Jump to content


WG and their strange RNG calculation


  • Please log in to reply
28 replies to this topic

Maschinenkanone #1 Posted 10 January 2018 - 08:43 AM

    Sergeant

  • Player
  • 35471 battles
  • 236
  • [MBEV] MBEV
  • Member since:
    05-17-2012

Just an anectode which should make us think:

 

 

Recently I tried to craft the tree topper for Modern, tier4. There are two possible tree toppers so the chance to create a specific one is 50%.

After 11 (eleven!) tries and investing around 5000 shards I gave up. All eleven tries created the same tree topper (the one I already have).


 

This means:

a) I have been incredible unlucky and encountered a next to zero probability.

or b) there is a bug in WGs RNG calculation


 


Edited by Maschinenkanone, 10 January 2018 - 08:44 AM.


_EXODUZ_ #2 Posted 10 January 2018 - 08:49 AM

    Lieutenant

  • Player
  • 34635 battles
  • 1,936
  • [RGT] RGT
  • Member since:
    11-05-2014

I vote for c) you're not very smart.

My worst streak was 4. I suggest re-taking Statistics or Probability Theory in school.



Phobos4321 #3 Posted 10 January 2018 - 08:49 AM

    General

  • Player
  • 43112 battles
  • 8,241
  • Member since:
    09-27-2011

if you talk about RNG normally everyone would thing about the battle related RNG that would be used for penetration and damage values

 

let me ask you was there any playce they stated  that there was a RNG in place for loot even more something with equal distribution ?



Homer_J #4 Posted 10 January 2018 - 08:51 AM

    Field Marshal

  • Beta Tester
  • 27652 battles
  • 29,000
  • [WJDE] WJDE
  • Member since:
    09-03-2010

View PostMaschinenkanone, on 10 January 2018 - 07:43 AM, said:

 the chance to create a specific one is 50%.

 

What makes you think that?

 

Some items could be more rare than others, just because there are two possible outcomes does not mean it's a 50:50 chance.  One could be set to only appear 10% of the time and the other 90%.

 

Even if it was 50:50 though you could still get the same one 11 times in a row.


Edited by Homer_J, 10 January 2018 - 08:52 AM.


Maschinenkanone #5 Posted 10 January 2018 - 09:01 AM

    Sergeant

  • Player
  • 35471 battles
  • 236
  • [MBEV] MBEV
  • Member since:
    05-17-2012

View Post_EXODUZ_, on 10 January 2018 - 07:49 AM, said:

I vote for c) you're not very smart.

My worst streak was 4. I suggest re-taking Statistics or Probability Theory in school.

 

Well, if you had indeed some knowledge about stochastics you would know how low a "50%" chance truely is with 10 preconditions.

Edited by Maschinenkanone, 10 January 2018 - 09:08 AM.


Maschinenkanone #6 Posted 10 January 2018 - 09:03 AM

    Sergeant

  • Player
  • 35471 battles
  • 236
  • [MBEV] MBEV
  • Member since:
    05-17-2012

View PostHomer_J, on 10 January 2018 - 07:51 AM, said:

What makes you think that?

 

Some items could be more rare than others, just because there are two possible outcomes does not mean it's a 50:50 chance.  One could be set to only appear 10% of the time and the other 90%.

 

Even if it was 50:50 though you could still get the same one 11 times in a row.

 

I agree, this could indeed be so. But isnt it rational to consider all items within one class and tier to be of the same probability until other information are available?

Edited by Maschinenkanone, 10 January 2018 - 09:05 AM.


HassenderZerhacker #7 Posted 10 January 2018 - 09:53 AM

    Captain

  • Player
  • 25612 battles
  • 2,265
  • [1DPG] 1DPG
  • Member since:
    09-09-2015

View PostMaschinenkanone, on 10 January 2018 - 09:03 AM, said:

 

I agree, this could indeed be so. But isnt it rational to consider all items within one class and tier to be of the same probability until other information are available?

 

yeah, but you left one obvious information without consideration: programmed by WG.

 



Aikl #8 Posted 10 January 2018 - 10:18 AM

    Brigadier

  • Player
  • 25167 battles
  • 4,253
  • [-MM] -MM
  • Member since:
    04-13-2011
You can probably slaughter the decorations on the tree and whatnot on the last day if you need another shot, at least if you've spent your rewards.

Balc0ra #9 Posted 10 January 2018 - 10:23 AM

    Field Marshal

  • Player
  • 64467 battles
  • 15,475
  • [WALL] WALL
  • Member since:
    07-10-2012
50% is still 50%. Just like you can lose 11 games in a row that xvm said you had a 50% chance to win. As it don't make it 60% or 90% on the next one. It's still 50% to get one of two. And when it's 50% flat, there is still a % chance of you getting the wrong one 100 times in a row, just like there is a % chance of you to get the right one on the first one.

KLguns #10 Posted 10 January 2018 - 10:36 AM

    Captain

  • Player
  • 46878 battles
  • 2,012
  • [T-D-U] T-D-U
  • Member since:
    08-07-2012
They want you to buy more loot boxes, and still not get the right ones...

trispect #11 Posted 10 January 2018 - 10:37 AM

    Lieutenant

  • Beta Tester
  • 33277 battles
  • 1,765
  • [HSOP] HSOP
  • Member since:
    01-16-2011

View PostBalc0ra, on 10 January 2018 - 11:23 AM, said:

50% is still 50%. Just like you can lose 11 games in a row that xvm said you had a 50% chance to win. As it don't make it 60% or 90% on the next one. It's still 50% to get one of two. And when it's 50% flat, there is still a % chance of you getting the wrong one 100 times in a row, just like there is a % chance of you to get the right one on the first one.

 

Odds for getting 11 same 50/50 rolls in a row is 0,0488% or 1:2058

 

...no I don’t think the odds are the same with all decorations.



TankkiPoju #12 Posted 10 January 2018 - 10:44 AM

    Lieutenant General

  • Player
  • 20014 battles
  • 6,192
  • Member since:
    05-20-2011

Statistical analysis and probality is hard.

 

For example, 99.95% of people think WoT has a linear 25% RNG. In reality it's not linear 25% RNG but a bell curve, meaning most rolls will be within 10% middle segment. Which means for example min and max rolls damage rolls are quite unlikely.



StinkyStonky #13 Posted 10 January 2018 - 10:44 AM

    Lieutenant

  • Player
  • 26982 battles
  • 1,921
  • [H5VOC] H5VOC
  • Member since:
    11-02-2015

View PostMaschinenkanone, on 10 January 2018 - 07:43 AM, said:

WG and their strange RNG calculation

 

View PostHomer_J, on 10 January 2018 - 07:51 AM, said:

Some items could be more rare than others, just because there are two possible outcomes does not mean it's a 50:50 chance.  One could be set to only appear 10% of the time and the other 90%.

That would be an example of the "Strangeness" that the OP suggests.  It would also be extraordinarily deceptive of WG as there has been no mention of rarity within a tier and the implication is that the "rarity" is paid for by paying 1440 to craft tier Vs rather than 60 to craft tier Is.

 

View PostHomer_J, on 10 January 2018 - 07:51 AM, said:

Even if it was 50:50 though you could still get the same one 11 times in a row.

Yes, there would be a 1 in 2024 chance.

 

If WG had any sense they would ensure that it didn't happen.

 



StinkyStonky #14 Posted 10 January 2018 - 10:48 AM

    Lieutenant

  • Player
  • 26982 battles
  • 1,921
  • [H5VOC] H5VOC
  • Member since:
    11-02-2015

View PostTankkiPoju, on 10 January 2018 - 09:44 AM, said:

Statistical analysis and probality is hard.

 

For example, 99.95% of people think WoT ...

 

Did you know that 47.3% of statistics are just made up ? :P

 

Your point on linear RNG is spot on, But ... that characteristic has been explained by WG (and read by 18.4% of players).

WGs only hint on rarity of decorations is that tier 5 are rarer that 4, which are rarer than 3, which ....



HassenderZerhacker #15 Posted 10 January 2018 - 12:35 PM

    Captain

  • Player
  • 25612 battles
  • 2,265
  • [1DPG] 1DPG
  • Member since:
    09-09-2015

View PostTankkiPoju, on 10 January 2018 - 10:44 AM, said:

Statistical analysis and probality is hard.

 

For example, 99.95% of people think WoT has a linear 25% RNG. In reality it's not linear 25% RNG but a bell curve, meaning most rolls will be within 10% middle segment. Which means for example min and max rolls damage rolls are quite unlikely.

 

and add to that, the skewness of the curve might be different depending on the tank....

HassenderZerhacker #16 Posted 10 January 2018 - 12:39 PM

    Captain

  • Player
  • 25612 battles
  • 2,265
  • [1DPG] 1DPG
  • Member since:
    09-09-2015

View PostBalc0ra, on 10 January 2018 - 10:23 AM, said:

50% is still 50%. Just like you can lose 11 games in a row that xvm said you had a 50% chance to win. As it don't make it 60% or 90% on the next one. It's still 50% to get one of two. And when it's 50% flat, there is still a % chance of you getting the wrong one 100 times in a row, just like there is a % chance of you to get the right one on the first one.

 

yes, correct.

and for ten tries there is a chance of 0.097% that it happened in this case.

 

it would be interesting to compile a list of other cases, i.e. how many tries until the right deco is made



eldrak #17 Posted 10 January 2018 - 12:43 PM

    Second Lieutenant

  • Player
  • 46578 battles
  • 1,007
  • [NE-VO] NE-VO
  • Member since:
    01-27-2011

If I remember to do it I can try some crafting tonight to check outcomes.

I'm fairly certain odds for each item is not the same.



Homer_J #18 Posted 10 January 2018 - 12:57 PM

    Field Marshal

  • Beta Tester
  • 27652 battles
  • 29,000
  • [WJDE] WJDE
  • Member since:
    09-03-2010

View PostMaschinenkanone, on 10 January 2018 - 08:03 AM, said:

 

I agree, this could indeed be so. But isnt it rational to consider all items within one class and tier to be of the same probability until other information are available?

Other information is available - you are dealing with WG.  That's all the information you need.

 

View PostStinkyStonky, on 10 January 2018 - 09:44 AM, said:

 

 It would also be extraordinarily deceptive of WG a

 

Unless WG specifically say there is a 50:50 chance then I would assume otherwise.



sifis_21 #19 Posted 10 January 2018 - 01:30 PM

    Corporal

  • Player
  • 32470 battles
  • 172
  • [HS_D] HS_D
  • Member since:
    04-13-2013

View PostTankkiPoju, on 10 January 2018 - 10:44 AM, said:

Statistical analysis and probality is hard.

 

For example, 99.95% of people think WoT has a linear 25% RNG. In reality it's not linear 25% RNG but a bell curve, meaning most rolls will be within 10% middle segment. Which means for example min and max rolls damage rolls are quite unlikely.

 

it was a bell curve for aim. then wg saw that guns were too accurate and people could aim for weakspot so decided to change it (make a hole in the middle of the bell) so you have to use gold to pen anything with armor

Edited by sifis_21, 10 January 2018 - 01:30 PM.


sifis_21 #20 Posted 10 January 2018 - 01:38 PM

    Corporal

  • Player
  • 32470 battles
  • 172
  • [HS_D] HS_D
  • Member since:
    04-13-2013

View PostMaschinenkanone, on 10 January 2018 - 08:43 AM, said:

Just an anectode which should make us think:

 

 

Recently I tried to craft the tree topper for Modern, tier4. There are two possible tree toppers so the chance to create a specific one is 50%.

After 11 (eleven!) tries and investing around 5000 shards I gave up. All eleven tries created the same tree topper (the one I already have).


 

This means:

a) I have been incredible unlucky and encountered a next to zero probability.

or b) there is a bug in WGs RNG calculation


 

 

or c) wg made this missions as most of them in last 3 or more years so they can earn a Christmas bonus, they dont give a sh:t about game community. missions deigned so you get close to end but its impossible to finish unless you pay. Thats why you dont see any marathons any more without "luck" factor. cause every rng factor made by wg scammers is rigged.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users