Jump to content


Random battles are not random!


  • Please log in to reply
291 replies to this topic

KurtKnispel___ #41 Posted 13 October 2017 - 08:36 AM

    Private

  • Player
  • 41770 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostK_A, on 13 October 2017 - 07:17 AM, said:

 

The chance of first flip being heads is 50%. The chance of second flip being heads is 50%. The chance of third flip being heads is 50%. The chance of fourth flip being heads is 50%. The chance of fifth flip being heads is 50%. I can go on.

 

The chance of getting HHHHHHHHHH is the same as the chance of getting HHTTHTHHHT, and also the same as getting HTHTHTHTHT, and also the same as getting TTTTTHTTTT. Every single combination has the same percentage chance of happening, and it is down to pure random chance which combination you will end up with for any given number of battles over any given time period.

 

Now, in WoT that chance is influenced slightly either to the positive or negative by your own actions, which alone makes your calculations of chance obsolete, and also makes any attempt of "rigging" highly impractical and useless.

 

Excactly, the probability of each single try is always 50%. But I am not talking about this. I talk about probability of getting the same (or whatever but predicted) results in all tries! That is getting lower with increasing number of tries. 

 

The probability of HHTTHTHHHT is exactly the same as HHHHHHHHHH, but its significantly lower then probability of HHT or HHH.

 

Common, this is elementary statistics of first semester at bachelor studies..

 

View PostMerton15, on 13 October 2017 - 07:18 AM, said:

Congratulations, you have provided empirical evidence that MM is indeed random.

 

No that does not proove it. But I admit that my conclusion can not be proven as well. It happens statisticaly exactly one times of 130000 series of 17 rolls (= means you need to play 2 210 000 battles to meet STATISTICALY once such result (that all 17 tries are loss), but okay, whatever it is unprobable, it does not definitely proove my therory, I may be just the unlucky winner... :-)
 

Edited by VMX, 17 October 2017 - 04:10 PM.


Jigabachi #42 Posted 13 October 2017 - 08:50 AM

    Field Marshal

  • Player
  • 17946 battles
  • 19,311
  • [T-D-U] T-D-U
  • Member since:
    08-12-2011

View PostKurtKnispel___, on 13 October 2017 - 08:00 AM, said:

With 17 tries you have only 0,000762939453125% to get exactly 17 eagles.

But playing WoT isn't a coinflip, so your probability there is wrong.


Edited by Jigabachi, 13 October 2017 - 08:50 AM.


KurtKnispel___ #43 Posted 13 October 2017 - 08:55 AM

    Private

  • Player
  • 41770 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostJigabachi, on 13 October 2017 - 07:50 AM, said:

But playing WoT isn't a coinflip, so your probability there is wrong.

 

It basicaly is - if MM is random. Because your influence on team is just 1/15, you must be extremely good to make your statistical chance much higher than 50% (60% or more) or extremely bad (doing teamdamage or being often AFK) to have chances much lower than 50% (40% or less). My longterm win-rate is 54% so talking about 50% chance is not so far from reality (if I calculated with 54%, the probability of getting 17 losses of 17 plays would be even lower).

cragarion #44 Posted 13 October 2017 - 08:56 AM

    Major

  • Beta Tester
  • 44285 battles
  • 2,598
  • Member since:
    07-10-2010

Random is not 50% it is random, that's how you can get the same map 4 times in a row then 5 different maps straight after it is random,

you could have numbers 1 to 10 and pick 1 out at a time randomly and believe it or not you could pick the same number 100 times or you may never get the same number twice in a row.

OP i suggest you go research exactly what random means.


Edited by cragarion, 13 October 2017 - 08:58 AM.


KurtKnispel___ #45 Posted 13 October 2017 - 08:59 AM

    Private

  • Player
  • 41770 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View Postcragarion, on 13 October 2017 - 07:56 AM, said:

Random is not 50% it is random, that's how you can get the same map 4 times in a row then 5 different maps straight after it is random,

OP i suggest you go research exactly what random means.

 

4 times in a row is statisticaly relatively probable, but did you get the same map 17 times in a row? Isn't that improbable? 

 

OKAY, I learn about randomness, you learn about statistics and probability.



Jigabachi #46 Posted 13 October 2017 - 09:08 AM

    Field Marshal

  • Player
  • 17946 battles
  • 19,311
  • [T-D-U] T-D-U
  • Member since:
    08-12-2011

View PostKurtKnispel___, on 13 October 2017 - 08:55 AM, said:

It basicaly is - if MM is random. 

No, it isn't. It would be if the MM was rigged, or, in other words, if there was a kind of skillbased MM that balances both teams.



cragarion #47 Posted 13 October 2017 - 09:11 AM

    Major

  • Beta Tester
  • 44285 battles
  • 2,598
  • Member since:
    07-10-2010

View PostKurtKnispel___, on 13 October 2017 - 08:59 AM, said:

 

4 times in a row is statisticaly relatively probable, but did you get the same map 17 times in a row? Isn't that improbable? 

 

OKAY, I learn about randomness, you learn about statistics and probability.

 

When you consider the millions of battles that take place every single day not totally improbable and i already know about statistics and probability,

some days you win 17 in a row and other days you lose 17 in a row,

however you have to take into consideration more than just MM,

look at weekends for example the skill level of players drops every weekend,

even though nothing changes on my game play my stats drop every single weekend and go back up again during the week, it's the same at certain times of day as well,

 

you can take 2 teams of identical tanks stick 15 fair players in 1 team and 1 good player with 14 crap players in the other team, the fair team of players will win every single time,

it has nothing to do with mm its the player base, and there is no way to add player skills to mm,

you have your stat padders that give stats a false reading,

and you have people who don't give a damn about stats who give stats a false reading, 

WG know this that's why they have never added any kind of player stats or xvm or anything to MM.

 

No matter how random something is and no matter what statistics or probabilities you take into account, there is always the one factor that can never be taken into consideration,

The Human Factor.

 



Wintermute_1 #48 Posted 13 October 2017 - 09:27 AM

    Lieutenant

  • Player
  • 44420 battles
  • 1,616
  • Member since:
    11-25-2013
I agree with you, random games aren't random but I also seriously doubt that they are rigged. And your speculative last paragraph is more about finding excuses than explanations. If a good player was switched team every game: A) The better you got the more it would effect you and it would become impossible to get beyond X winrate, but there are players with stupidly high winrate. B) You'd never see all red teams facing all Yellow/Green/Blue teams, but you do.

Merton15 #49 Posted 13 October 2017 - 09:55 AM

    Staff Sergeant

  • Player
  • 13869 battles
  • 464
  • Member since:
    04-08-2011

View PostKurtKnispel___, on 13 October 2017 - 07:49 AM, said:

 

No that does not proove it. But I admit that my conclusion can not be proven as well. It happens statisticaly exactly one times of 130000 series of 17 rolls (= means you need to play 2 210 000 battles to meet STATISTICALY once such result (that all 17 tries are loss), but okay, whatever it is unprobable, it does not definitely proove my therory, I may be just the unlucky winner... :-)

I did not say it was proven but that you had provided empirical evidence. You claimed a 17 game losing streak was proof of rigging which is untrue. I asserted that a 17 game losing streak is evidence of randomness which it is. The common give away of rigged statistics is an absence of outliers.



KurtKnispel___ #50 Posted 13 October 2017 - 09:58 AM

    Private

  • Player
  • 41770 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostMerton15, on 13 October 2017 - 08:55 AM, said:

I did not say it was proven but that you had provided empirical evidence. You claimed a 17 game losing streak was proof of rigging which is untrue. I asserted that a 17 game losing streak is evidence of randomness which it is. The common give away of rigged statistics is an absence of outliers.

 

This is good reply. I understand your point. 

Merton15 #51 Posted 13 October 2017 - 10:40 AM

    Staff Sergeant

  • Player
  • 13869 battles
  • 464
  • Member since:
    04-08-2011

View PostKurtKnispel___, on 13 October 2017 - 08:58 AM, said:

 

This is good reply. I understand your point. 

 

Thank you for your gentlemanly response

Coldspell #52 Posted 13 October 2017 - 10:55 AM

    Captain

  • Player
  • 19487 battles
  • 2,126
  • Member since:
    08-12-2013
The problem with the rigging theory is it means WG would have put time and money into developing something that won't make them money.

Long_Range_Sniper #53 Posted 13 October 2017 - 11:09 AM

    General

  • Player
  • 32961 battles
  • 9,029
  • [-MM] -MM
  • Member since:
    04-04-2011

View Postcragarion, on 13 October 2017 - 08:11 AM, said:

some days you win 17 in a row and other days you lose 17 in a row

 

 I've told you a million times not to exaggerate. 

 

View Postcragarion, on 13 October 2017 - 08:11 AM, said:

I already know about statistics and probability.

 

I'm not 100% certain this statement is correct. So you tell me the chance of getting 17 wins or losses back to back for a 50% player if you click battle now?

 

Spoiler

 

View Postcragarion, on 13 October 2017 - 08:11 AM, said:

you can take 2 teams of identical tanks stick 15 fair players in 1 team and 1 good player with 14 crap players in the other team, the fair team of players will win every single time,

it has nothing to do with mm its the player base, and there is no way to add player skills to mm,

you have your stat padders that give stats a false reading,

and you have people who don't give a damn about stats who give stats a false reading, 

WG know this that's why they have never added any kind of player stats or xvm or anything to MM.

 

So good stats are stat padders, and bad stats are people who don't give a damn about stats........mmmmmm?

 

I'm not convinced this "logic" is connected to skill based match making either. 

 


Mimos_A #54 Posted 13 October 2017 - 11:25 AM

    Captain

  • Player
  • 24562 battles
  • 2,028
  • [QSF-L] QSF-L
  • Member since:
    05-30-2015

View PostLong_Range_Sniper, on 13 October 2017 - 11:09 AM, said:

 

 

So good stats are stat padders, and bad stats are people who don't give a damn about stats........mmmmmm?

 

I'm not convinced this "logic" is connected to skill based match making either. 

 

No that's not what he says. He says there are enough people who's stats do not reflect their actual skill.



Excavatus #55 Posted 13 October 2017 - 11:44 AM

    Lieutenant

  • Player
  • 18908 battles
  • 1,656
  • [T-D-U] T-D-U
  • Member since:
    07-28-2013

View PostKurtKnispel___, on 13 October 2017 - 10:00 AM, said:

 

No it's not. Learn some basics about binomial distribution. If you flip coin 1 times, you have 50% chance getting 1 eagle. If you flip 2x times, you have 25% chance to get 2 eagles. It you flip 3 times, your chance of getting 3 eagles is 12,5% etc... The more tries you make, the less chance is that you get eagle in ALL OF THEM. With 17 tries you have only 0,000762939453125% to get exactly 17 eagles.

 

Ok I accept defeat..

There is no way to teach someone who have no idea they have it wrong...

I wrote a wall of text here.. but then.. what is the point..

 

Yes pal,

you are right..

I am wrong..

Thank you..

 


Edited by Excavatus, 13 October 2017 - 11:49 AM.


jabster #56 Posted 13 October 2017 - 12:02 PM

    Field Marshal

  • Beta Tester
  • 12555 battles
  • 23,687
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View PostKurtKnispel___, on 13 October 2017 - 07:00 AM, said:

 

No it's not. Learn some basics about binomial distribution. If you flip coin 1 times, you have 50% chance getting 1 eagle. If you flip 2x times, you have 25% chance to get 2 eagles. It you flip 3 times, your chance of getting 3 eagles is 12,5% etc... The more tries you make, the less chance is that you get eagle in ALL OF THEM. With 17 tries you have only 0,000762939453125% to get exactly 17 eagles.

 

But that’s not really the probability you’ve calculated. Firstly you have why is 17 the magic number so what you really looking at is the probability of a ‘large’ lose streak indeed if you had 17 loses with a single win in there would that still count. Next it’s over all your battles not just one day. Finally it’s the probability that any player had that loss streak not just you.

malachi6 #57 Posted 13 October 2017 - 12:06 PM

    Colonel

  • Player
  • 50211 battles
  • 3,585
  • Member since:
    04-14-2011

So the only way we can prove the MM is random is to rig it.  So you as you hit your statistical probability?

 

Irony?



KurtKnispel___ #58 Posted 13 October 2017 - 12:11 PM

    Private

  • Player
  • 41770 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View Postjabster, on 13 October 2017 - 11:02 AM, said:

 

But that’s not really the probability you’ve calculated. Firstly you have why is 17 the magic number so what you really looking at is the probability of a ‘large’ lose streak indeed if you had 17 loses with a single win in there would that still count. Next it’s over all your battles not just one day. Finally it’s the probability that any player had that loss streak not just you.

 

Nope. I had 17 losses in a row, with no win at all. All during one day. Then I gave up.

shishx_the_animal #59 Posted 13 October 2017 - 12:18 PM

    Colonel

  • WGL PRO Player
  • 31072 battles
  • 3,861
  • [MEME] MEME
  • Member since:
    04-06-2013

View PostKurtKnispel___, on 13 October 2017 - 11:11 AM, said:

 

Nope. I had 17 losses in a row, with no win at all. All during one day. Then I gave up.

 

What kind of a madman doesn't stop after 3?!

unhappy_bunny #60 Posted 13 October 2017 - 12:19 PM

    Major

  • Player
  • 18377 battles
  • 2,769
  • [-OC-] -OC-
  • Member since:
    08-01-2012

would it be fair to say that with the example of 17 wins or losses in a row, that the chances of it happening are “Possible”, but not neccessarily “Probable”?

 

To my simple mind, all this statistical evidence, is like conducting experiments in an enclosed labratory and then comparing it with what happens in the real world. Like car exhaust emmisions or car fuel comsumption. Nice results in the lab or test benches, but completely different when on the road under real life driving conditions. 

 

Or like a Lottery ticket. Any one lottery ticket has a possibility of 2 outcomes. It could win or it could not. 50% possibility, but if there are 14 million possible combinations then its probability is 1 in 14 million. That probability does not effect its possibility.

 

Mathematical stats go out the window when human beings are involved. People add their own randomness into the equation, and here we are talking about 30 human beings being involved in each equation. 






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users