Random

isprite
Citoyen
Posts: 36
Joined: Tue 13 Sep 2016, 11:28
Realm: -

Random

Postby isprite » Fri 12 Jan 2018, 08:06

Gala-mining at 25% chance: 48 total, 5 successes. Why do you license Chinese random?
Image
Image
Image

User avatar
Céline
Mystique
Posts: 2286
Joined: Fri 20 Nov 2015, 20:42
Realm: -

Re: Random

Postby Céline » Fri 12 Jan 2018, 13:38

And sometimes you will have 3 or 4 "win" in a row.
Don't know the English for "tirage sans remise"... it's not 48 and 5.
It's 1 and 1 and 1 and 1. With the same % for each.

And 48 is not a lot for such "tirage aléatoire". If you do it 1000 (minima) times and have 10%... ok there is a problem.

But, and it's not your fault, Human have a better memory for what's "not good" and quickly forget all the "too good" things ;)



(or maybe... we pay underpaid Chinese children to roll dices each time a character do something ? :D )
---
~Contact : support@virtys.com~

IF_
Coordinateur
Posts: 171
Joined: Wed 11 May 2016, 11:43
Realm: Vodekan

Re: Random

Postby IF_ » Fri 12 Jan 2018, 14:28

Use the Force, Luke!
Ваши байты - биты!

Cosette
Citoyen
Posts: 18
Joined: Fri 24 Nov 2017, 23:49
Realm: -

Re: Random

Postby Cosette » Thu 25 Jan 2018, 23:07

Just to you know, things like this happening, are why people quit playing. There are already far too many time sinks in the game. I think its time you guys take a step back and really decide if this is good for the game or not. The Nostria + game is not fun at all, just time consuming and annoying for a lot of things. So many people have quit on the Steam Server after reaching Nostria. It makes it really hard to log in and spend money on your game when you get the foreboding feeling there wont be anyone left to play with soon on that server. I have actually stopped buying BC for a while now because of it.

Last night on Steam Server, there were 7 players online at login and 4 of those where me. That is not good at 6pm.

Borgratz
Défricheur
Posts: 8
Joined: Thu 15 Jun 2017, 08:09
Realm: -

Re: Random

Postby Borgratz » Sun 28 Jan 2018, 13:42

@Cosette
What exactly would you prefer instead? The way I see it, you can either have people quit, because they ran out of stuff to do, or because the grinds become too frustrating for them.
Everything else requires the continent after Nostria to exist.

Cosette
Citoyen
Posts: 18
Joined: Fri 24 Nov 2017, 23:49
Realm: -

Re: Random

Postby Cosette » Thu 01 Feb 2018, 19:58

No I understand its a Catch 22. I just hate it when you finally find a person interested and help them out, to have them just quit. Makes it hard to help out the next guy and the next etc when most quit. Wish there was a better answer.

Rufio
Citoyen
Posts: 16
Joined: Fri 02 Feb 2018, 08:05
Realm: Nostria

Re: Random

Postby Rufio » Fri 02 Feb 2018, 08:11

Céline wrote:And sometimes you will have 3 or 4 "win" in a row.
Don't know the English for "tirage sans remise"... it's not 48 and 5.


FWIW I think that's "sampling without replacement" - each draw is independent. Humans are great at seeing patterns in things even when they don't exist.

That said, I do think there's something fishy in some of the 'runs' I've seen. In my early days I stopped using BCs to rush things because a run of ~20 failures to skillup when the chance was >50% convinced me that %skillup was based on real time elapsed (i.e. time skipped when rushing was ignored). I never really started doing that again so I still don't know if I had a (literally) one-in-a-million unlucky streak or if it was a bug.

Tiny Hippo
Bourlingueur
Posts: 99
Joined: Wed 24 May 2017, 10:15
Realm: Tortolion

Re: Random

Postby Tiny Hippo » Fri 02 Feb 2018, 11:11

Rufio wrote:That said, I do think there's something fishy in some of the 'runs' I've seen. In my early days I stopped using BCs to rush things because a run of ~20 failures to skillup when the chance was >50% convinced me that %skillup was based on real time elapsed (i.e. time skipped when rushing was ignored). I never really started doing that again so I still don't know if I had a (literally) one-in-a-million unlucky streak or if it was a bug.

Yes, this problem exist. When you use acceleration, the probability of skill growth, even with a high chance, is much lower. I think this is a problem on the server side. With acceleration, many problems.
I'm not rude to you, I'm just hurtful
:lol:

User avatar
Balok
Bâtisseur
Posts: 140
Joined: Wed 06 Jan 2016, 16:45
Realm: Fauconnia

Re: Random

Postby Balok » Fri 02 Feb 2018, 19:11

Last night I hunted something 24 times at a 50% chance. Got 7, should have got 12. And that's far from the first time. Either the chance of success is not being accurately displayed or the random number generator isn't too good. They really ought to run a chi-square on it to see.
Gentlemen! You can't fight in here!! This is the War Room!

Cordelia
Citoyen
Posts: 19
Joined: Fri 05 Feb 2016, 04:40
Realm: Fauconnia

Re: Random

Postby Cordelia » Sat 03 Feb 2018, 01:03

Balok wrote:I hunted something 24 times at a 50% chance. Got 7

24 independent trials at 50% are supposed to result in 7 or fewer successes 3.2% of the time. So while this run of results is unlucky, it is still more common than rolling double-six at Monopoly.

Tuf
Bâtisseur
Posts: 130
Joined: Thu 12 Jan 2017, 21:59
Realm: -

Re: Random

Postby Tuf » Sat 03 Feb 2018, 08:04

The best way is not care about such results.
U only mention them because they happened. It is common Survivorship bias https://en.wikipedia.org/wiki/Survivorship_bias
I don't care about this things - u will get result earlier or later but will get it anyway.
e.g. i gather Gala from November nonstop 5 chars. And i can get from 90 to 200 ore per day (it is average from 5 chars - remember) So random processes are always random.

User avatar
Balok
Bâtisseur
Posts: 140
Joined: Wed 06 Jan 2016, 16:45
Realm: Fauconnia

Re: Random

Postby Balok » Sat 03 Feb 2018, 16:19

Cordelia wrote:24 independent trials at 50% are supposed to result in 7 or fewer successes 3.2% of the time. So while this run of results is unlucky, it is still more common than rolling double-six at Monopoly.

Sure. But we know they're bad at rounding (use different methods different places) because aknoth's attempts to simulate crafting revealed this. If they can't do rounding math consistently, they might be inconsistent at other kinds of math, too. Hence my suggestion that they run a test on their random number generator to see how accurately it simulates randomness.

They won't, of course, because they're too busy and because there's no benefit for them in doing so. Nor do I really expect them to. I would just find the results interesting, academically.
Gentlemen! You can't fight in here!! This is the War Room!

Gopol
Mystique
Posts: 781
Joined: Wed 06 Jan 2016, 16:42
Realm: Agonedus

Re: Random

Postby Gopol » Mon 05 Feb 2018, 16:43

We know how the rounding for the crafting is done. And indeed it is a bit weird. As far as I know, I was the first one to detail how it works. (See previous posts on universal formula and on roundings if there are still around in the forums)
It cannot be fully properly simulated because the rounding in the game depends on how exactly the sub results are represented in floating point. Well, you could simulate it but that would require much much more knowledge about the exact coding and full specs of the mathematical libraries used in the game : not worth the effort. We can however check which crafts *might* have a rounding issue which is precisely what I do in my own simulators.

They won't change it however because that would change the stats of existing objects...

For harvesting, the impact of rounding should be very limited. The harvest rate displayed on screen is rounded whereas the true rate isn't but that is only a 0.5% difference.

Cyram
Emperator
Posts: 41
Joined: Wed 06 Jan 2016, 16:02
Realm: Filgar

Re: Random

Postby Cyram » Thu 08 Feb 2018, 16:16

Tiny Hippo wrote:Yes, this problem exist. When you use acceleration, the probability of skill growth, even with a high chance, is much lower. I think this is a problem on the server side. With acceleration, many problems.


If that's actually the case, than that's an issue that's a lot worse than just a streaky random number generator, because people would be getting a lot less than they are led to believe for their money.

User avatar
Cloudigger
Mystique
Posts: 1491
Joined: Fri 20 Nov 2015, 20:43
Realm: -

Re: Random

Postby Cloudigger » Thu 08 Feb 2018, 17:04

The up possibilities are calculated on the fixed time in design and not on the duration in second of the skill. So reducing the asked times in second does not influence the probabilities of upping.


To be clear, there is no way that the random is affected by any acceleration.

Cordelia
Citoyen
Posts: 19
Joined: Fri 05 Feb 2016, 04:40
Realm: Fauconnia

Re: Random

Postby Cordelia » Thu 08 Feb 2018, 23:00

Cloudigger wrote:To be clear, there is no way that the random is affected by any acceleration.

...provided that the pseudo-random number generator behaves the same way when you take a large number of samples on a single schedule, compared to taking them at 2 hour intervals.

As an aside, I worked on a rogue-like game in the mid 1990s. Replacing the inbuilt system random() with a 'better' PRNG dramatically altered the look and feel of the procedurally-generated room layouts. We hadn't changed any of the probabilities we were testing against, but the behaviour of the program was materially changed. It was surprising.

Tiny Hippo
Bourlingueur
Posts: 99
Joined: Wed 24 May 2017, 10:15
Realm: Tortolion

Re: Random

Postby Tiny Hippo » Fri 09 Feb 2018, 03:53

Cloudigger wrote:
To be clear, there is no way that the random is affected by any acceleration.

You talk about formulas and theory, I'm talking about practice. About the fact that the acceleration affects the chances of increasing skills, said two independent observers.
Another example. My character performs the task in his house at the plains of hawkoria. 15 minutes left before completion. I add to the tasks to move in guild domain which's placed at the whitestone fields (+1.5 min). After that I do not want to wait and I press the "speed up" button. According to your formulas and theories, my character should be directly in the domain, because 17 minutes less than 20, but in practice my character is at the point from which he is still moving for 5 seconds toward the domain. checkmate ;)
I'm not rude to you, I'm just hurtful
:lol:

User avatar
nobukita
Citoyen
Posts: 24
Joined: Fri 06 May 2016, 19:38
Realm: Alchimideon

Re: Random

Postby nobukita » Fri 09 Feb 2018, 05:21

I can read English, but for sure, I can't write it good enough, so I asked Google to translate this (see spoiler)

Quelques remarques
- vous vouliez une vérification, c'est fait : le générateur de hasard va bien et l'accélération n'est pas lié à lui
- pensez-vous que les joueurs "d'avant" étaient plus bêtes que vous et n'ont rien remarqué si le problème est là depuis toujours ? Ou que les devs ont tripatouillé le code ces derniers temps juste pour vous ennuyez ? Dans quel but ?
- tu parles de ton expérience personnelle, la mienne ne montre aucun lien "pratique" entre les deux (et j'accélère souvent les crafts que je fais pour up : je veux pouvoir faire le craft qui nécessite ces ups !)

Mais surtout qu'espérez-vous (tous) de plus dans cette discussion ? Que les devs vous donnent des up parce que vous avez l'impression que le jeu n'est pas gentil ? Qu'ils vérifient une fois de plus ce qui a déjà été vérifié ?

(*) Devs, on est combien de joueurs actifs en ce moment ? vous avez on a une idée du nombre de crafts fait par jour ? du nombre de crafts accélérés ?


English version, by Google (hem)
A few comments
- you wanted a check, it's done: the random generator is fine and the acceleration is not linked to him
- do you think that the players "from before" were more stupid than you and did not notice anything if the problem is always there? Or that Devs have tinkered with the code recently just to annoy you? What purpose ?
- you talk about your personal experience, mine shows no "practical" link between the two (I often accelerate the crafts that I do for up: I want to be able to do the craft that requires these ups!)

But above all, what more do you expect in this discussion? That Devs give you ups because you feel that the game is not nice? That they check once more what has already been checked?

(*) Devs, how many players are active right now? do you have an idea of ​​the number of crafts made per day? the number of accelerated crafts?

Tiny Hippo
Bourlingueur
Posts: 99
Joined: Wed 24 May 2017, 10:15
Realm: Tortolion

Re: Random

Postby Tiny Hippo » Fri 09 Feb 2018, 20:21

nobukita wrote:A few comments
- you wanted a check, it's done: the random generator is fine and the acceleration is not linked to him
- do you think that the players "from before" were more stupid than you and did not notice anything if the problem is always there? Or that Devs have tinkered with the code recently just to annoy you? What purpose ?
- you talk about your personal experience, mine shows no "practical" link between the two (I often accelerate the crafts that I do for up: I want to be able to do the craft that requires these ups!)

- one big lol. You need to re-read again slowly and thoughtfully. after using it, something happens (probably at server side) which is affects skills up. One more reason to make fun of you, it's cause you simply ignored part with the described bug with moving and acceleration
- I'm player from before.
- https://en.wiktionary.org/wiki/practice go read no.7

Behind your words is emptiness. Trying to disprove me with no arguments, in a fit of a maternal sacrifice full of love for devs. It looks pathetic :roll:
I'm not rude to you, I'm just hurtful
:lol:

User avatar
Céline
Mystique
Posts: 2286
Joined: Fri 20 Nov 2015, 20:42
Realm: -

Re: Random

Postby Céline » Fri 09 Feb 2018, 20:55

Keep calm.
Take a deap breath.

Or I'll lock this topic.
---
~Contact : support@virtys.com~


Return to “General”

Who is online

Users browsing this forum: No registered users and 1 guest