Jump to content


Random battles are not random!


  • Please log in to reply
291 replies to this topic

KurtKnispel___ #1 Posted 12 October 2017 - 10:56 PM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

Hi guys, something about matchmaker. The random battles, though being declared and named as "random", are not really random and I can prove that.

Today (12.10.2017) I played 17 random battles in a row with two vehicles (Emil II and Strv 103-0) trying to get daily double experience. I was not able to achieve one single win. I was not the cause of the losses (no team damage etc., about 1500 average damage to enemy tanks, which is not super-amazing, but still fair enough to show that my influence on the battles was positive for my team, not negative).

If random battles were really random, an average player’s chance to win has to be about 50% each battle. My long-term stats and result show that I am at least average (actually somewhat better than average player, my long term win rate is about 54%), but OK, let’s not complicate it and say that I have basically only 50% chance.

What is the probability of 17 losses in a row, where chance in each battle is statistically 50%? Easy to calculate, it’s a binomical distribution of n independent experiments, each asking a yes–no question, and each with its own boolean-valued outcome. The result for 17 experiments (battles) is:

= 0,0007629% chance that 17 battles will all result as loss with 50% chance in each single battle.

That calculated to human perspective means that if I play 17 battles every day, I need statistically 131072 days to get one result like this (= 17 losses in a row). That is 359 years of playing WOT every day!

For the details of the calculation please refer to https://en.wikipedia.org/wiki/Binomial_distribution.

Yes, you may say it’s VERY IMPROBABLE, but still possible, it’s just a coincidence and I have solid chance that it does not happen again (or at least probably not in the nearest 300 years), but I do not accept this explanation, because I had several similar improbable loosing streaks in the recent months. If we all admit that mathematics as we know it, is generally valid in our universe, we must see that something strange happens here – is it our universe changing and our mathematic rules are to be not valid any longer?

Let’s forget about theories of changing universe. I consider proven that the explanation lies in the Wargaming’s servers. What happens is that mm logic secretly balances the two teams and takes one of the better players from random created teams and puts him to the worst to equalize chances of both teams. Of course this “feature” may be nice for the beginners to help them get wins easier, but guess how frustrating it is for the “balancing” player to be constantly in defeating teams. The most important outcome is that WOT’s main game mode, RANDOM BATTLES, ARE NOT REALY RANDOM (!) and WARGAMING IS HIDING THIS INFORMATION FROM PUBLIC (!).



Lil_Dimitry #2 Posted 12 October 2017 - 10:58 PM

    Sergeant

  • Player
  • 2318 battles
  • 209
  • Member since:
    12-07-2014
https://www.youtube.com/watch?v=9Deg7VrpHbM

wsatnutter #3 Posted 12 October 2017 - 11:01 PM

    Lieutenant General

  • Beta Tester
  • 24781 battles
  • 7,456
  • [WSAT] WSAT
  • Member since:
    08-25-2010

huge wall of poo poo and this is what happens with poo poo

 



Aikl #4 Posted 12 October 2017 - 11:04 PM

    Brigadier

  • Player
  • 25284 battles
  • 4,349
  • [-MM] -MM
  • Member since:
    04-13-2011

Now that you've calculated the probability of a single player getting this result, what is the chance of one player overall having seventeen losses in a row one day? Point being that if you e.g. gave every man in Germany a coin and made them flip it e.g. twenty times a day and to call you if they got seventeen tails in a row, chances are you'd be busy writing stuff down.

http://www.drdobbs.c...-odds/229300217

 

Statistics are weird stuff, but you're not particularly special - besides, unless Wargaming/CIA/KGB is targeting only you (in which case we wouldn't give a damn), both why and how could they rig games like that? Doesn't make any sense. I prefer the simpler explanation. Occam's razor and all that jazz.


Edited by Aikl, 12 October 2017 - 11:08 PM.


Homer_J #5 Posted 12 October 2017 - 11:05 PM

    Field Marshal

  • Beta Tester
  • 28291 battles
  • 29,578
  • [WJDE] WJDE
  • Member since:
    09-03-2010

View PostKurtKnispel___, on 12 October 2017 - 10:56 PM, said:

 What happens is that mm logic secretly balances the two teams and takes one of the better players from random created teams and puts him to the worst to equalize chances of both teams. 

 

So, if it's rigged then why did they rig it for you to win more than you lost?



K_A #6 Posted 12 October 2017 - 11:14 PM

    Brigadier

  • Player
  • 13641 battles
  • 4,665
  • [T-D-U] T-D-U
  • Member since:
    04-18-2013
The odds of winning the Powerball lottery in USA is 0,0000000034% and yet people win it.

KurtKnispel___ #7 Posted 12 October 2017 - 11:14 PM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostAikl, on 12 October 2017 - 10:04 PM, said:

Now that you've calculated the probability of a single player getting this result, what is the chance of one player overall having seventeen losses in a row one day? Point being that if you e.g. gave every man in Germany a coin and made them flip it e.g. twenty times a day and to call you if they got seventeen tails in a row, chances are you'd be busy writing stuff down.

 

Statistics are weird stuff, but you're not particularly special - besides, unless Wargaming/CIA/KGB is targeting only you (in which case we wouldn't give a damn), both why and how could they rig games like that? Doesn't make any sense. I prefer the simpler explanation. Occam's razor and all that jazz.

 

I see your point. I may be just the one "lucky bastard". But as I wrote, this happened to me not first time (well, not for 17 cases, but streak of 9 losses I had recently, which is again statisticaly quite improbable). So, why would wargaming do this stuff? Obvious, they need to attract new players. If they "throw them into water" between all the sealclubbers with no mm optimization, no new players (payers) come and stay in the game. That's why.

 

And to your example: the exact statistical probabilty is that 1 of 1 048 576 people does it, so some 80 people would daily win it in Germany if all germans do that. But the probabilty that one person wins it multiple times?.. Yes, I can calculate that, but lets just say it's very improbable.. :-)

 

View PostHomer_J, on 12 October 2017 - 10:05 PM, said:

 

So, if it's rigged then why did they rig it for you to win more than you lost?

 

I declare that they changed behaviour of mm in the recent months. Before random mm was random, imhb.
 

Captain_Kremen0 #8 Posted 12 October 2017 - 11:20 PM

    Second Lieutenant

  • Player
  • 37041 battles
  • 1,353
  • Member since:
    06-04-2011

I was not the cause of the losses.

Arsene Wengers WoT account



MattieW #9 Posted 12 October 2017 - 11:21 PM

    Sergeant

  • Player
  • 13483 battles
  • 217
  • [SA_HB] SA_HB
  • Member since:
    10-11-2013

The last week I've really been struggling with wins with no idea what's changed.  Team mates have been playing beyond stupid, like not even bot level of activity.

Then after about 7 losses in a row I'll be on a team that rolls through destroying the enemy, knife through butter for maybe 5 games in a row.

So no idea what's happening in the MM but short of sticking to tier 5 and 6 going forward I hope something in the next patch fixes it.

Tin foil hat off.



KurtKnispel___ #10 Posted 12 October 2017 - 11:24 PM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostCaptain_Kremen0, on 12 October 2017 - 10:20 PM, said:

I was not the cause of the losses.

Arsene Wengers WoT account

 

I know it sounds bad.. :-) But even if I try to loose.. cant make it , 17 losses in a row, just try it ... :-)

Captain_Kremen0 #11 Posted 12 October 2017 - 11:28 PM

    Second Lieutenant

  • Player
  • 37041 battles
  • 1,353
  • Member since:
    06-04-2011
I give up after 5

TungstenHitman #12 Posted 12 October 2017 - 11:39 PM

    Brigadier

  • Player
  • 21504 battles
  • 4,006
  • Member since:
    08-28-2016

View Postwsatnutter, on 12 October 2017 - 10:01 PM, said:

huge wall of poo poo and this is what happens with poo poo

 

 

Do you know that the OP is Polish? As in that place from which was born and owes both heritage and lineage from? and yet it is also the same designation one who speaks of English language for which shares that of a viscous product that shines ones shoes all nice and shiny so that one can both carry a certain classiness and reflection of self esteem before others while also carrying the added bonus of carrying two highly air coefficient weapons in which to insert ones foot without much air nor indeed trouser resistance, all the way up the desired personages rectum all the way to the ankle without the slightest resistance...

Of course such matters rely heavy on both the application of shoe Polish, the quality of both shoe and shoe Polish, the weight of leg, power of muscle tissue, wind resistance, altitude and just how clenched and aware that persons rear end is to receive that Polished to perfection, mirrored leather shoe traveling at near mach2 towards the nether regions



Mimos_A #13 Posted 12 October 2017 - 11:45 PM

    Lieutenant

  • Player
  • 23518 battles
  • 1,950
  • [QSF-L] QSF-L
  • Member since:
    05-30-2015

You seem to discard a number of things:

 

While your overall wr is 54%, your average tier played is 5.96. If you look at your high tier tanks your winrate drops off quite a bit and is rather spotty, so it's possible you're struggling more to get the same amount of wins at higher tiers.

 

If you look at the Emil II, you also didn't do well winrate wise in the Emil I, so maybe you just didn't figure out how to win in those tanks yet. Same goes for the STRV and the UDES.

 

It's easy to discard loss streaks as bad luck, and some of them definitely are. But if you're honest to yourself you'll often find you're just not playing a tank right, making the wrong plays at the wrong time and getting tilt and making even more mistakes. Especially if you're focusing on daily doubles 3 tiers above your average tier.

 

I had a pretty awful streak yesterday and while the teams definitely didn't help, I found myself just getting caught out of position, taking bad shots, not anticipating well enough, getting near tilt, not reading the teamlist properly, forgetting about an unspotted tank etc etc.


Edited by Mimos_A, 12 October 2017 - 11:52 PM.


Gremlin182 #14 Posted 12 October 2017 - 11:50 PM

    General

  • Player
  • 47939 battles
  • 8,064
  • Member since:
    04-18-2012

17 games is far too small a sample to prove anything and if you had won all 17 of those games would you still have posted.

Has anyone in the history of wot come here and complained because they won too many games and this proved random games were not random.

 

I had a teacher once ask me to generate a page of random numbers on the computer then he complained because sometimes the same number came up in sequence ie 12345556789 He claimed they were not random.

 

 



KurtKnispel___ #15 Posted 12 October 2017 - 11:55 PM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostMimos_A, on 12 October 2017 - 10:45 PM, said:

You seem to discard a number of things:

 

While your overall wr is 54%, your average tier played is 5.96. If you look at your high tier tanks your winrate drops off quite a bit and is rather spotty, so it's possible you're struggling more to get the same amount of wins at higher tiers.

 

If you look at the Emil II, you also didn't do well winrate wise in the Emil I, so maybe you just didn't figure out how to win in those tanks yet. Same goes for the STRV and the UDES.

 

It's easy to discard loss streaks as bad luck, and some of them definitely are. But if you're honest to yourself you'll often find you're just not playing a tank right, making the wrong plays at the wrong time and getting tilt and making even more mistakes. Especially if you're focusing on daily doubles 3 tiers above your average tier.

 

I had a pretty awful streak yesterday and while the teams definitely didn't help, I found myself just getting caught out of position, taking bad shots, not anticipating well enough, getting near tilt, not reading the teamlist properly, forgetting about an unspotted tank etc etc.

 

I partly agree with you. Even though that before today, my win rate on Strv was 58% and on Emil II 57%... But I agree I cant play these tanks well, I have that feeling about it. But does it explain 17 losses in a row? Even if I try to do losses, I cant do 17 in a row purposefully... 

Mimos_A #16 Posted 12 October 2017 - 11:58 PM

    Lieutenant

  • Player
  • 23518 battles
  • 1,950
  • [QSF-L] QSF-L
  • Member since:
    05-30-2015

View PostKurtKnispel___, on 12 October 2017 - 11:55 PM, said:

 

I partly agree with you. Even though that before today, my win rate on Strv was 58% and on Emil II 57%... But I agree I cant play these tanks well, I have that feeling about it. But does it explain 17 losses in a row? Even if I try to do losses, I cant do 17 in a row purposefully... 

 

Why wouldn't it be possible?

 

If you don't play well and support well it only takes a few others to mess up to lose a game, especially at tier 9 and 10 you often get punished pretty brutally for not contributing enough.



KurtKnispel___ #17 Posted 12 October 2017 - 11:59 PM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostGremlin182, on 12 October 2017 - 10:50 PM, said:

17 games is far too small a sample to prove anything and if you had won all 17 of those games would you still have posted.

Has anyone in the history of wot come here and complained because they won too many games and this proved random games were not random.

 

I had a teacher once ask me to generate a page of random numbers on the computer then he complained because sometimes the same number came up in sequence ie 12345556789 He claimed they were not random.

 

 

 

Go in casino a guess if you can see a roulette serie of 17x in a row falling red, for example. That would start a panic that the roulette is broken...
 

View PostMimos_A, on 12 October 2017 - 10:58 PM, said:

 

Why wouldn't it be possible?

 

If you don't play well and support well it only takes a few others to mess up to lose a game, especially at tier 9 and 10 you often get punished pretty brutally for not contributing enough.

 

Well, I was between top three players in most of these games.. Cant prove that, if you dont want to believe me, you just dont, np.
 

Mimos_A #18 Posted 13 October 2017 - 12:06 AM

    Lieutenant

  • Player
  • 23518 battles
  • 1,950
  • [QSF-L] QSF-L
  • Member since:
    05-30-2015

View PostKurtKnispel___, on 13 October 2017 - 12:01 AM, said:

 

Well, I was between top three players in most of these games.. Cant prove that, if you dont want to believe me, you just dont, np.

 

Being in the top three doesn't mean you did what was needed to win. Especially in the strv I can say from experience that it's pretty damn easy to throw games by being too passive. You'll end up top of damage easy in that thing, doesn't mean you did your damage at the right time.

 

I'm not saying you played bad or are a bad player, but seeing your performance in the tanks leading up to these ones and the fact that you were playing way above your average tier I think some self reflection might be in order and  I don't see how this "proves" that randoms are not random.



KurtKnispel___ #19 Posted 13 October 2017 - 12:17 AM

    Private

  • Player
  • 41031 battles
  • 32
  • [PRDEC] PRDEC
  • Member since:
    03-26-2014

View PostMimos_A, on 12 October 2017 - 11:06 PM, said:

 

Being in the top three doesn't mean you did what was needed to win. Especially in the strv I can say from experience that it's pretty damn easy to throw games by being too passive. You'll end up top of damage easy in that thing, doesn't mean you did your damage at the right time.

 

I'm not saying you played bad or are a bad player, but seeing your performance in the tanks leading up to these ones and the fact that you were playing way above your average tier I think some self reflection might be in order and  I don't see how this "proves" that randoms are not random.

 

I understand your point. I admit I am just average player in high tiers (and I wrote in my article, that calculation is done for average player with 50% win chance). I agree especially with that about Strv and not-winning damage. But still.. could you just flip coin and get 17x times one side? Do you realy think it can all be explained with my self reflection, that I just didnt help team enough?

Edited by KurtKnispel___, 13 October 2017 - 12:21 AM.


Mimos_A #20 Posted 13 October 2017 - 12:23 AM

    Lieutenant

  • Player
  • 23518 battles
  • 1,950
  • [QSF-L] QSF-L
  • Member since:
    05-30-2015

View PostKurtKnispel___, on 13 October 2017 - 12:17 AM, said:

 

I understand your point. I admit I am just average player in high tiers. I agree especially with that about Strv and not-winning damage. But still.. could you just flip coin and get 17x times one side? Do you realy think it can all be explained with my self reflection, that I just didnt help team enough?

 

Not all 17, you might have just had crap teams for 10 of them and then neglected to tip the balance in 7 others. And even if that wasn't the case, if you flip a coin 33345 times, yes, it is totally possible that it will land on the same side 17 times in a row.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users