Colin Johanson on Loot Drops

Yesterday, Colin made a post on the forums concerning loot drops. Apparently, there has been a glitch in the code that has prevented players from being rewarded in certain cases. This has been obvious to many for a while now as champion level mobs don't always drop loot even though Arena Net has stated that they should. The official word now is that this problem has been located and fixed. Here is what Colin had to say,
The rules by which you qualify for credit for an event or for experience when killing a mob are different rules than those used to help determine if you qualify for loot. Some of these were set up to be unintentionally restrictive, and as such you could kill a creature with a lot of health (this was most noticeable on champions) and not qualify for loot, despite qualifying for all other credit. 
For the February Flame & Frost: The Gathering Storm release we will be repairing this issue to ensure more players are capable of earning loot credit for kills. This should fix the issue such that you got credit for killing a champion, but received no loot from it. Once fixed, all champion mobs will correctly always drop 1 blue or better loot item, and all veterans will have a chance of dropping loot better than regular mobs as outlined in the November release. Please note, it is still possible to kill a creature and not do enough damage to qualify for loot (which is the only way you won’t get loot for a champion), it is also still possible for veterans to not drop loot at all, since they simply have better odds of dropping loot than normal creatures but no guaranteed drops.
This is great news! I am glad this is finally being addressed and fixed. Not getting loot from a champion is extremely frustrating. As with everything on the Internet, this post has turned into a controversial topic. A lot of players are frustrated after weeks of telling Arena Net this problem existed without receiving proper communication. This may be true, but in the end though I think we can applaud Arena Net's choice to admit to the mistake and quickly work towards getting a fix in game. As if anticipating this bad reaction Colin attempts to placate the masses with promises of more loot.
On top of this, we’re also updating some of the loot in various areas of the game for the February release that should make specific areas more rewarding. We’ll cover this is in detail with the February release notes.
And he wins because you can't argue with more loot. At least I won't.

Still a lot of players seem to believe that there are further issues in the game relating to a change made in the November patch which reduced the amount of loot players were receiving. The official word on this is that it is now an "X-Files level conspiracy". Although a horrible choice of words, since in the end the X-Files "conspiracies" turned out to be true, we can get what Colin means by reading his words in context.
Just wanted to give a quick update on this topic. We've completed verifying every update from the November release and there were zero changes to anything what so ever that would have negatively affected loot in any way/shape/form. We can officially confirm this as an X-files level conspiracy at this point.  
We’re in the midst of evaluating every loot table in the game and running massive random roll evaluations table by table, as well as evaluating every system game wide that causes a player to qualify for items to determine if any issues exist in those systems from launch.
So, what Colin is saying is that there was nothing changed in the November patch. They have looked at the code and confirmed that this is the case. You have two choices now. Firstly, you can believe Arena Net and move on and enjoy the game as is. Secondly, you can choose to not believe Arena Net and be upset and bitter. I choose to be happy. There are answers to why a large number of players think something changed in November when in fact nothing did. You do not have to resort to thinking Arena Net is lying or incompetent.

What a player might be observing is a form of the confirmation bias. A good example of this is that many people believe that there are more emergency room visits during a full moon then on any other day. The only evidence for this comes anecdotally from nurses and doctors. They are more likely to remember and report a surge in patients on a night with a full moon than any other night. Thus, it may seem like the full moon causes people to go a little crazy. Statistics do not lie and it can be shown that there is no increase in emergency room traffic on nights with a full moon. This equates to the situation in Guild Wars 2 where you have, firstly, players only noticing when they are not getting good loot, and secondly, only players getting bad loot are reporting their problems. Thus, it may seems like there is a problem when in fact one does not exist.

The other thing that may be a factor here is that people are terrible judges of true randomness. As an example take the following two images. Which do you think was generated by the most random process?


It turns out that the image on the left was generated by simply placing 100 random stars with in the fixed area using a random number generator. The image on the right was generated by first dividing the entire area into 100 squares and then randomly placing a star inside each of those squares. See for yourself in the image below. No two stars are in the same box.


It is hard for a lot of people to accept that the image with the black stars is in fact generated by a more random process than the image with the blue stars. This has a lot to do with how the human brain is constantly looking for patterns. When the brain sees these patterns it attempts to correlate them to a cause even if a cause does not exist. Essentially, this is the illusion of luck. It is why people can believe that they are on a "hot streak" or why they might believe an object gives them an increased chance at success. Some call this the Gambler's fallacy. In the end it is all the same thing. People are terrible judges of randomness. That is why we invented statistics.

So, what does this have to do with Guild Wars 2? Imagine your character is walking around in the image with the black stars above. Each time it encounters a star you can equate that to "getting good loot". Now if your character was in an area of that image with a clump of stars before the November patch then afterwards you managed to find yourself in a void you might say to yourself, "Hey, I am getting less loot after then November patch then I was before. The patch must have changed something." This is a perfectly reasonable hypothesis, especially if you think the world looks like the image with the blue stars. This hypothesis has been shot down by Arena Net who have looked at the bigger picture. They can see the entire image and use statistical techniques to determine the randomness of the spread of stars. Unfortunately, as players all we can see is what is around us. Thus, in the end I am going to trust Arena Net when they say that there is no difference pre and post patch.

I hope this quick look at probability, math, and psychology has given you something to think about. I have attempted to back up my statements with links to sites with more information if you are interested in exploring these topics further. All-in-all my advice to everyone would  be to be optimistic about the situation. More loot is coming in February and an outstanding bug has been corrected. Lets move on and enjoy the game.

What do you guys think? Are you happy with these changes/fixes? Do you buy my theories as why these conspiracy theories started? Let me know in the comments.

11 comments:

  1. I think the conspiracy theories are caused by ANet's slow response to multiple "bad/bug" comments.Response speed is the key.

    ReplyDelete
  2. Excellent article, and this points out exactly why I've been pretty happy. Most of the complaints I've found are subjective and fallacious.

    Also, I have to point out "The patch must of changed something" should read "must have".

    All in all, great stuff :)

    ReplyDelete
    Replies
    1. Glad you enjoyed the article, and thank you for the correction :)

      Delete
  3. Reminds me of the http://en.wikipedia.org/wiki/Seattle_Windshield_Pitting_Epidemic
    It's collective disillusion.

    ReplyDelete
  4. A good article. I like the part about true randomness. I hope that this "fix", when it hits, will fix the "no chest" from the dragon fights that have been going around. Colin's comments seem to lean toward this but they only mention Champion mobs so I can only hope at this point. If this was the update at the beginning of february and not the one coming next week, then there are still problems.

    ReplyDelete
  5. Still does not explain why I always get only white items when I am in full MF gear and MF food/boosts running but when I wear my normal armor and activate no food or boosts I will get blues/greens and the occasional yellow. To top it off everyone in my 230 person guild has received rare drops that had high value on the TP and I have never had a drop that is worth more than 40 silver call it bad luck but I call it bad code.

    ReplyDelete
    Replies
    1. You should read the part about randomness again..and again..and again...

      Delete
  6. Loot was reduced circa Nov 15. See
    http://www.gw2spidy.com/gem
    for proof (select Zoom->all).

    ReplyDelete
  7. I often say "RNG hates me." I know this to be inherently false, but that is up to this point my reality. I played GW1 since launch, I id'd every rare item I ever got. I only achieved the first tier Wisdom (100 rares id'd) title. I sit at 123 in 8 years.

    Move forward to Guild Wars 2. Ignoring guaranteed exotic drops (Mad King tower, zone completion, etc.) I've gotten 3 total exotics, Galrath's boots in WvW, Tsunami from Karka finale, Carrion Ring from during Cathedral of Silence (story). Rares (again excluding guaranteed drops) are at a count of 9. I have played every single day since early access (8/25 for me) for at least 2 hours each day. That means at a bare-minimum I've put in 352 hours since launch. This is a low estimate since there have been numerous days where I've spent 4 to as many as 8 hours in game.

    So let's take the 352 hour estimate and divide by my 12 valuable drops. 29.3 hours per "good loot". That is down-right disheartening.

    I KNOW the RNG bears me no ill-will, but I wander the wastelands between your black stars and that makes me a little bit sad. Fortunately, I enjoy the gameplay itself and the people I play with and that keeps me playing. Were I the type solely motivated by acquiring gold I would not be playing this game still.

    For every player that gets Dawn after Zap after Charged Lodestone after random black star, there's another one like me that is over in that white area. Some day we might switch places, but personally that hasn't happened yet and there's no guarantee that it ever will (random is random). It's an imperfect system that all things being equal randomly values one player's time over another's, but I have yet to see an alternative that doesn't come with its own problems. And, as such, I just deal with it and try not to get sad when people are spamming their awesome drops in mapchat when I just looted yet another glob of gloppy goop.

    ReplyDelete
  8. My brain hurts...i feel dizzy...can't concentrate..
    Ah there's solution..i'm going to a programmer. He will fix it..

    People are always talking about randomness on a computer.Is there a mysterious device in it...producing some random numbers. No it isn't.

    Computer:"The user clicked word, super, so i'm opening paint."
    You don't want that.
    It's a logical device..no choices, no randomness at all. Thats the way it is constructed.

    So now talk about randomness in such a system....how is it possible to create?
    1)U can compute a complicated mathematical row, take the result and fit it to your problem.
    -> numbers appear to be random
    But everytime u run the algorithm, it will produce the same "random-numbers"
    2)In such a system there's only one (and i mean one) source of "randomness". That is another system interacting with it. Right guess..it is called "user" here.
    There are a lot of approaches to use the user. For example u can count the milliseconds your Windows is running into the algorithm mentioned above. The time the computer was started will most likely be different every day. Another way would be the interactions with a keyboard or the mouse.

    There is no "true" randomness on a computer...only a more or less clever algorithm pretending it. The more effective it is, the more processing time it will need. In computer games, your choice will most likely a very simple one..obvious why...

    ps: if my brain hurts..i won't go to a programmer. I prefer to go to a guy knowing a lot about "brain-patterns"

    ReplyDelete
    Replies
    1. The other option is to have a large pre generated list of random numbers that you select from. Rand Corp still does this I think. See, http://en.wikipedia.org/wiki/A_Million_Random_Digits_with_100,000_Normal_Deviates

      Delete