buds going forward, but I hope we can interact with civility and respect. Sorry I lost my temper.
    "Very little is needed to make a happy life; it is all within yourself, in your way of thinking." ~ Marcus Aurelius

  8. #18
    Mayor Perwyn's Avatar
    Join Date
    Feb 2012
    Location
    The Shire
    Posts
    1,559
    World
    Zeus
    le sigh...... Of course this simply means I will track and post my results till I'm 90 years old and hope someone's paying attention.
    Last edited by BB_Sarrunah; 04-25-16 at 08:27 am. Reason: content from Support
    "Very little is needed to make a happy life; it is all within yourself, in your way of thinking." ~ Marcus Aurelius

  9. #19
    Community Manager BB_Sarrunah's Avatar
    Join Date
    Dec 2015
    Posts
    111
    World
    Zeus
    Hello,

    Thank you for posting your findings. The drop rates should be working correctly, but our Team will ensure it's working properly. If something is found to be amiss and information comes available, I'll update this thread and let you know.

    Happy Settling,
    BB_Sarrunah

  10. #20
    Mayor Raubhautz's Avatar
    Join Date
    Aug 2013
    Location
    Comporellon
    Posts
    2,203
    World
    Zeus
    Removing content from support? Seems harmless enough if it is subject related. More information available for the rest of is us a good thing.
    “Many that live deserve death. And some die that deserve life. Can you give it to them? Then be not too eager to deal out death in the name of justice, fearing for your own safety. Even the wise cannot see all ends.”
    J. R. R. Tolkien

Reply to Thread

Posting Permissions

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