Jump to content


Official Matchmaking Discussion Thread


  • Please log in to reply
1787 replies to this topic

GaijinLOver #1521 Posted 11 December 2020 - 10:24 PM

    Junior Sergeant

  • Players
  • 2558 battles
  • 143
  • [GRWO]
  • Member since:
    08-16-2020
This game's nub matchmaking is killing my IS-3..  Whenever matchmaking is on my side I kill so many on IS-3 , and whenever matchmaking is against me I die in 20 seconds. Its gotten to the point where I don't even know if I am a nub or a pro. Thx a lot wg :arta:

 

 


_Crusader6_ #1522 Posted 13 December 2020 - 07:16 PM

    Please don’t go to town.

  • Players
  • 78470 battles
  • 18,641
  • [III-P]
  • Member since:
    12-08-2014

View PostGaijinLOver, on 11 December 2020 - 05:24 PM, said:

This game's nub matchmaking is killing my IS-3..  Whenever matchmaking is on my side I kill so many on IS-3 , and whenever matchmaking is against me I die in 20 seconds. Its gotten to the point where I don't even know if I am a nub or a pro. Thx a lot wg :arta:


Let’s look at this objectively:  

 1) You have a .5-.6 Damage Ratio Meaning you take 2x the damage you deal.  
 2) You don’t spot much at all 

 3) Your survival rate is terrible. 
 4) Your average damage / tier is 127, a treading water DR is around 200 per tier. 
 

 

 

 


I hate Annihilator spammers...  

 
Tank Hoarder: 463 tanks in Garage:  481/481 Played Tanks Aced
 
 
I need more tanks...
 
    Wallet Warrior: Loyal Original M60 owner
 

 


Bellatormonk #1523 Posted 14 December 2020 - 04:59 PM

    Tinfoil Knight

  • Players
  • 38482 battles
  • 1,803
  • [MMMM]
  • Member since:
    04-12-2011

So did the patch break the MM even more than usual?  Preface.  10.5k people online currently.  T7 match green except me, less 400 battles for every player on green, not re-rolls.  Red, every player is 1k-115k battles (yes...one red player had 115k battles).  WR not withstanding in this instance, I am talking about the 5k rule were told about and have talked about for the better part of 1-2 years?

Obviously this MM mechanic is not longer in effect.  Can you seriously tell me that it was "random", other than working as intended to place a player with 30k battles on a team where everyone is 400 battles or less against a full red team of 1k+ battles players?

Out of 10.5k players there was not a "normal" match of 14 5k+ battles players to be joined?

 

Edit01-So I am trying to play it out to 10 battles at least today in T7 just to see how much of a fluke this is and so far it's no fluke or random.  I continue to get greens with players at 500 or less battles and of course they know nothing of the game/meta, where to go etc.  Accepting that this is the reality for right now as opposed to just the usual "bad" players, I thought maybe a new batch of Switch players came in or something and I set my Control Mode to "Same" (I am Steam/PC) for only the 5th time since I have played from Beta even.

 

Turns out...I am getting the same player types.  These are seemingly new PC players even though I don't see anything that would drawing droves of new PC players in Steam.  I mean there is what seems like a fairly decent $20 pack that contains and FCM/30 Days Premium and 1 million silver for a new player....but still..that's not a massive draw is it?

 

Edit02-So I ran 10 in a row and after that first 3 losses in a row with nothing but new players on the teams..a "switch" flipped and I got the regular crowd back of recognizable clans/names and won the next 7.  I mean it was just so odd to get those 1st 3 teams in a row like..that..then..boom..regular playerbase for the next 7?


Edited by Bellatormonk, 14 December 2020 - 07:57 PM.


IronBehemoth #1524 Posted 15 December 2020 - 09:53 PM

    Official Society Game Play Publicist.

  • Players
  • 14089 battles
  • 598
  • [-EQX-]
  • Member since:
    01-17-2015

View PostBellatormonk, on 14 December 2020 - 04:59 PM, said:

So did the patch break the MM even more than usual?  Preface.  10.5k people online currently.  T7 match green except me, less 400 battles for every player on green, not re-rolls.  Red, every player is 1k-115k battles (yes...one red player had 115k battles).  WR not withstanding in this instance, I am talking about the 5k rule were told about and have talked about for the better part of 1-2 years?

Obviously this MM mechanic is not longer in effect.  Can you seriously tell me that it was "random", other than working as intended to place a player with 30k battles on a team where everyone is 400 battles or less against a full red team of 1k+ battles players?

Out of 10.5k players there was not a "normal" match of 14 5k+ battles players to be joined?

 

Edit01-So I am trying to play it out to 10 battles at least today in T7 just to see how much of a fluke this is and so far it's no fluke or random.  I continue to get greens with players at 500 or less battles and of course they know nothing of the game/meta, where to go etc.  Accepting that this is the reality for right now as opposed to just the usual "bad" players, I thought maybe a new batch of Switch players came in or something and I set my Control Mode to "Same" (I am Steam/PC) for only the 5th time since I have played from Beta even.

 

Turns out...I am getting the same player types.  These are seemingly new PC players even though I don't see anything that would drawing droves of new PC players in Steam.  I mean there is what seems like a fairly decent $20 pack that contains and FCM/30 Days Premium and 1 million silver for a new player....but still..that's not a massive draw is it?

 

Edit02-So I ran 10 in a row and after that first 3 losses in a row with nothing but new players on the teams..a "switch" flipped and I got the regular crowd back of recognizable clans/names and won the next 7.  I mean it was just so odd to get those 1st 3 teams in a row like..that..then..boom..regular playerbase for the next 7?

Your daily winrate is 65% your 30d is 62% and your 90d is 59% so I'm not sure what your complaint is



_Wikipedia #1525 Posted 15 December 2020 - 10:11 PM

    Official advocate for 40%ers

  • Players
  • 15291 battles
  • 807
  • [GUPPY]
  • Member since:
    10-28-2018

View PostGaijinLOver, on 11 December 2020 - 04:24 PM, said:

This game's nub matchmaking is killing my IS-3..  Whenever matchmaking is on my side I kill so many on IS-3 , and whenever matchmaking is against me I die in 20 seconds. Its gotten to the point where I don't even know if I am a nub or a pro. Thx a lot wg

Thats because you are a utter bot! How the hell can you get away with 685 damage in ALL of your tanks. Yes even in your IS-3 where your alpha is 400. This is the first time I seen a 30%er schmuck cry about ANYTHING when it’s usually your fault, and your 30 day aside from WN8 is stagnant. If you die in 20 seconds, THATS ON YOU, UTILIZE YOUR INVINCIBLE TURRET AND COVER! DON’T OVEREXPOSE LIKE A CLOWN! We all love to give Wargaming some hate but do so with logic. NOW BEFORE YOU CONTINUE YOUR MAD BECAUSE BAD RANT...

 

 

https://www.blitzstars.com/player/com/GaijinLOver

 


The legend himself: https://www.twitch.tv/dsteves

Do not have an average tier of 6-8 if you want to cry.


BrassGun #1526 Posted 17 December 2020 - 03:46 PM

    Senior Sergeant

  • Players
  • 6818 battles
  • 869
  • [NT123]
  • Member since:
    03-27-2016
Am I the only one who's noticed that there's something bizarre about how the game raises and lowers winrate? I've noticed, over and over again, that at times a single win does not raise your winrate. Instead, you need to get two in a row to push it up, whereas a loss, any loss, will always lower it. This makes raising winrate a massive pain, especially if you get either a trade off of wins and losses.

"If the tanks succeed, then victory follows."

-Heinz Guderian.

"Your task will not be an easy one. Your enemy is well trained, well equipped and battle-hardened. He will fight savagely."

-Dwight D. Eisenhower


_Crusader6_ #1527 Posted 17 December 2020 - 06:03 PM

    Please don’t go to town.

  • Players
  • 78470 battles
  • 18,641
  • [III-P]
  • Member since:
    12-08-2014

View PostBrassGun, on 17 December 2020 - 10:46 AM, said:

Am I the only one who's noticed that there's something bizarre about how the game raises and lowers winrate? I've noticed, over and over again, that at times a single win does not raise your winrate. Instead, you need to get two in a row to push it up, whereas a loss, any loss, will always lower it. This makes raising winrate a massive pain, especially if you get either a trade off of wins and losses.


Math? 
 

WR is a Percentage of Wins in Battles played.  
   To move WR up, it takes more wins the more battles you have.   
 

 


I hate Annihilator spammers...  

 
Tank Hoarder: 463 tanks in Garage:  481/481 Played Tanks Aced
 
 
I need more tanks...
 
    Wallet Warrior: Loyal Original M60 owner
 

 


j_rod #1528 Posted 17 December 2020 - 07:55 PM

    Quit being a Karen!

  • Players
  • 33588 battles
  • 4,206
  • [III-C]
  • Member since:
    05-04-2011

View Post_Crusader6_, on 17 December 2020 - 12:03 PM, said:


Math? 
 

WR is a Percentage of Wins in Battles played.  
   To move WR up, it takes more wins the more battles you have.   
 

 

 

Maybe you should explain how decimal points and rounding work....


Edited by j_rod, 17 December 2020 - 11:50 PM.

Eventually given an infinite number of typewriters and an infinite number of Monkeys, they will come up with the complete works of Shakespeare. - C6 articulating the Infinite Monkey Theorem
It was never fair... but it was fun! - Krietenstein34 

 


BrassGun #1529 Posted 17 December 2020 - 11:42 PM

    Senior Sergeant

  • Players
  • 6818 battles
  • 869
  • [NT123]
  • Member since:
    03-27-2016

View Post_Crusader6_, on 17 December 2020 - 06:03 PM, said:


Math? 
 

WR is a Percentage of Wins in Battles played.  
   To move WR up, it takes more wins the more battles you have.   
 

 

 

View Postj_rod, on 17 December 2020 - 07:55 PM, said:

 

Maybe you should explain how decimal plays and rounding work....

 

Yeah, I get the feeling I'll need this.

 

 

 

 

"If the tanks succeed, then victory follows."

-Heinz Guderian.

"Your task will not be an easy one. Your enemy is well trained, well equipped and battle-hardened. He will fight savagely."

-Dwight D. Eisenhower


j_rod #1530 Posted 17 December 2020 - 11:55 PM

    Quit being a Karen!

  • Players
  • 33588 battles
  • 4,206
  • [III-C]
  • Member since:
    05-04-2011

View PostBrassGun, on 17 December 2020 - 05:42 PM, said:

Yeah, I get the feeling I'll need this.


Not sure if you’re being sarcastic, but what we’re saying is that your winrate is moving every time you play, it’s just that it’s reflected as a percentage to the second decimal point and the more games one has under one’s belt, the less likely it is to move the winrate as we see it in game.

 

Think about it this way - if you’ve played 10,000 games and won 5,000, you’re at 50.00% right? If you win one more, you’re at 50.005% right? If that number is rounded up, you’re at 50.01% and if it’s rounded down, you’re still at 50.00%


Eventually given an infinite number of typewriters and an infinite number of Monkeys, they will come up with the complete works of Shakespeare. - C6 articulating the Infinite Monkey Theorem
It was never fair... but it was fun! - Krietenstein34 

 


BrassGun #1531 Posted 17 December 2020 - 11:58 PM

    Senior Sergeant

  • Players
  • 6818 battles
  • 869
  • [NT123]
  • Member since:
    03-27-2016

View Postj_rod, on 17 December 2020 - 11:55 PM, said:


Not sure if you’re being sarcastic, but what we’re saying is that your winrate is moving every time you play, it’s just that it’s reflected as a percentage to the second decimal point and the more games one has under one’s belt, the less likely it is to move the winrate as we see it in game.

 

Think about it this way - if you’ve played 10,000 games and won 5,000, you’re at 50.00% right? If you win one more, you’re at 50.005% right? If that number is rounded up, you’re at 50.01% and if it’s rounded down, you’re still at 50.00%

 

Okay, thanks for that bit of an explanation.


"If the tanks succeed, then victory follows."

-Heinz Guderian.

"Your task will not be an easy one. Your enemy is well trained, well equipped and battle-hardened. He will fight savagely."

-Dwight D. Eisenhower


RussianMiracle #1532 Posted 27 December 2020 - 02:53 PM

    Private

  • Players
  • 294 battles
  • 1
  • [MORTM]
  • Member since:
    02-09-2020
This thread never disappoints

_Baroness #1533 Posted 28 December 2020 - 03:02 PM

    Left the forums

  • Players
  • 16604 battles
  • 422
  • [NHP]
  • Member since:
    10-17-2018

I posted this in another thread but thought I'd drop it here:

 

 

I think they are reading too much into a random system.  Some of this is just because you tend to notice the losses more than the wins (there is a name for this someone help me) the rest is just what's called gambler's fallacy.

The key here is believing you draw a certain team based on what has happened previously.  But that's not the case.  It may be that certain times the server may produce certain results, but it has nothing to do with your win rate.

You can see this using the coin toss example:

 

If a fair coin is flipped 21 times, the probability of 21 heads is 1 in 2,097,152. The probability of flipping a head after having already flipped 20 heads in a row is 50%. 
 

Each time you press battle, MM only looks at what it has.  Not where you have been.


http://www.blitzstars.com/player/com/_Baroness

Well, I supposed that's it for me.  It's been a fun ride and I truly appreciate all the tanking advice I've been given.  I'll be taking my Dr. Seuss books and Mr. Potato head and moving on to other pastures.  I have not tried the discord, but I think I'll give that a go next.  I wish everyone here the best and perhaps I'll see you on the battlefield.  o7


Bellatormonk #1534 Posted 28 December 2020 - 04:41 PM

    Tinfoil Knight

  • Players
  • 38482 battles
  • 1,803
  • [MMMM]
  • Member since:
    04-12-2011

View Post_Baroness, on 28 December 2020 - 03:02 PM, said:

I posted this in another thread but thought I'd drop it here:

 

 

I think they are reading too much into a random system.  Some of this is just because you tend to notice the losses more than the wins (there is a name for this someone help me) the rest is just what's called gambler's fallacy.

The key here is believing you draw a certain team based on what has happened previously.  But that's not the case.  It may be that certain times the server may produce certain results, but it has nothing to do with your win rate.

You can see this using the coin toss example:

 

If a fair coin is flipped 21 times, the probability of 21 heads is 1 in 2,097,152. The probability of flipping a head after having already flipped 20 heads in a row is 50%. 
 

Each time you press battle, MM only looks at what it has.  Not where you have been.

Got proof?



j_rod #1535 Posted 28 December 2020 - 05:00 PM

    Quit being a Karen!

  • Players
  • 33588 battles
  • 4,206
  • [III-C]
  • Member since:
    05-04-2011

View Post_Baroness, on 28 December 2020 - 09:02 AM, said:

I posted this in another thread but thought I'd drop it here:

 

 

I think they are reading too much into a random system.  Some of this is just because you tend to notice the losses more than the wins (there is a name for this someone help me) the rest is just what's called gambler's fallacy.

The key here is believing you draw a certain team based on what has happened previously.  But that's not the case.  It may be that certain times the server may produce certain results, but it has nothing to do with your win rate.

You can see this using the coin toss example:

 

If a fair coin is flipped 21 times, the probability of 21 heads is 1 in 2,097,152. The probability of flipping a head after having already flipped 20 heads in a row is 50%. 
 

Each time you press battle, MM only looks at what it has.  Not where you have been.

 

This is true IF it’s known that there’s a 50% chance of an event happening. For high versus low tiers, that percentage is most certainly not 50% contingent on what tier one is playing and when they are playing.

 

When we say “random” MM, we’re really just saying that MM isn’t taking into consideration any factors other than what has been stated - time in queue, tier (+1/-1) and tank class. There is no evidence at this point in time that MM is weighting teams by skill (winrate, avg damage, whatever) or any other factor, but I think that most of us are open to the possibility of that happening. We just want to see evidence if someone says it is, through the form of a decent sample size of data that can be analyzed. 


Eventually given an infinite number of typewriters and an infinite number of Monkeys, they will come up with the complete works of Shakespeare. - C6 articulating the Infinite Monkey Theorem
It was never fair... but it was fun! - Krietenstein34 

 


_Baroness #1536 Posted 28 December 2020 - 05:08 PM

    Left the forums

  • Players
  • 16604 battles
  • 422
  • [NHP]
  • Member since:
    10-17-2018

View PostBellatormonk, on 28 December 2020 - 11:41 AM, said:

Got proof?

It's just math.  But helps to show your point of view. 

 

Assume we did flip a coin and it landed on heads 20 times in a row.  We are going to flip it one more time.  Which of the following is you:

 

1. There's no way that coin will land heads again.  The chances have to be very low because it just happened 20 times.

2. 50/50 it could be heads.

 

If you are number 1 then you probably do think the system is rigged against you.  And it probably seems that way.

 

View Postj_rod, on 28 December 2020 - 12:00 PM, said:

 

This is true IF it’s known that there’s a 50% chance of an event happening. For high versus low tiers, that percentage is most certainly not 50% contingent on what tier one is playing and when they are playing.

 

When we say “random” MM, we’re really just saying that MM isn’t taking into consideration any factors other than what has been stated - time in queue, tier (+1/-1) and tank class. There is no evidence at this point in time that MM is weighting teams by skill (winrate, avg damage, whatever) or any other factor, but I think that most of us are open to the possibility of that happening. We just want to see evidence if someone says it is, through the form of a decent sample size of data that can be analyzed. 

 

Yeah you said it much better than me.  Agree.

 

My only point is that MM only looks at what it has right now.  Not what has been happening to any of the players.  The system simply looks at what is in que and puts teams together based on tiers, tanks, and battle preference.


http://www.blitzstars.com/player/com/_Baroness

Well, I supposed that's it for me.  It's been a fun ride and I truly appreciate all the tanking advice I've been given.  I'll be taking my Dr. Seuss books and Mr. Potato head and moving on to other pastures.  I have not tried the discord, but I think I'll give that a go next.  I wish everyone here the best and perhaps I'll see you on the battlefield.  o7


Bellatormonk #1537 Posted 29 December 2020 - 03:00 PM

    Tinfoil Knight

  • Players
  • 38482 battles
  • 1,803
  • [MMMM]
  • Member since:
    04-12-2011

View Post_Baroness, on 28 December 2020 - 05:08 PM, said:

It's just math.  But helps to show your point of view. 

 

Assume we did flip a coin and it landed on heads 20 times in a row.  We are going to flip it one more time.  Which of the following is you:

 

1. There's no way that coin will land heads again.  The chances have to be very low because it just happened 20 times.

2. 50/50 it could be heads.

 

If you are number 1 then you probably do think the system is rigged against you.  And it probably seems that way.

 

 

Yeah you said it much better than me.  Agree.

 

My only point is that MM only looks at what it has right now.  Not what has been happening to any of the players.  The system simply looks at what is in que and puts teams together based on tiers, tanks, and battle preference.

 

Not proof about Math and coins tosses...proof about the MM not looking at previous matches, etc.  Your statement is no better than people who say it does look at something like previous matches, wins, or whatever.  Neither person has proof, only speculation and a belief in what the Devs claim yet have never provided actual information on.



_Crusader6_ #1538 Posted 29 December 2020 - 04:48 PM

    Please don’t go to town.

  • Players
  • 78470 battles
  • 18,641
  • [III-P]
  • Member since:
    12-08-2014

View PostBellatormonk, on 29 December 2020 - 10:00 AM, said:

 

Not proof about Math and coins tosses...proof about the MM not looking at previous matches, etc.  Your statement is no better than people who say it does look at something like previous matches, wins, or whatever.  Neither person has proof, only speculation and a belief in what the Devs claim yet have never provided actual information on.

WG has said the only think MM looks at for previous matches is maps.  
  The idea behind that is so that it doesn’t keep dumping you on the same maps.  


I hate Annihilator spammers...  

 
Tank Hoarder: 463 tanks in Garage:  481/481 Played Tanks Aced
 
 
I need more tanks...
 
    Wallet Warrior: Loyal Original M60 owner
 

 


_Baroness #1539 Posted 29 December 2020 - 05:00 PM

    Left the forums

  • Players
  • 16604 battles
  • 422
  • [NHP]
  • Member since:
    10-17-2018

View PostBellatormonk, on 29 December 2020 - 10:00 AM, said:

 

Not proof about Math and coins tosses...proof about the MM not looking at previous matches, etc.  Your statement is no better than people who say it does look at something like previous matches, wins, or whatever.  Neither person has proof, only speculation and a belief in what the Devs claim yet have never provided actual information on.

 

Ah, I must have misunderstood your question.  That kind of proof would divulge proprietary information and I'm sure WG would not be on board with that.  I suppose in the meantime we can only speculate about how a system works.  I speculate that there is much less to it than it seems.  While I'm sure there is a thought process to help engage players more in the game (less skilled win more, make it harder for higher skilled, etc.), I don't think there is enough player base on the NA servers to make this work.  There are already enough known rules in place.

 

They do have job openings in relevant departments in NA.  Perhaps a mole could be inserted....:trollface:


http://www.blitzstars.com/player/com/_Baroness

Well, I supposed that's it for me.  It's been a fun ride and I truly appreciate all the tanking advice I've been given.  I'll be taking my Dr. Seuss books and Mr. Potato head and moving on to other pastures.  I have not tried the discord, but I think I'll give that a go next.  I wish everyone here the best and perhaps I'll see you on the battlefield.  o7


dangerousdan26 #1540 Posted 31 December 2020 - 01:18 AM

    Lance-corporal

  • Players
  • 37548 battles
  • 66
  • [5TH-D]
  • Member since:
    06-14-2018
The matchmaker sucks

Edited by dangerousdan26, 31 December 2020 - 02:01 AM.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users