Old Phoenix artifacts

[Sir Derf]

"A strange game. The only winning move is not to play. How about a nice game of chess?"

 

[Deleted User - 1759805]

@MinMax Gamer I looked at your US server thread and it felt a bit like being in an episode of The Big Bang Theory (I love the show btw)
For us not so versed in maths and statistics: Is the conclusion that there are 2 pools one with frequent recipes and one with those more rare? Will you make a list of with recipe goes in which pool? And is there any way to know how recipes get picked from the pools? I have notice to frequently have a lot of constitutive selections of only those frequent recipes and then get all the rare items in the same selection. Wouldn't it be expected to have something like 4 recipes from the frequent pool mixed with 1 recipe from the not frequent pool?

Not enough data to answer most of that, unfortunately. We only have about 1500 observations between basically 2.5 cities so far, and given that there are ~80 core recipes that's not quite enough to be certain about most of the things you're asking. One conclusion that can be made is that hypothesis of all recipes having the same probability is very, very unlikely with the data we observed so far. And apparently community manager on French forums confirmed as much, even though I wouldn't necessarily put much weight into that by itself.

2-pool hypothesis is the one that does work quite well with the data we have so far. It does not mean it is correct We can have all recipes have their own individual frequencies that are different from each other - we just don't have enough data to differentiate between them yet. But we can differentiate between high-frequency and low-frequency recipes as these are already quite different from the ones in the other group. As to which is which, you can take a look at the last couple of posts in the thread, I explain how you can tell which recipe is which (and some are less certain than others).

BTW, all the analysis there assumes that recipe selection is driven entirely by random process with some fixed parameters. If, say, Inno exerts some more specific influence on the recipe selection (based on some city parameters, timing etc), then all this stuff can get thrown out of the window. I mean, data would still be correct, but it would be really hard to generalize it.

 

[DeletedUser9591]

Observation matters. Another important thing is sample size. Example with the spire: genie has a probability of 5%. Say, given sample size is 100, then if I will do spire 100 times, I should get 5 genies. But if sample size is 100000, then I might play the spire 1000 times and get 0. I stopped paying attention to the numbers describing the chance, because my observation is that most likely the spire is built using second scenario Library has a similar probability, but in reality it is won more often.

Same with the phoenix recipes. Some players should wait ages for something to appear, others are getting one after another. Universal evolve would be answer.

At times it looks like astonishing lack of balance in this game. I personally do not find this entertaining...

 

[ANIKHTOS]

Observation matters. Another important thing is sample size. Example with the spire: genie has a probability of 5%. Say, given sample size is 100, then if I will do spire 100 times, I should get 5 genies. But if sample size is 100000, then I might play the spire 1000 times and get 0. I stopped paying attention to the numbers describing the chance, because my observation is that most likely the spire is built using second scenario Library has a similar probability, but in reality it is won more often.

Same with the phoenix recipes. Some players should wait ages for something to appear, others are getting one after another. Universal evolve would be answer.

At times it looks like astonishing lack of balance in this game. I personally do not find this entertaining...

Fixed it for you.



Is your account going backwards?



Register where? As a gambling site? For a game that gives no tangible rewards?

And I thought the problem was whether the game mechanics favored random vs. skill?

you still not get anythign about probabilities i
but if you did you would not make the comments you make.

the thing is that as the number of repetation increase it will also increase the cluster effect
aka getting the same result in a row

which while you may have the 50% in our simpel coin scenario that means nothing if you look at a specific area of the repetition

you may and will get more head and more tales and they will constant flip which one is more. but in the bigger picture you will still have the 50%
when the repetition become extreme big the clusters that will also become bigger will not practically change the 50% ratio

and the problem is that you do not know when this cluster will happen but you know they will happen
and there is formulas to predict how big cluster you can expect to find in a given number of repetitions.

and once more even if you have 100 cluster very likely to happen lets say 99.99% to happen in the number of repetition you have it did happen or in other cased you may even end up with even bigger cluster.

and is so nice for you to flip flop all the time
you say we need to talk about expecting rate and then you say game ins unfair.

and yes is bad game mechanics if you are classified as a game of skill and in fact you have elements of luck that completely de rail some players.

and also to make things worse it does not matter how many times you got the recipe but when you got the recipe
you may have got the recipe more times than the expected drop rate but got it when you did not have artefacts to change so it was not use for you

so what in this scenario are you lucky or unlucky?
what if you got 2 times more the recipe than others but it was all conentrate in the beginign of the event
and at the beginning of the fellowship which meant you had no chance to play and win the artefacts to convert to begin with.

so are you lucky or unlucky? if you see the raw nubmers you were extreme lucky judging the nubmer of recipes you won but if you analyze the data also when they where given you may conclude that you where actuall extreme unlucky
because in the end when you got the artefacts and wanted the recipe you simple did not get the recipe which is what you will expect to happen as the repetition increase you expect to get the %

and as i said to you and you ignored one day i spent lots of currency to open chest to get the daily prize and i manage to get 8 that day. another players with more spent them and got 0 and lets be serious me and him mostly played the event for the daily offer not interest about the new phoenix and in the first time i did not know if i should even try to convert the new artefacts into the old ones and try to build a fire phoenix half way the event i made my mind to go to build a fire phoenix.

also you did not paid any attention when i said to you that probabilities is the hardest part of maths and is a massacre to calculate things.
the only secure way we have to be sure we have the right answer is brute force the problems and create and count every possible outcome.
thats how hard this part is, and i was not jocking when i said that even mathematician have hard time in this area
the majority of problems are counter intuition and most probably you will start and try to use the wrong tools and formulas to solve a problem
thats why the brute force comes into play, thank to computer power at least we can have the answer to a problem that we know is 100% correct

and i said to you your mentality of probabilities are the same as the gamblers i will go with expected % and that why all of gamblers end up poor and destroyed
because the cluster will happen and financial destory them


and once more no one is going to keep playing a game and getting the short stick every single time.
and once more the person who start the thread agrees with me and not with you
that person feels betrayed and wants to stop playing the game and i do tot blame them

 

[ANIKHTOS]

Not enough data to answer most of that, unfortunately. We only have about 1500 observations between basically 2.5 cities so far, and given that there are ~80 core recipes that's not quite enough to be certain about most of the things you're asking. One conclusion that can be made is that hypothesis of all recipes having the same probability is very, very unlikely with the data we observed so far. And apparently community manager on French forums confirmed as much, even though I wouldn't necessarily put much weight into that by itself.

2-pool hypothesis is the one that does work quite well with the data we have so far. It does not mean it is correct We can have all recipes have their own individual frequencies that are different from each other - we just don't have enough data to differentiate between them yet. But we can differentiate between high-frequency and low-frequency recipes as these are already quite different from the ones in the other group. As to which is which, you can take a look at the last couple of posts in the thread, I explain how you can tell which recipe is which (and some are less certain than others).

BTW, all the analysis there assumes that recipe selection is driven entirely by random process with some fixed parameters. If, say, Inno exerts some more specific influence on the recipe selection (based on some city parameters, timing etc), then all this stuff can get thrown out of the window. I mean, data would still be correct, but it would be really hard to generalize it.

one question does the % remain constant?
it will not be reduced in event like this one to accomodate the 2 new recipes which means all others will have to have lower %
and also the increased the rate of the old phoenix buildings?

 

[Deleted User - 1759805]

one question does the % remain constant?
it will not be reduced in event like this one to accomodate the 2 new recipes which means all others will have to have lower %
and also the increased the rate of the old phoenix buildings?

The analysis ignores all the temporary (like artifact conversion) and one-off recipes (like chess set and evolving buildings), as these would be different for different people/times. It only compares core recipes that everyone gets no matter what.

 

[Sir Derf]

@ANIKHTOS, I'm having a hard time responding.

Firstly, we are spreading out with more and more parts to in each reply, as I point out three aspects in your post, then you address each of those three with two each, etc., etc., It's the argument form of a hydra.

I've tried to craft a reply, and I've literally spent an hour and a half scrapping five separate attempts. So, I'm throwing the ball back into your court.

Please explain your point about clustering as it applies to recipe-less weeks and 50k players, and please do it with some actual numbers. I understand what clustering is. I understand the scenarios as described with the coin flips. But I'm not seeing what aspect of clustering you are trying to apply to our scenario.

 

[DeletedUser]

dear friends,
due to the unravelling of the discussion I see no other option but to lock the thread
@ANIKHTOS @Sir Derf this message goes out to you both in particular.