Jump to content


RNG. How is it decided?


  • Please log in to reply
83 replies to this topic

FritzSpiegal #1 Posted 03 May 2020 - 05:11 PM

    Corporal

  • Player
  • 26651 battles
  • 156
  • Member since:
    10-13-2016

I wonder how exactly RNG is calculated.

 

Is it per shot, per person for the whole game or per team for the whole game?

 

Most of us have been at the wrong end of 15-1 losses (and of course had 15-1 wins too) yet mostly, we have the same ability (or lack of) every time we play.

 

Recently, Claus Kellerman did a quick analysis of his teams drubbings and found that, even though both teams had similar levels of skilled players, the losing team in each case had only10-20% penetrations whereas the winning teams were 70-80% yet both had similar total shots. In that case, it would appear that RNG was team generated for that battle.

 

It might be interesting if this were done over a larger sample as Claus only did it for three games but, on his admittedly limited number, it would seem win or lose might be decided before we even start to play. Clearly too, in some cases, a really skilled player can over come this (though  that doesn't apply to me!)

 

Of course, some players can achieve 60% + wins regardless but even they must have these whitewash games now & then?


Edited by FritzSpiegal, 03 May 2020 - 05:26 PM.


Homer_J #2 Posted 03 May 2020 - 05:12 PM

    Field Marshal

  • Moderator
  • 33872 battles
  • 38,229
  • [WJDE] WJDE
  • Member since:
    09-03-2010
It's per shot.

jabster #3 Posted 03 May 2020 - 05:14 PM

    Field Marshal

  • Beta Tester
  • 12848 battles
  • 27,702
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View PostFritzSpiegal, on 03 May 2020 - 04:11 PM, said:

I wonder how exactly RNG is calculated.

 

Is it per shot, per person for the whole game or per team for the whole game?

 

Most of us have been at the wrong end of 15-1 losses (and of course had 15-1 wins too) yet mostly, we have the same ability (or lack of) every time we play.

 

Recently, Claus Kellerman did a quick analysis of his teams drubbings and found that, even though both teams had similar levels of skilled players, the losing team in each case had only10-20% penentrations whereas the winning teams were 70-80% yet both had similar total shots. In that case, it would appear that RNG was team generated for that battle.

 

It might be interesting if this were done over a larger sample as Claus only did it for three games but, on his admittedly limited number, it would seem win or lose might be decided before we even start to play. Clearly too, in some cases, a really skilled player can over come this (though  that doesn't apply to me!)

 

Of course, some players can achieve 60% + wins regardless but even they must have these whitewash games now & then?


Now there’s your problem, you’re listening to Kellerman.



wEight_Tanker #4 Posted 03 May 2020 - 05:16 PM

    Second Lieutenant

  • Player
  • 13049 battles
  • 1,479
  • [D_L_Z] D_L_Z
  • Member since:
    11-13-2016

View Postjabster, on 03 May 2020 - 05:14 PM, said:


Now there’s your problem, you’re listening to Kellerman.

 

rng = wallet.open() ? hit : miss



tank276 #5 Posted 03 May 2020 - 05:22 PM

    Warrant Officer

  • Player
  • 56297 battles
  • 970
  • [KEKE] KEKE
  • Member since:
    08-30-2012

View Postjabster, on 03 May 2020 - 04:14 PM, said:


Now there’s your problem, you’re listening to Kellerman.


Lol, I really like Claus .

He is not the best player out there ofc and , lately, he has become a conspiracy theorist of sorts.



NUKLEAR_SLUG #6 Posted 03 May 2020 - 05:25 PM

    Lieutenant General

  • Player
  • 36972 battles
  • 6,119
  • [FISHY] FISHY
  • Member since:
    06-13-2015
Klaus is utterly clueless. Do yourself a favour and stop listening to his gibberish. 

jabster #7 Posted 03 May 2020 - 05:27 PM

    Field Marshal

  • Beta Tester
  • 12848 battles
  • 27,702
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View Posttank276, on 03 May 2020 - 04:22 PM, said:


Lol, I really like Claus .

He is not the best player out there ofc and , lately, he has become a conspiracy theorist of sorts.


Of sorts?



Balc0ra #8 Posted 03 May 2020 - 05:38 PM

    Field Marshal

  • Player
  • 77299 battles
  • 23,316
  • [WALL] WALL
  • Member since:
    07-10-2012

View PostFritzSpiegal, on 03 May 2020 - 05:11 PM, said:

Recently, Claus Kellerman did a quick analysis...

 

I wanted to stop reading there, as he does that once a week, with the smallest sample size and calls if solid proof... Because that's his audience. Tinfoils. It still doesn't prove that it's per game vs per shot. As there are too many variables on each of the 30 guns each game. Because he has no clue what the other 29 did that game. Nor does everyone have the same armor, same gun, fought at the same range, or did see flat armor all the time to even have a chance to get even results vs others every game. If they did, he might have been on to something. But they did not do that, did they?

 

Even with equally skilled players, if one team had a 50 100 that went the medium lane, and the other team had a Vk 100.01 that went to face a stock T29 and a Kv-1S. What team do you think will get more pens? The only way to prove it would be to have two teams of the same tanks and armor, aiming at the exact same spots with the same ammo at the same range and the exact same impact angle. If you did see 80% vs 20% more of them then sure... But in randoms? Not really.

 

View PostFritzSpiegal, on 03 May 2020 - 05:11 PM, said:

 Of course, some players can achieve 60% + wins regardless but even they must have these whitewash games now & then?

 

Sure. Just as 45% players can get something done right now and them and win vs good players. As again, the variables before RNG were more correct, or incorrect. I mean even Circon have games where he yells at his gun vs the team.

 

We can use the M&M example again. As if you have a bowl of red and blue M&M's. Red is low rolls, blue are high rolls, and you pick them blind. At times you will blind pick more red then blue. That's when Circon yells at his gun. Just as you will get the odd round of more blue M&M's. Chances are still there for that to happen. Even with no rigging.

 


Edited by Balc0ra, 03 May 2020 - 06:00 PM.


gunslingerXXX #9 Posted 03 May 2020 - 06:17 PM

    Brigadier

  • Player
  • 14791 battles
  • 4,254
  • [GUNSL] GUNSL
  • Member since:
    11-16-2014

Rigged number generator is based on players behavior (e.g. playing a stock tank) and game conditions (e.g. a x4 weekend). Sometimes the whole team is affected to ensure you will hardly get any wins the first 10 or so games. Keep on playing and you get the improved RNG to stomp other players. Also this is to make it less obvious and make it more difficult to prove its rigged. 

 

Spoiler

 



Gkirmathal #10 Posted 03 May 2020 - 06:26 PM

    Lieutenant

  • Player
  • 8692 battles
  • 1,877
  • [2VTD] 2VTD
  • Member since:
    01-14-2013

View PostFritzSpiegal, on 03 May 2020 - 04:11 PM, said:

I wonder how exactly RNG is calculated.

 

Is it per shot, per person for the whole game or per team for the whole game?

 

From one of the last Q&A's a dev explained that RNG values are predetermined by the server, it keeps a generating these numbers and puts them in a database.

These are used in battle and randomly picked.

 

The theory is that each entry in the db corresponds to a set of RNG values, determining: shot distribution inside aim circle, pen rng multiplier and dmg rng multiplier (and probabky some more things)

In a battle when you fire a shot, the server randomly picks an entry from the list and applies it on the shot.

 

What is also still unclear to me is if the systems works from the main constantly generated RNG db, or wether each battle gets assinged their own list populated with RNG sets from the main db.

If the latter then the question becomes does each individual team get their own assigned with RNG sets. Cause that could explain why some teams, like fully red ones, magically outperform an enemy that has much better players.

One team got lucky and got a better sets of RNG then the other team.

 

Dunnow.



jabster #11 Posted 03 May 2020 - 06:32 PM

    Field Marshal

  • Beta Tester
  • 12848 battles
  • 27,702
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View PostGkirmathal, on 03 May 2020 - 05:26 PM, said:

 

From one of the last Q&A's a dev explained that RNG values are predetermined by the server, it keeps a generating these numbers and puts them in a database.

These are used in battle and randomly picked.

 

The theory is that each entry in the db corresponds to a set of RNG values, determining: shot distribution inside aim circle, pen rng multiplier and dmg rng multiplier (and probabky some more things)

In a battle when you fire a shot, the server randomly picks an entry from the list and applies it on the shot.

 

What is also still unclear to me is if the systems works from the main constantly generated RNG db, or wether each battle gets assinged their own list populated with RNG sets from the main db.

If the latter then the question becomes does each individual team get their own assigned with RNG sets. Cause that could explain why some teams, like fully red ones, magically outperform an enemy that has much better players.

One team got lucky and got a better sets of RNG then the other team.

 

Dunnow.


A team can get a bad ‘set’ of RNG for a battle regardless as to whether the are pre-generated or dynamically generated when required. Both scenarios are as random as each other.


Edited by jabster, 03 May 2020 - 06:36 PM.


Slyspy #12 Posted 03 May 2020 - 06:46 PM

    Field Marshal

  • Player
  • 15327 battles
  • 17,984
  • [T-D-U] T-D-U
  • Member since:
    12-07-2011

View PostGkirmathal, on 03 May 2020 - 06:26 PM, said:

 

From one of the last Q&A's a dev explained that RNG values are predetermined by the server, it keeps a generating these numbers and puts them in a database.

These are used in battle and randomly picked.

 

The theory is that each entry in the db corresponds to a set of RNG values, determining: shot distribution inside aim circle, pen rng multiplier and dmg rng multiplier (and probabky some more things)

In a battle when you fire a shot, the server randomly picks an entry from the list and applies it on the shot.

 

What is also still unclear to me is if the systems works from the main constantly generated RNG db, or wether each battle gets assinged their own list populated with RNG sets from the main db.

If the latter then the question becomes does each individual team get their own assigned with RNG sets. Cause that could explain why some teams, like fully red ones, magically outperform an enemy that has much better players.

One team got lucky and got a better sets of RNG then the other team.

 

Dunnow.

 

Even if they had someone rolling dice that could still happen so I'm not sure what difference it makes. 



qpranger #13 Posted 03 May 2020 - 06:54 PM

    Lieutenant General

  • Player
  • 38889 battles
  • 6,442
  • [HAMMY] HAMMY
  • Member since:
    12-25-2013
RNG is calculated per rigged shot.

malachi6 #14 Posted 03 May 2020 - 06:56 PM

    Brigadier

  • Player
  • 51526 battles
  • 4,597
  • Member since:
    04-14-2011

View PostgunslingerXXX, on 03 May 2020 - 06:17 PM, said:

Rigged number generator is based on players behavior (e.g. playing a stock tank) and game conditions (e.g. a x4 weekend). Sometimes the whole team is affected to ensure you will hardly get any wins the first 10 or so games. Keep on playing and you get the improved RNG to stomp other players. Also this is to make it less obvious and make it more difficult to prove its rigged. 

 

Spoiler

 

 

Thus, you cannot prove it is rigged, yet it is rigged?  Paradox or irony?  There is a very little used answer to such questions: "I don't know.  I need more evidence before I can draw a valid conclusion".



Gkirmathal #15 Posted 03 May 2020 - 07:07 PM

    Lieutenant

  • Player
  • 8692 battles
  • 1,877
  • [2VTD] 2VTD
  • Member since:
    01-14-2013

View Postjabster, on 03 May 2020 - 05:32 PM, said:


A team can get a bad ‘set’ of RNG for a battle regardless as to whether the are pre-generated or dynamically generated when required. Both scenarios are as random as each other.

View PostSlyspy, on 03 May 2020 - 05:46 PM, said:

 

Even if they had someone rolling dice that could still happen so I'm not sure what difference it makes. 

 

First since that Q&A from a few days ago we now know WoT uses of predetermined rng numbers.

Secondly yes of course it could makes 'a' difference, as a better argument against some rng claims. Like that of Claus.

 

That is sort of the question I asked in the second half of my post. Might not have stated it all to clearly.

Is it 2 pools 1 battle (don't mind the an...terminology)...or is it 1 pool for both teams per battle.

 

On that I would really like to get an answer, as it would partially explain some rng related observations, on which all those idiotic tinfoil hat theories are based.

 

To know this can act as a better and more understandable counter then the current arguments used IMO. But he that just my logic and I know I'm a bit of am odd ball ;)

 


Edited by Gkirmathal, 03 May 2020 - 07:08 PM.


LethalWalou #16 Posted 03 May 2020 - 07:14 PM

    Lieutenant Сolonel

  • Player
  • 35446 battles
  • 3,232
  • [OISPA] OISPA
  • Member since:
    09-17-2012

View Postjabster, on 03 May 2020 - 04:27 PM, said:

View Posttank276, on 03 May 2020 - 04:22 PM, said:


Lol, I really like Claus .

He is not the best player out there ofc and , lately, he has become a conspiracy theorist of sorts.

Of sorts?

 

And lately?



NUKLEAR_SLUG #17 Posted 03 May 2020 - 07:19 PM

    Lieutenant General

  • Player
  • 36972 battles
  • 6,119
  • [FISHY] FISHY
  • Member since:
    06-13-2015

View PostGkirmathal, on 03 May 2020 - 07:07 PM, said:

Secondly yes of course it could makes 'a' difference, as a better argument against some rng claims. Like that of Claus.

 

No it doesn't make a difference. It's the same thing no matter how you do it. 

 

Try it yourself if you want proof. Roll some dice 10 times and write the result down on a sheet of paper. That's your pregenerated RNG list. Now roll your dice 10 times again and that's your 'live'. RNG.

 

Now which of those is 'more' random than the other and why? 



shikaka9 #18 Posted 03 May 2020 - 07:22 PM

    Captain

  • Player
  • 78498 battles
  • 2,404
  • [-SJA-] -SJA-
  • Member since:
    02-27-2013
RNG = Really Nothing Generated

jabster #19 Posted 03 May 2020 - 07:44 PM

    Field Marshal

  • Beta Tester
  • 12848 battles
  • 27,702
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View PostGkirmathal, on 03 May 2020 - 06:07 PM, said:

 

First since that Q&A from a few days ago we now know WoT uses of predetermined rng numbers.

Secondly yes of course it could makes 'a' difference, as a better argument against some rng claims. Like that of Claus.

 

 

No it doesn't make a difference to how random they values are. Two different methods, I roll a dies ten times and note down the sequence. You ask me for the roll of a die. I take the first in the list and tell you. You ask me again, I take the second in the list and so on. Now repeat it except when you ask for the roll of a die, I roll it and tell you.

 

Why would either method be different in terms of randomness?

 

Edit: Note to self, don’t start a post, go and do something else and then comeback to finish it. Someone may have already made the point in the meantime.

 


Edited by jabster, 03 May 2020 - 08:24 PM.


wsatnutter #20 Posted 03 May 2020 - 07:53 PM

    Field Marshal

  • Beta Tester
  • 27425 battles
  • 10,265
  • [WSAT] WSAT
  • Member since:
    08-25-2010
such tinfoilery




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users