• Dear forum reader,

    To actively participate on the forum by joining discussions or starting your own threads or topics, you need a game account and to REGISTER HERE!

Discussion New Spire of Eternity rewards!

little bee

Well-Known Member
While all of the above is accurate, it is also a bit of a red herring. No one implements their own pseudorandom algorithms (outside of very, very specific cases, and online gaming is definitely not one of them). Modern PRNG algorithms, while indeed deterministic, are indistinguishable from the true random for all practical purposes, including passing statistical randomness tests. The same PRNG algorithms are used in variety of fields where having a good source of randomness is crucial (e.g. various simulations), and they are definitely good enough even in those cases.
Well, I think it would be an interesting experiment to see, if the RNG in Elvenar could pass a statistical randomness test while looking only at a single city at a time (rather than averaging over all cities). But that was not my main point. My main point is that the RNG of a game creates the illusion of randomness while actually being predetermined. It is possible to implement it in such a way as to be practically random or it can be implemented in a way that prevents streaks of good or bad luck.
Which one is better depends on the situation and on your point of view. Some people like randomness and others don't. Personally I don't and it seems I am not alone. And I think that the randomness is particullary frustrating when applied to something like artifacts as they are extremly valuable in small and useless in large numbers. In the events they are always given as grand prices on a deterministic rotation. I think that system works well and I would like to see something similar for the artifacts in spire. That is why I suggested changing the RNG for the spire rewards to a form of hidden shuffle deck and thereby creating the illusion of randomness while actually giving everyone the same predetermined number of artifacts.

There are of course downsides to this suggestion. First of all, some people might like the randomness and not want to remove it. Secondly all probabilities in the spire come as multiples of 5% so it would need at least 20 weeks to go through a single cycle, which is a lot longer than the 6 weeks for which the phoenix artifacts have been announced. And finally, such a change could make the spire rewards become predictable. If the plan is to rotate between diffrent artifacts than it could allow some players to manipulate which artifacts they get by possibly skipping the spire for a week. But personally, I think it would be an improvement.
 

Earwen1

Well-Known Member
Happy for you new players, game is rewarding you while punishing us for playing 3 years without possibility to get universal artifact.
Merry Christmas.
Happy for you too, oldie. As you have had the benefit of several Phoenix' and Bears while we got punished with just the eggs in the MA.
Happy New Year!
:diamond:
 

Arthus

Well-Known Member
Happy for you too, oldie. As you have had the benefit of several Phoenix' and Bears while we got punished with just the eggs in the MA.
Happy New Year!
:diamond:

Then you could play as long as we did so you could earn them, nobody forbid you.
 

CrazyWizard

Well-Known Member
Well, I think it would be an interesting experiment to see, if the RNG in Elvenar could pass a statistical randomness test while looking only at a single city at a time (rather than averaging over all cities).
How would that be possible?

how can you compare a limited number of rolls vs the average?
the whole point of RNG with a limited dataset is that it's random.

There is no wat you can properly compare any limited dataset to anything.
 

maxiqbert

Well-Known Member
How would that be possible?

how can you compare a limited number of rolls vs the average?
the whole point of RNG with a limited dataset is that it's random.

There is no wat you can properly compare any limited dataset to anything.
you can compare a large dataset to expected results, it's called statistics
 

little bee

Well-Known Member
How would that be possible?

how can you compare a limited number of rolls vs the average?
the whole point of RNG with a limited dataset is that it's random.

There is no wat you can properly compare any limited dataset to anything.
You could take a few cities and write down what each one wins in the spire for a few month and then compare how many of these cities are close to the expected value. If too many cities have good or bad luck streaks then something is probably wrong with the RNG. Of course I am not an expert on statistics and this would require a lot of datacollection. But I hope that INNOs experts try experiments like this rather than just checking that the average comes out right.
But again, my main point is that the variance seems too large even if the RNG is working perfectly.
 

CrazyWizard

Well-Known Member
you can compare a large dataset to expected results, it's called statistics
Yes but little bee specifically said a single city

You could take a few cities and write down what each one wins in the spire for a few month and then compare how many of these cities are close to the expected value. If too many cities have good or bad luck streaks then something is probably wrong with the RNG. Of course I am not an expert on statistics and this would require a lot of datacollection. But I hope that INNOs experts try experiments like this rather than just checking that the average comes out right.
But again, my main point is that the variance seems too large even if the RNG is working perfectly.
Lol you are already backtracking what you wrote earlier,
The larger the dataset the more it gets in line with the "average", this is what we call RNG.
The smaller the dataset the more random your results the bigger the dataset the more you get towards the average.

Adding multiple cities opver many months is increasing your dataset and therefore making it more stable.
How many data you need to get a good picture of all the odds, I would say ask a mathmathician.

I assume it's smaller then most of us would expect but how small you can go no clue and not in the mood to figure it out as it doesn't matter to me.
 

Deleted User - 81190

Guest
Lol you are already backtracking what you wrote earlier,
The larger the dataset the more it gets in line with the "average", this is what we call RNG.
The smaller the dataset the more random your results the bigger the dataset the more you get towards the average.
Tests for randomness (and that's what @little bee is asking about) have very little to do with averages. I may have an experiment with fully random (uniform) coin, or the one that alternates heads and tails on every flip. The averages would be the same, but the second one would have no randomness whatsoever. You won't even need a lot of observations to figure that out with high degree of certainty.

Of course, most RNG are not quite as obvious as alternating coin, so will require more observations to detect randomness failures. And if it takes long enough, then it is good enough for practical purposes.
 

Pauly7

Well-Known Member
You could take a few cities and write down what each one wins in the spire for a few month and then compare how many of these cities are close to the expected value. If too many cities have good or bad luck streaks then something is probably wrong with the RNG.
This is one thing that I believe Inno are doing all the time, to see whether people are receiving what is expected. Lots of purple chests this week, anyone?
 

little bee

Well-Known Member
So the 6th week of the new spire rewards will end tomorrow. Any news on what happens now or when this hits live?
 

Enevhar Aldarion

Well-Known Member
It has been announced on the EN servers, but not yet on the US servers, so either everyone is getting it or EN is doing another extended beta test. If anyone not on EN gets the announcement, let us know.
 

Lelanya

Well-Known Member
US Forum got the Announcement as well now too.
So no word yet, on if we're getting changes, or what they will be?
 

Arthus

Well-Known Member
In my opinion it should be 10 weeks instead of 6.
Of course if you plan to add some old Artifacts to the Ashen Phoenix event (like last year with Coldfire) then it changes whole perspective.
 
Top