Jump to content


RibbleStripe - does matchmaker analyze player skill?

MM Matchmaker

  • Please log in to reply
289 replies to this topic

minitel_NA #21 Posted 15 February 2018 - 06:39 AM

    First Sergeant

  • Players
  • 454 battles
  • 5,729
  • [EUREF]
  • Member since:
    12-19-2015

View Postreluctanttheist, on 15 February 2018 - 05:35 AM, said:

Hey Ribble

 

There's a long time argument on the forums that MM makes games more difficult for players over a number of games.  The idea behind this is that MM looks at the skill of the players on the teams and puts higher-skilled players on the opposing team to make it harder for you to win.  

 

There are other players who have read before that MM does NOT analyze player skill, but may in fact place you as low-tier to offer you a greater challenge in a certain percentage of games that you play.  In this scenario, MM remembers how often you are low tier / high tier and manipulates how often you are low or high tier.

 

Finally, there's those who believe that MM is entirely random, and that MM simply grabs you and your tank to make up the team needed at that very millisecond that MM starts building teams for the next match for which you're available.

 

Who's right, and who's buying the other guys a beer?  If you can answer this question, I will heartily thank you!

 

here is my bet : you have an oversimplified view of the situation.

 

i have personally believed that the game is rigged in some ways from even before I was 10 000 games in, That was 2014 then... and I’m 50.000 now. I have since undergone tests spanning hundreds of games. The results are very difficult to interpret, and when they are clear, they either lack good description of the mechanics, or the exposed mechanics has nothing to do with winning or game difficulty. So in my results I have unticked some boxes as mere impressions... and I found things that were past the disturbing threshold. But in most cases what i  found was lacking what is needed to support the main claim.

 

What I took out from these tests is that the game easily starts to make oneself paranoid. And I think a game or movie shaking ur psycho should immediately receive a good rating. Maximal, i mean. . And that if the game is rigged, it is not in the sense or with the tools that most people are willing to think of. 

 

And I agree with above, rigged might not be as good of a word as “weighted” is. Using “weight” might get us in a more conclusive debate, but I have only faint hopes of being able to have the word rigged abandoned lol... maybe I should start using it myself ? :)

 

now to the people who think unicums are artificially pushed up, while their demise isn’t due to a lack of skills and understanding of the game ... <<<<< delusional award.      W o r k   H a r d e r   A n d   G o o d   L u c k .


Edited by minitel_NA, 15 February 2018 - 06:44 AM.

Minitelrose visiting from EU, occasional player/forumer in the NA

 


NoobieDude #22 Posted 15 February 2018 - 06:43 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View Postreluctanttheist, on 14 February 2018 - 10:34 PM, said:

 

 ​So I guess you live in the woods and sit on a log?  Really dude.  I call that a case of chronic progressive rectal-cranial inversion

 

I'm not an English teacher or grammarian or anything, but to me, the difference is that "intelligent" would refer to someone (or something) that is smart and able to figure things out.  "Intellectual", on the other hand, would refer to someone who likes to pontificate spew from there mouths...you do you boo.. 

M00niac #23 Posted 15 February 2018 - 06:44 AM

    Senior Sergeant

  • Players
  • 46099 battles
  • 623
  • Member since:
    08-08-2014

Mudkip,

ive answered this many, many times. 

 

The “game” is not rigged.  Individual matches are rigged. The game is not rigged if in match 1 we help you, and in match 2 we help Reluctant. 

 

As for speculation, you are correct. Many people are speculating against all evidence, that MM is nothing but random assignments. I’m unsure why they do so, but they often resort to name calling when you challenge their “truth”. :)



NoobieDude #24 Posted 15 February 2018 - 06:51 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View PostM00niac, on 14 February 2018 - 10:44 PM, said:

Mudkip,

ive answered this many, many times. 

 

The “game” is not rigged.  Individual matches are rigged. The game is not rigged if in match 1 we help you, and in match 2 we help Reluctant. 

 

As for speculation, you are correct. Many people are speculating against all evidence, that MM is nothing but random assignments. I’m unsure why they do so, but they often resort to name calling when you challenge their “truth”. :)

like the above post of mine..sorry bout that...but man that dude is a trip.



reluctanttheist #25 Posted 15 February 2018 - 06:54 AM

    Canuck Didactics

  • Players
  • 27135 battles
  • 6,152
  • [III-H]
  • Member since:
    01-12-2015

View PostNoobieDude, on 14 February 2018 - 10:43 PM, said:

I'm not an English teacher or grammarian or anything, but to me, the difference is that "intelligent" would refer to someone (or something) that is smart and able to figure things out.  

 

Perhaps one day you will figure out a dictionary.  (Heck, even Grammarly for your browser can compensate for still using rocks as tools!)

 

View Postminitel_NA, on 14 February 2018 - 10:39 PM, said:

here is my bet : you have an oversimplified view of the situation.

 

If MM doesn't use player skill, then the recipe it uses is of less concern.  Other than tier X games, MM still has to decide if I'm top-tier or low-tier.  And it doesn't keep any of us top-tier all the time.  If MM remembers that I was top-tier in the last three games, will it make sure I'm low tier in the next game?  Or is tier assignment a mix of tier memory and how many tanks are in the queue in the tiers above and below my tank?  Again, that's not a weighting against me, it's just a form of balancing.  Let's say you have a poor player - MM could try to make them high-tier all the time to give them a chance to win more often.  And I don't think we have any proof of that either.

Tanks:  _X: T110E5, T110E3, FV215b(183), IS-7, Obj.140  _IX: M103, T-54  _VIII: IS-6, T34, Lowe, T-44, IS-6, IS-3D  _VII: T-43, Comet, E25, AT-15A, SU-122-44
Usually on in the evenings Pacific time.  Intake Contact for Triarii Clan (PM for details)
Be a better player  |  Click here if you have lag  |  Graphics Settings for iOS  |  Check your ping with Pingplotter  |  Get good: watch Bushka!  |  Check out tanks on Tank Compare  and  BlitzHangar


9_V0LT #26 Posted 15 February 2018 - 07:01 AM

    Cunning Linguist

  • Players
  • 10285 battles
  • 2,939
  • [OGRE]
  • Member since:
    06-26-2017

View Postreluctanttheist, on 15 February 2018 - 05:35 AM, said:

RibbleStripe - does matchmaker analyze player skill?

 

 

Hey NSA, you guys don't listen in on my private phone calls, do you?


Yeah, I'm back. Let's not make a big deal about this.


Dan_Deerso #27 Posted 15 February 2018 - 07:11 AM

    Tankologist

  • Players
  • 28180 battles
  • 4,030
  • [III]
  • Member since:
    11-13-2013

View PostNoobieDude, on 15 February 2018 - 01:32 AM, said:

irdc what he has to say. a corporate hack is not to be trusted.

 

irdc what a failed reroll has to say on the forums, so I guess we're even.

NoobieDude #28 Posted 15 February 2018 - 07:12 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View Postreluctanttheist, on 14 February 2018 - 10:34 PM, said:

 

 ​So I guess you live in the woods and sit on a log?  Really dude.  I call that a case of chronic progressive rectal-cranial inversion

nice

View PostNoobieDude, on 14 February 2018 - 10:43 PM, said:

 

I'm not an English teacher or grammarian or anything, but to me, the difference is that "intelligent" would refer to someone (or something) that is smart and able to figure things out.  "Intellectual", on the other hand, would refer to someone who likes to pontificate spew from there mouths...you do you boo.. 

and lets put every thing  in context shall we?

View Postreluctanttheist, on 14 February 2018 - 10:54 PM, said:

 

Perhaps one day you will figure out a dictionary.  (Heck, even Grammarly for your browser can compensate for still using rocks as tools!)

                                                                                                              Grammar    fixed this for you.

you seem to have a real psychosis for grammar. are you O.C.D? you seem to just want to attack. is it because you are having trouble pontificating?

i never attacked you until provoked by you. what gives dude?

NoobieDude #29 Posted 15 February 2018 - 07:13 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View PostDan_Deerso, on 14 February 2018 - 11:11 PM, said:

 

irdc what a failed reroll has to say on the forums, so I guess we're even.

 

wow cool. believe what ya want..sad really sad.lol but thanx for the reroll  comment....cant handle civil discourse i guess..oh well

Edited by NoobieDude, 15 February 2018 - 07:17 AM.


9_V0LT #30 Posted 15 February 2018 - 07:14 AM

    Cunning Linguist

  • Players
  • 10285 battles
  • 2,939
  • [OGRE]
  • Member since:
    06-26-2017

https://patents.goog...atent/US8425330

 

"According to another aspect, the matchmaking server may store a win/loss percentage for each user (or vehicle) at a given battle level. As the player’s win/loss ratio decreases, the player becomes more likely to be placed in battles having battle levels at the lower end of the allowable range, whereas as the player’s win/loss ration increases, the player becomes more likely to be placed in battles having battle levels at the upper end of the allowable range. Thus, when a player has been repeatedly put into too many difficult battles, the balancing is done in favor of easier battle sessions, thereby encouraging the player by providing an easier game environment. Similarly, when the player has been repeatedly put into too many easy battles, the balancing is done in favor of harder battle sessions, thereby keeping the player challenged instead of letting the player become bored with easy games"

 

 


Yeah, I'm back. Let's not make a big deal about this.


Dan_Deerso #31 Posted 15 February 2018 - 07:19 AM

    Tankologist

  • Players
  • 28180 battles
  • 4,030
  • [III]
  • Member since:
    11-13-2013

View PostNoobieDude, on 15 February 2018 - 02:13 AM, said:

Wow, cool. Believe what you want. Sad, really sad. lol But thanks for the reroll comment.

 

Any time. 

 

Anyways, what RT, the rest of III, me, most other forumers, etc. are waiting for is your personally signed data set that shows a trend towards rigging. It's dirt simple to do.



MrNellis #32 Posted 15 February 2018 - 07:19 AM

    Senior Sergeant

  • Players
  • 14054 battles
  • 594
  • [CHEF]
  • Member since:
    12-10-2016


NoobieDude #33 Posted 15 February 2018 - 07:20 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View Post9_V0LT, on 14 February 2018 - 11:14 PM, said:

https://patents.goog...atent/US8425330

 

"According to another aspect, the matchmaking server may store a win/loss percentage for each user (or vehicle) at a given battle level. As the player’s win/loss ratio decreases, the player becomes more likely to be placed in battles having battle levels at the lower end of the allowable range, whereas as the player’s win/loss ration increases, the player becomes more likely to be placed in battles having battle levels at the upper end of the allowable range. Thus, when a player has been repeatedly put into too many difficult battles, the balancing is done in favor of easier battle sessions, thereby encouraging the player by providing an easier game environment. Similarly, when the player has been repeatedly put into too many easy battles, the balancing is done in favor of harder battle sessions, thereby keeping the player challenged instead of letting the player become bored with easy games"

 

 

+1



reluctanttheist #34 Posted 15 February 2018 - 07:22 AM

    Canuck Didactics

  • Players
  • 27135 battles
  • 6,152
  • [III-H]
  • Member since:
    01-12-2015

Noobie, your youth is evident in your posts.  Don't be surprised that you're not being taken seriously.  Even your post above is a fail (Grammarly is a common browser add-on).  Your statements are generally indefensible.

 

"A house built on lies will fall."  A company that employs deception as a business practice - well, even its employees wouldn't be able to trust each other.  The whole thing goes down.  If WG employs skill-based MM when it says it does not, then the company cannot be trusted - internally or externally.  You call Ribble a corporate hack instead of acknowledging his competence as part of an industry-respected gaming company whose coding methodology is understood to be leading edge in the industry.  They have Youtube videos on their coding techniques.  It's wickedly boring.  It's also best in class.  Your derisive approach shows that you're too young and foolish to even comprehend the caliber the company you attempt to insult.  Your insults - and the logic you bring to the table - are so far short of the discussion on this topic that it shows that you don't even belong here.


Edited by reluctanttheist, 15 February 2018 - 07:23 AM.

Tanks:  _X: T110E5, T110E3, FV215b(183), IS-7, Obj.140  _IX: M103, T-54  _VIII: IS-6, T34, Lowe, T-44, IS-6, IS-3D  _VII: T-43, Comet, E25, AT-15A, SU-122-44
Usually on in the evenings Pacific time.  Intake Contact for Triarii Clan (PM for details)
Be a better player  |  Click here if you have lag  |  Graphics Settings for iOS  |  Check your ping with Pingplotter  |  Get good: watch Bushka!  |  Check out tanks on Tank Compare  and  BlitzHangar


Dan_Deerso #35 Posted 15 February 2018 - 07:24 AM

    Tankologist

  • Players
  • 28180 battles
  • 4,030
  • [III]
  • Member since:
    11-13-2013

View Post9_V0LT, on 15 February 2018 - 02:14 AM, said:

https://patents.goog...atent/US8425330

 

"According to another aspect, the matchmaking server may store a win/loss percentage for each user (or vehicle) at a given battle level. As the player’s win/loss ratio decreases, the player becomes more likely to be placed in battles having battle levels at the lower end of the allowable range, whereas as the player’s win/loss ration increases, the player becomes more likely to be placed in battles having battle levels at the upper end of the allowable range. Thus, when a player has been repeatedly put into too many difficult battles, the balancing is done in favor of easier battle sessions, thereby encouraging the player by providing an easier game environment. Similarly, when the player has been repeatedly put into too many easy battles, the balancing is done in favor of harder battle sessions, thereby keeping the player challenged instead of letting the player become bored with easy games"

 

Gee, everyone uses every single patent they file.

 

I'm sure they're also using all 67 of their other patents, too.


Edited by Dan_Deerso, 15 February 2018 - 07:26 AM.


9_V0LT #36 Posted 15 February 2018 - 07:26 AM

    Cunning Linguist

  • Players
  • 10285 battles
  • 2,939
  • [OGRE]
  • Member since:
    06-26-2017

View PostMrNellis, on 15 February 2018 - 07:19 AM, said:

 

"If you win, it's because you are awesome...if you lose, it must have been the coach or weather or..."

 

This is why we keep battle statistics.

If I do the most damage, spotting, scoring, in a battle and we win, then it's fair to say I had a hand in the win. Conversely, if my team loses the next battle and I'm still the one with the most damage, spotting and scoring, then it's also fair for me to say, my team sucked.


Yeah, I'm back. Let's not make a big deal about this.


NoobieDude #37 Posted 15 February 2018 - 07:28 AM

    First Sergeant

  • Players
  • 3191 battles
  • 1,888
  • [CYRUS]
  • Member since:
    11-29-2017

View PostDan_Deerso, on 14 February 2018 - 11:19 PM, said:

 

Any time. 

 

Anyways, what RT, the rest of III, me, most other forumers, etc. are waiting for is your personally signed data set that shows a trend towards rigging. It's dirt simple to do.

 

i thought we talked about this before.? in fact i  stood up for your opinion on MM. i didn't agree with it. but some other forumer left a nasty comment  on your post. and i think we agreed to disagree  civilly. and i have never used the word  rigged MM. ther is all kinds of data on weighted MM look it up.

minitel_NA #38 Posted 15 February 2018 - 07:31 AM

    First Sergeant

  • Players
  • 454 battles
  • 5,729
  • [EUREF]
  • Member since:
    12-19-2015

View Postreluctanttheist, on 15 February 2018 - 06:54 AM, said:

 

Perhaps one day you will figure out a dictionary.  (Heck, even Grammarly for your browser can compensate for still using rocks as tools!)

 

 

If MM doesn't use player skill, then the recipe it uses is of less concern.  Other than tier X games, MM still has to decide if I'm top-tier or low-tier.  And it doesn't keep any of us top-tier all the time.  If MM remembers that I was top-tier in the last three games, will it make sure I'm low tier in the next game?  Or is tier assignment a mix of tier memory and how many tanks are in the queue in the tiers above and below my tank?  Again, that's not a weighting against me, it's just a form of balancing.  Let's say you have a poor player - MM could try to make them high-tier all the time to give them a chance to win more often.  And I don't think we have any proof of that either.

 

Last year I was trying to grind tier9, solo, from scratch, like I used to. T95, E75, and two more, all at once. 

 

My winrate after nearly 100 battles in each of them was around 38-45%.  worst : I still didn’t reach even half of the xp needed, simply because all of the losses and 7 green dead before I can even hit enemy twice.

 

so I started some study and here is the result : there are more people with tier10 in the queue, than tier9.

I looked up these numbers for months in the queue, and in blitzstars, and it was constantly the case.

Tier9 are further split to meet tier8 and tier 10. It made your chance of meeting tier8 very slim, and meeting tier10 meant that in that game there was a lot more tier 10 than tier 9. with stock tank and crew it was impossible to carry. 

 

This wasnt due to a curse against me : the  MM was unbalanced. but that was all the server population’s fault. Too many noobs flood tier10. (I know, it should make it to the evening news)...

 

After this I went to check other tiers, and there are such unbalances over there as well, though less drastic, but still to be noted for exemple at tier5 and 6 due to supremacy on/off, etc..


Edited by minitel_NA, 15 February 2018 - 07:37 AM.

Minitelrose visiting from EU, occasional player/forumer in the NA

 


Dan_Deerso #39 Posted 15 February 2018 - 07:32 AM

    Tankologist

  • Players
  • 28180 battles
  • 4,030
  • [III]
  • Member since:
    11-13-2013

View PostNoobieDude, on 15 February 2018 - 02:28 AM, said:

 

i thought we talked about this before.? in fact i  stood up for your opinion on MM. i didn't agree with it. but some other forumer left a nasty comment  on your post. and i think we agreed to disagree  civilly. and i have never used the word  rigged MM. ther is all kinds of data on weighted MM look it up.

 

I agreed to get my data set. My data set so far shows no trend between battle tier and winstreaks, and no trend between winstreaks and losing streaks.

 

And if you can point me to a csv of weighted MM data, go right ahead.



9_V0LT #40 Posted 15 February 2018 - 07:32 AM

    Cunning Linguist

  • Players
  • 10285 battles
  • 2,939
  • [OGRE]
  • Member since:
    06-26-2017

View PostDan_Deerso, on 15 February 2018 - 07:24 AM, said:

Gee, everyone uses every single patent they file.

I'm sure they're also using all 67 of their other patents, too.

 

Maybe they do, maybe they don't.

They've never confirmed or denied it -which only encourages speculation.

I can't prove God exists and you can't prove he doesn't, 'cause God aint taking.

Without the ability to prove either position, both opinions are equally valuable and equally worthless.


Yeah, I'm back. Let's not make a big deal about this.






Also tagged with MM, Matchmaker

1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users